[PDF] CHAPTER 1 METALLIC CONSTRUCTION SYSTEMS: YESTERDAY AND TODAY

1 2 Source: METAL BUILDING SYSTEMS CHAPTER 1 METAL BUILDING SYSTEMS: YESTERDAY AND NOW 1.1 THE ORIGINS What's in a name...

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 1

METALLIC BUILDING SYSTEMS: YESTERDAY AND TODAY

1.1 THE ORIGIN 1.1.1 What is in a name? Some readers may not be clear on the exact topic of our discussion. In fact, even some design professionals tend to get confused by the term metal construction system. “Are we talking about a steel structure? What is this building? Is it a modular building? Or prefabricated? Or maybe paneled? Is it the same as a prefabricated building?” – you hear a lot. While all of these terms refer to some type of structure designed and partially assembled by the manufacturer in the workshop, they refer to entirely different concepts. Before proceeding, the distinctions must be resolved. Modular buildings are made up of three-dimensional segments that are fabricated into blueprints and shipped to a site for assembly and final assembly by a field worker. One of the most popular materials for modular buildings is wood, and these factory-made units are common in residential construction. Another common application is precast concrete formed into stackable modular cells that are fully pre-wired and prefabricated. These modules consist of four walls and a ceiling that also serves as the floor for the upper unit. In the 1960s and 1970s, modular steel systems consisting of three-dimensional columns and beam modules bolted together on site were commercialized with limited success. However, modern metal construction systems cannot be called modular. Panel systems include two-dimensional components such as wall, floor and roof sections that are manufactured at the factory and assembled on site. In addition to “traditional” precast concrete, modern exterior wall panels can be made from materials such as metals, brick, stone and composites known as EIFS (Exterior Insulation and Finish System). While the outer "skins" of metal buildings often use panels, the term panels does not capture the essence of metal building systems and should not be used to describe them. Prefab buildings are crafted and essentially assembled at the factory. While metal construction has its roots in prefabricated construction, this type of construction today primarily includes small structures that are transported to the construction site in one piece, such as toll booths, kiosks and house sheds. Modern metal buildings are not prefabricated in this sense. As we shall see, changes in terminology come along with the evolution of the industry itself.

1.1.2 The first metal buildings The first iron framed building was the Ditherington Flax Mill built in 1796 at Shrewsbury, England.1 In a calico mill built 3 years ago near Derby, cast iron columns were replaced by the usual wooden columns. These iron experiments were triggered by the frequent devastating fires in British cotton mills. After the metal's refractory properties were demonstrated in buildings, wrought and cast iron structures gradually became commonplace. 1 Retrieved from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

METALLIC BUILDING SYSTEMS: YESTERDAY AND TODAY 2

CHAPTER ONE

Eventually, in the mid-19th century, experimentation with rolled iron beams culminated in the construction of the Cooper Union Building in New York City, the first building to use hot-rolled steel beams. In 1889, the Rand McNally Building in Chicago became the first skyscraper entirely framed in steel.2 Around the same time, the first prefabricated metal buildings appeared. As early as the mid-19th century, "portable iron houses" were being marketed by Peter Naylor, a New York metal roofer, to meet the housing needs of the fortune hunters of the 1848 California Gold Rush. At least several hundred of these structures have been sold. A typical iron house measured 15 feet by 20 feet and, according to the advertisements, could be assembled by a single man in less than a day. Naylor's advertisements claimed his structures were cheaper than log homes, fireproof, and more comfortable than tents.1 Eventually, of course, California's lumber industry caught on, and Naylor's invention lost its appeal. In the first two decades of the 20th century, prefabricated metal components were mainly used for garages. Founded in 1901, Butler Manufacturing Company developed its first prefabricated building in 1909 to provide garage space for the ubiquitous Model T. This curved-top building used timber frames covered with corrugated iron. Eventually, to improve the fire resistance of its buildings, the company switched to all-metal construction with frames made of corrugated, bent steel sheets. The domed design, inspired by cylindrical grain elevators, influenced many other prefabricated metal buildings.3 In 1917, the Austin Company of Cleveland, Ohio began marketing 10 standard designs of factory building to choose from a catalog. The framework of these early metal structures consisted of previously designed and detailed steel supports and trusses. The Austin buildings were true precursors to what would later come to be known as prefabricated construction, a new concept that allowed material to be shipped several weeks in advance, eliminating the need for post-sale design time. Austin sold its buildings through a newly formed network of district sales offices.4 In the early 1920's, the Liberty Steel Products Company of Chicago offered a prefabricated factory building that could be erected quickly. LIBCO's ad featured the building and boasted, “10 men built this building in 20 hours. Just ordinary help, and the only tools needed were wrenches...”1 Steel was then an established competitor to other building materials. The first edition of the Standard Specification for the Design, Fabrication and Erection of Structural Steel for Buildings was published in 1923 by the newly formed American Institute of Steel Construction. In the 1920s and 1930s several steel fabrication companies were formed to meet the needs of the oil industry, constructing buildings to store equipment; Some of these companies also produced farm buildings. For example, Star Building Systems was founded in 1927 to meet the needs of oil drillers during the Oklahoma oil boom. These early metal buildings were fairly small—8 feet by 10 feet, or 12 feet by 14 feet in plan—and framed by trusses spanned between lattice columns. The wall panels, typically 8 feet by 12 feet and extending vertically, were made of sections of corrugated galvanized iron bounded by riveted steel angles. 1.1.3 The war years and beyond During World War II, larger versions of these metal buildings were used as aircraft hangars. Its columns were made of interlocking angles, perhaps 6 by 4 by 3Ⲑ8 in cross section, and the roof structure was of arched trusses. Military manuals were typically used for design criteria. Unlike their predecessors, these buildings relied on intermediate beams for lateral support. The most famous prefabricated building during World War II was Quonset Hut, which became a household word. Quonset huts were mass-produced in the hundreds of thousands to meet the need for cheap, standardized housing (Fig. 1.1). These structures required no special skills, were easily assembled with hand tools, and—with little effort—could be easily disassembled, moved, and reassembled elsewhere. The primary manufacturer of Quonset huts was Stran-Steel Corp., a pioneering metal fabrication company that went on to develop many "firsts". Quonset huts followed soldiers wherever they went, testament to the legendary advantages of American mass production. Yet these utterly utilitarian, simple, and uninspiring structures were widely perceived as cheap and ugly. This impression is still in the memory of many, although some Quonset huts have survived for over half a century.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

METAL CONSTRUCTION SYSTEMS: YESTERDAY AND TODAY METAL CONSTRUCTION SYSTEMS: YESTERDAY AND TODAY

FIGURE 1.1 Quonset Hut, Quonset Point, R.I. (Photo: David Nacci.)

The negative connotation of the term prefabricated building increased after the end of the war and the next generation of metal buildings emerged. Like the Quonset smelter, this new generation filled a specific need: the post-war economic boom required more factory space to meet pent-up demand for consumer goods. The huge, well-organized and efficient sheet metal industry had just lost its biggest customer - the military. Could the earlier prefabricated sheet metal buildings and the Quonset shack, as well as the legendary Liberty Ship, which was quickly mass-produced at the Kaiser shipyard in California, be a lesson in the rapid manufacture of factory buildings? The answer was clear: "Yes!" With the new generation of sheet metal clad buildings, fast and inexpensive construction was once again the priority, not aesthetics. After all, it was the contents of these early metal structures that mattered, not the design of the building. With patterned sheet metal sides and roof supported by triangular steel trusses and columns - a roof pitch of 4:12 was common - the required building volumes could be created relatively quickly. In this galvanized and corrugated environment, windows, insulation and extensive mechanical systems were seen as unnecessary embellishments. The sheer number of these prefabricated buildings, cloned in the less imaginative spirit of mass production, was overwhelming. Eventually the economic boom subsided, but the buildings stayed. Its simple appearance has never been an advantage. As time passed and these buildings weathered, they conveyed an image of weathered and out of place. After all, prefabricated buildings were frowned upon by almost everyone. The cheap, inferior impression that characterized the Quonset smelter was powerfully reinforced by "booming factories". This double whammy has removed the seriousness of “prefab” and possibly forever burdened the term with negative connotations. The metal construction industry has understood the problem. I looked for another name. 1.1.4 Pre-Engineered Buildings The scientifically based term pre-engineered buildings originated in the 1960s. Buildings were "pre-engineered" because, like their ancestors, they relied on standard construction for a limited number of pre-engineered configurations.

METALLIC BUILDING SYSTEMS: YESTERDAY AND TODAY 4

CHAPTER ONE

Several factors made this period significant in the history of metal construction. First, improved technology was constantly expanding the maximum free spans of metal buildings. The first rigid frame buildings, introduced in the late 1940s, could only span 40 feet. Within a few years, 50, 60, and 70 foot buildings became possible. By the late 1950s, rigid structures with spans of 100 feet were being manufactured.5 Second, ribbed metal panels became available in the late 1950s, allowing buildings to look different from the old corrugated look and tired. Third, colored plates were obtained from Stran-Steel Corp. introduced. in the early 1960s, which allowed for a certain design individuality. Around the same time, continuous-span, cold-formed Z-purlins were invented (also by Stran-Steel), the first factory-insulated panels were developed by Butler, and the first UL-approved metal roof appeared.1 And, last but not least, the first am Computer designed metal buildings began to appear in the early 1960's and with the advent of computerization the design possibilities became almost limitless. All of these factors led to a new metal building boom in the late 1950s and early 1960s. As long as the buyer could confine themselves to standard designs, the buildings could be said to be prefabricated. As the industry began offering custom metal buildings to meet the specific needs of each customer, the name prefabricated building became something of a misnomer. In addition, this term was uncomfortably close and easily confused with the conservative prefabricated buildings with which the new industry did not want to be associated. Although the term prefabricated building is still widely used and will appear frequently in this book, the industry now prefers to call its products metal building systems.

1.2 METALLIC CONSTRUCTION SYSTEMS Why "Systems"? Is this just another use of cyber language used indiscriminately to describe anything that has more than one component? Nowadays, even the words paint system or floor cleaning system no longer elicit a smile. In fairness, the metal construction system fits the classical definition of a system as an interdependent group of elements forming a unified whole. In a modern metal building, components such as walls, roof, main and substructure, and bracing are designed to work together. A typical arrangement of a metal structure is shown in Fig. 1.2. In addition to a brief discussion at this point, Chap. 3. A building's first line of defense against the weather is its wall and roof materials. These elements also withstand structural loads such as wind and snow and transfer loads to the supporting secondary structure. The secondary structure - wall joists and roof purlins - takes the loads from the wall and roof covering and distributes them to the main building structures, giving them a valuable lateral boundary along the way. The main structures, consisting of columns and rafters, carry snow, wind and other loads to the building's foundations. Wall and roof supports give stability to the entire building. Even the fasteners are chosen to be compatible with the materials to be fastened and are designed by the manufacturers. The system approach is therefore clearly recognizable. The term metal construction system is apt and well deserved. Over time, it will no doubt replace the still-used name "prefabricated buildings."

1.3 SOME STATISTICS Today, metal building systems dominate the slow-growing non-residential market. According to the Metal Building Manufacturers Association (MBMA), in 1995, prefabricated structures accounted for 65% of all new one- and two-story buildings up to 150,000 square feet. 355 million square feet of space was created. Large industrial buildings over 150,000 square feet added another 34.3 million square feet of new space.6 Revenue in 2000 was $2.5 billion.

METAL CONSTRUCTION SYSTEMS: YESTERDAY AND TODAY METAL CONSTRUCTION SYSTEMS: YESTERDAY AND TODAY

FIGURE 1.2 Typical components of a metal construction system. (VP building.)

Metal building systems serve many applications. Commercial uses have historically accounted for 30 to 40 percent of metal building sales. This category includes not only the well-known beige warehouses (Fig. 1.3), but also office buildings, garages, supermarkets and retail stores (Figs. 1.4, 1.5 and 1.6). Another 30 to 40 percent of metal structural systems are found in manufacturing applications - factories, material recycling plants, automotive and chemical industries (Figs. 1.7, 1.8 and 1.9). About 10 to 15 percent of prefabricated buildings are used for community purposes: schools, town halls and even churches (Figs. 1.10 and 1.11). Extensive "other" use includes everything else, particularly farm buildings such as grain stores, farm machinery, sheds, storage buildings, and livestock sheds.

1.4 THE BENEFITS OF METALLIC BUILDING SYSTEMS Most metal structures are purchased by the private sector, which seems to appreciate the benefits of proprietary prefabricated structures more than public entities. What are these benefits? ●

●

Ability to travel long distances. There are not many other types of triangular structures that can economically span 100 feet or more. The competition consists primarily of trusses, which require significant design and manufacturing time. (Special tensioned fabrics can also cover the distance, but are in a class of their own.) Faster occupancy. Anyone who has ever attempted to assemble a piece of furniture remembers the frustration and time it took to understand the various components and assembly methodology. The second time, the process is much faster. A similar situation occurs at a construction site when erecting a structure made of sticks. The first time takes a little longer...

METALLIC BUILDING SYSTEMS: YESTERDAY AND TODAY 6

CHAPTER ONE

FIGURE 1.3 A familiar beige metal warehouse. (Photo: Maguire Group, Inc.)

FIGURE 1.4 Metal construction system in a commercial application. (Photo: Bob Cary Construction.)

●

but there is no second time to use the learning curve. However, with ready-made standard components, an experienced fitter is always on familiar ground and very efficient. According to estimates, the use of metal construction systems can save up to a third of the construction time. This time is definitely hard cash, especially for private customers, who can achieve significant savings simply by shortening the term of expensive construction financing. It is not uncommon for small metal fabrication projects (approximately 10,000 square feet) to be completed in 3 months. At this time, many structures built on stilts rise above the ground. cost efficiency. In a true systems approach, well-matched pre-engineered components are assembled by one or just a few builders; faster assembly means cheaper fieldwork. In addition, each structural element is designed for near-complete efficiency, minimizing material waste. Less labor and less material means lower costs. Estimates of this cost-efficiency vary, but it is generally accepted that prefabricated buildings are 10-20% cheaper than traditional buildings. However, as in Chap. 3, some carefully designed stick-built structures can successfully compete with metal building systems.

METAL CONSTRUCTION SYSTEMS: YESTERDAY AND TODAY METAL CONSTRUCTION SYSTEMS: YESTERDAY AND TODAY

FIGURE 1.5 Prefabricated office building. (HCI Structural Steel Systems, Inc.)

FIGURE 1.6 Car dealership installed in a metal building. (Photo: Metallbausysteme.)

METALLIC BUILDING SYSTEMS: YESTERDAY AND TODAY 8

CHAPTER ONE

FIGURE 1.7 A modern manufacturing facility enabled by metal construction systems. (Photo: Bob Cary Construction.)

●

expansion flexibility. Metal buildings are relatively easy to extend by stretching, which requires dismantling bolted connections in the front wall, removing the wall, and installing additional tensioning structure in its place. The removed bulkhead molding can often be reused at the new location. Matching roof and wall panels are then added to complete the expanded building envelope. Low maintenance. A typical metal building system with prefabricated metal panels and vertical seam roof is easy to maintain: metal surfaces are easy to clean and modern metal surfaces offer excellent resistance to corrosion, fading and discoloration. Some of the permanent surface treatments available on the market today are discussed in Chap. 6. Responsibility from a single source. Having one person responsible for the entire building envelope is one of the key benefits of metallic building systems. At least in theory, everything is compatible and well thought out. The building owner or construction manager does not have to keep track of many different suppliers or worry that one of them will fail mid-construction. Owners of small occupied buildings especially appreciate the convenience of dealing with an entity if something goes wrong during occupancy. This convenience is an important selling point of the systems.

METAL CONSTRUCTION SYSTEMS: YESTERDAY AND TODAY METAL CONSTRUCTION SYSTEMS: YESTERDAY AND TODAY

FIGURE 1.8 Material recycling plants generally use metal construction systems. (Photo: Maguire Group Inc.)

FIGURE 1.9 A large production facility in a prefabricated building. (Photo: Varco Pruden Building.)

METALLIC BUILDING SYSTEMS: YESTERDAY AND TODAY 10

CHAPTER ONE

FIGURE 1.10 A church in a prefabricated building.

FIGURE 1.11 This community building uses a metal construction system. (Photo: Metallbausysteme.)

METAL CONSTRUCTION SYSTEMS: YESTERDAY AND TODAY METAL CONSTRUCTION SYSTEMS: YESTERDAY AND TODAY

1.5 SOME DISADVANTAGES OF METAL STRUCTURES An objective look at the industry cannot be complete without mentioning some of its disadvantages. As with any type of construction, metallic construction systems have a negative side that needs to be clearly understood and anticipated to avoid unwanted surprises. ●

●

Variable build quality. Most people familiar with prefabricated buildings will no doubt have noticed that not all manufacturers and their builders are the same. Large manufacturers typically belong to a trade association or certification program that promotes certain design and manufacturing quality standards. Some other providers may not accept the same restrictions and occasionally provide less than adequate or inferior buildings. In fact, a frame can be assembled using separately purchased metal components, but without any engineering - or much thought -. These pseudo-prefab buildings are prone to failure and give the industry a bad name. Therefore, it is important to know how to specify a specific level of performance, rather than assuming that every manufacturer will deliver the desired quality for the project. Lack of reserve power. The downside to the metal fabrication industry's legendary efficiency is the difficulty in adapting existing prefabricated buildings to new loading requirements. With every ounce of “excess” metal removed to make the structure as economical as possible, any future load changes must be approached with extreme care. Even the relatively small additional weight imposed by a modest rooftop air conditioner or light monorail can theoretically push the designed structure “to the limit” unless structural changes are considered. Possible ignorance of the manufacturer with local codes. When a metal building is shipped from a distant part of the country, its fabricator may not be as familiar with the nuances of applicable building codes as a local contractor. While most large manufacturers maintain a library of national and local building codes and train their dealers to communicate local codes to them, some smaller operators may not. Homeowners must ensure that the building they purchase conforms in every respect to applicable building codes, a task that requires some knowledge of the regulations and manufacturing practices. (Of course, some local codes may be based on obsolete model code changes.)

The many advantages of metal building systems clearly outweigh some shortcomings, a fact that explains the systems' popularity. Still, the specification of prefabricated buildings is not an easy process; contains many potential pitfalls for the unwary. Some of them are described in this book.

LITERATUR 1. „MBMA: 35 Years of Leading the Industry“, eine Sammlung von Artikeln, Metal Construction News, Juli 1991. 2. Leslie H. Gillette, The First 60 Years: The American Institute of Steel Construction, Inc. 1921–1980, AISC, Chicago, IL, 1980. 3. Design/Specifiers Manual, Butler Manufacturing Company, Roof Division, Kansas City, MO, 1995. 4. Metal Building Systems, 2. Aufl., Building Systems Institute, Inc., Cleveland, OH, 1990. 5. "Ceco Celebrates 50th Year...", Metal Construction News, Februar 1996. 6. "1995 Metal Building Systems Sales Top 2 Mrd. $ in Low Construction", Metal Construction News, Juli 1996.

METALLIC BUILDING SYSTEMS: YESTERDAY AND TODAY 12

CHAPTER ONE

REVIEW QUESTIONS 1. What are the two types of building occupancies currently used most commonly in metal building systems in the United States? What were the first uses? 2 When was the first prefabricated house built? From which company? 3 Which company started offering standard plans for factory buildings? 4 Name the best-known prefab building of the Second World War. 5 What factors enabled the transition from prefabricated buildings to metallic building systems? 6 What are the advantages of metallic construction systems?

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 2

INDUSTRY GROUPS, PUBLICATIONS AND WEBSITES

2.1 INTRODUCTION Metal building systems dominate the market for low-rise non-residential buildings, as shown in Chap. 1. For dozens, if not hundreds, of manufacturers offering proprietary frame, profile, and material systems, competing on an equal footing is impossible without a common thread of standardization and consistency. This unifying role has traditionally been performed by industry associations and trade associations. A designer who is serious about specifying metalwork systems and wants to become familiar with common industry practices will at some point need to review the design manuals and specifications issued by these groups. The planner may even wish to keep up with the latest industry developments by becoming a member of the trade association or subscribing to some of its publications. Sooner or later, the designer will be faced with a question about the availability of a particular metal component or the feasibility of a non-traditional design approach—questions that only an industry representative can answer. In addition, should disagreements arise during construction, the manufacturer and contractor would likely consult the designers' literature to substantiate their position. For all of these reasons, it is important to become familiar with industry groups and their publications. While our list is not exhaustive - new organizations are formed every year in this dynamic industry - it should be helpful to anyone looking for more information on metal building systems.

2.2 METAL BUILDING MANUFACTURERS ASSOCIATION (MBMA) 2.2.1 The organization In the 1950's the metal building industry was still unorganized and faced with a range of problems ranging from building codes and insurance restrictions to union disputes. The idea that the fledgling industry needed commercial organization came from Wilbur Larkin of Butler Manufacturing Co., who invited his competitors to a meeting. The MBMA was founded on October 1, 1956 with 13 founding members. Founding members were Armco Drainage, Behlen Manufacturing, Butler Manufacturing, Carew Steel, Cowin & Company, Inland Steel, Martin Steel, Metallic Building, Pascoe Steel, Soule Steel, Steelcraft Manufacturing, Stran-Steel Corp. Butler Manufacturing's Wilbur Larkin was elected the first President of the MBMA.1 The new group set out to build industry consensus on how to address common issues such as code acceptance, design practices, security and insurance. 13 Transferred from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

INDUSTRY GROUPS, PUBLICATIONS AND WEBSITES 14

CHAPTER TWO

Today, the MBMA has around 30 members representing the best-known metal fabricators and over 9,000 builders. Together, the group's members account for about 9 out of 10 metal building systems built in the country. Recently, the MBMA opened its membership to industry suppliers, who can join as Associate Members for greater access to MBMA programs and information.2 One of the primary roles played by the Association is to provide technical leadership to manufacturers Offer. Prior to the formation of MBMA, each building supplier used its own engineering assumptions and analysis methods, a situation that resulted in varying reliability of metal buildings. The development of engineering standards was one of the first steps of the new organization. In fact, the MBMA Technical Committee was formed at the association's first annual meeting on December 4, 1956. Over the years, the MBMA Director of Research and Engineering has been the industry's key technical representative. The association's technical efforts, which had become particularly intense since the 1970s, were directly responsible for increasing the technical development of manufacturers' technical departments. In its early days, the industry had to deal with a lack of building code information on the behavior of one- and two-story buildings in the wind. Since low-rise buildings were the main product of the metal structure manufacturers, further research was urgently needed. MBMA took the opportunity to sponsor a wind tunnel research program at the University of Western Ontario (UWO), Canada in 1976 in partnership with the American Iron and Steel Institute (AISI) and the Canadian Steel Industries Construction Council. 3 Wind tunnel testing has been around for decades, but this program, led by Dr. Alan G. Davenport was its first extensive application in low-rise buildings. The results of this test were included in the 1986 edition of the MBMA Handbook5 and contributed to the development of wind load regulations in ANSI 58 (now ASCE 7), the Standard Building Code and other codes around the world. Most of the research work at the UWO was carried out between 1976 and 1985. More recently, a program called Random City was run at the UWO Boundary Layer Wind Tunnel Laboratory. Random City is a miniature model of a typical industrial city exposed to a hurricane; the goal is to measure the wind forces acting on a typical low-rise building.4 Similar studies are being conducted at Clemson University, where roof panels with vertical seams are tested for dynamic wind forces, and at Mississippi State University, the site of the experiment. Load simulation by electromagnets. Snow load research also receives some of the MBMA's attention. For example, the effects of an uneven snow load on front structures are being investigated at the Rensselaer Polytechnic Institute under the leadership of the MBMA. Another area of MBMA-funded research concerns the thermal effects that solar radiation produces on metal roofs and wall assemblies. Because metal building systems are not inherently fireproof, the establishment of UL Listed assemblies with system components is critical to industry acceptance by building officials. The MBMA facilitated progress on this front by sponsoring the fire resistance testing of tapered steel columns and steel roof structures. Another important MBMA publication came in 2000. The Metal Roof System Design Manual marked the culmination of a successful program developed to provide designers with the best design details for different types of metal roofs. In addition to its role in developing engineering standards for low-rise buildings, the MBMA serves as the promotional arm of the metal fabrication industry. The group publishes the MBMA Fact Book and Annual Market Review and offers videos, slide shows and other promotional materials that explain the benefits of metal building systems. The association was instrumental in extending the scope of the quality certification program administered by the American Institute of Steel Construction (AISC) to steel fabricators. The program was originally intended to certify structural steel manufacturers to ensure consistently high quality throughout the production process. The new MB (Metal Building Systems) certification category applies to manufacturers of prefabricated buildings “that incorporate engineering services as an integral part of the final product”. Program objectives include evaluation of manufacturer design and quality assurance procedures and practices, certification of qualified manufacturers, periodic audits of certified companies and

INDUSTRY GROUPS, PUBLICATIONS AND WEBSITES INDUSTRY GROUPS, PUBLICATIONS AND WEBSITES

fifteen

encourage others to embrace it. Certification by the renowned agency obviously improves the image of the manufacturer and makes it easier for the local building authorities to accept his system. MBMA has made AISC certification a condition of membership. Building on the success of its AISC certification program, MBMA has developed and actively promotes its umbrella certification program. This new program aims to further improve the standards of the metal roofing industry. The Metal Building Manufacturers Association is located at 1300 Sumner Avenue, Cleveland, OH 44115-2851; his phone number is (216) 241-7333; his website is www.mbma.com.

2.2.2 The MBMA Steel Building Systems Handbook Since its first edition in 1959, the Handbook5 has been a desktop reference resource for steel fabricators and their engineers and construction workers. The amount of useful material contained in this book - and its sheer volume - has steadily increased. The 1986 edition of the handbook (then called the Low Rise Building Systems Handbook) was only about 300 pages long. The following 1996 edition changed its look from a sleek, easy-to-carry gray volume to a thick three-ring binder. The 2002 edition was released in the same easy-to-update three-ring binder format, but otherwise signaled a change in direction. The publication has been renamed Metal Building Systems Manual to sharpen its focus and improve its recognition by designers. The first section of the guide, formerly called Design Practices, is now divided into three sections: Load Application, Crane Loads, and Maintainability. Rather than presenting its own unique design methods as before, the 2002 edition provides commentary on the relevant structural provisions of the International Building Code 2000 (IBC). In the "Upload application" area, there are now extensive design examples that illustrate the design process. Instead of providing its own load combinations, the manual now refers the reader to those of the IBC. Another important part of the MBMA Low Rise Building Systems Manual, Common Industry Practices, covers a variety of topics dealing with the sale, design, manufacture, supply and installation of metal building systems, as well as some insurance and legal issues. Metalwork designers should pay close attention to Section 2, “Selling a Metalwork System,” which details what parts and accessories are included in a standard metalwork system package and which are typically excluded. The next section of the handbook, Guide Specifications, is intended to provide guidance for creating contract specifications. The handbook also includes an overview of the AISC-MB certification regulations, a commentary on wind loads, representative fire ratings, US state load data, a glossary, an appendix, and the bibliography. It is important to remember that although the manual is widely used and respected, the information it contains is presented from the manufacturer's point of view and is primarily intended to guide them. The Handbook is not a building code with legally binding provisions; it is a commercial document and its use is voluntary. As with other similar business documents, the “General Industry Practices” can be modified by project-specific contract language as needed.

2.3 AMERICAN IRON AND STEEL INSTITUTE (AISI) The American Iron and Steel Institute emerged from the American Iron Association founded in 1855. Over the years, the Institute has occasionally been instrumental in the development of codes and design standards for a variety of member steel structures, intersecting with the American Institute of Steel Construction (AISC). In order to avoid duplication of work, the two institutes have agreed to split the applicability of their standards. Currently, the AISC manual covers the design of hot-rolled steel members, which include the well-known wide-flanged beams, angles, and channels. These elements are cast and rolled to their final cross-sectional dimensions in steel mills at elevated temperatures. The AISC manual also covers plate girders fabricated from plate thicknesses generally greater than 3Ⲑ16 inches.

INDUSTRY GROUPS, PUBLICATIONS AND WEBSITES 16

CHAPTER TWO

In contrast, steel components manufactured without the application of heat or cold formed belong to the AISI range. Today, AISI is a recognized authority on cold-formed construction. Cold-formed structures are made from steel sheet, plate, or flat bar by bending, rolling, or pressing, and are generally limited to thinner materials. Some examples of cold-formed shapes used in construction include metal decks, siding, steel beams, beams and purlins - the "meat and bones" of prefabricated buildings. These structural members are typically less than 3Ⲑ16 inches thick and are referred to as "lightweight" frames. Most components of a typical steel construction system, such as B. Secondary beams and wall and roof cladding are likely to be subject to AISI regulations; the main metallic structures, according to AISC specifications. While the AISC 6 manual can be found on most civil engineers' bookshelves, the AISI manual is less well known, perhaps because cold-formed structures have traditionally been designed outside of consulting engineering firms. In fact, most consulting engineers deal primarily with bar-built structures using common hot-rolled elements. At the heart of the AISI manual is what was formerly known as the specification for the design of cold-formed steel members. Project. The original specification was largely derived from AISI-funded research at Cornell University led by Dr. George Winter and other institutions in the late 1930s and early 1940s Development of the current code is in the hands of the Building Codes and Standards Committee. In 2002 the name of the specification was changed to the North American Specification for the Design of Cold-Formed Structural Members. Consequently, it applies to the design of cold-formed structures in the United States, Canada and Mexico. Design rules common to the three countries are included in the body of the specification; country-specific points are included in the appendices. The specification includes design methods for various lightweight reinforced and unreinforced members, provides detailed design criteria for connections and bracing, and describes tests required for special cases. The specification's equations are used by fabricators and fabricators of prefabricated buildings, steel decks, cladding, and steel studs, and are used in numerous non-construction related applications such as steel ships and automobile bodies. Some provisions of the specification are explained in chap. 5. The AISI manual also includes a commentary, reference data and design examples that explain and illustrate the specification. In addition to publishing the handbook, AISI is involved in technical education efforts and promotional activities. The institute's network of regional engineers is ready to answer technical questions from code specifiers and officials. The AISI Construction Marketing Committee actively promotes specific areas of steel construction. An extensive marketing program by the committee, which included direct mail, presentations at building conventions and individual marketing, was largely responsible for the enormous success of the metal roofing systems. In addition, the Institute performs many other tasks, such as representing the entire steel industry before legislators and the executive. The American Iron and Steel Institute is located at 1101 17th Street, NW, Suite 1300, Washington, DC 20036-4700; his phone number is (202) 452-7100 and his website is www.steel.org.

2.4 METAL BUILDING CONTRACTORS AND BUILDERS ASSOCIATION (MBCEA) As the name suggests, this professional group represents metal building contractors and builders. It was founded in 1968 as the Metal Building Dealers Association (MBDA); the name was later changed to System Builders Association (SBA). The latter sounded pompous but somewhat confusing, and the group's name was changed again in 2002 to better reflect its members' professions. The MBCEA offers multiple membership categories for builders, independent contractors, metal roofers, light steel fabricators, contractors and even design professionals.

INDUSTRY GROUPS, PUBLICATIONS AND WEBSITES INDUSTRY GROUPS, PUBLICATIONS AND WEBSITES

Many MBCEA activities take place in local chapters, where participants gather in the evenings during the day to discuss common challenges and share information. At the national level, the MBCEA offers contractors and builders of metal buildings legal assistance in contractual matters, encumbrances, billing issues and the like. It also publishes several standard legal forms, such as B. a standard contract form between contractor and customer, a subcontractor contract and a contract proposal. The MBCEA maintains a certification program that is awarded to companies deemed to have extensive knowledge and experience in the metal fabrication industry and as a demonstration of honesty and integrity. The association founded the Metallbauinstitut (MBI) as an independent non-profit educational and training organization. MBCEA sponsors annual trade shows, conferences, seminars and social events and publishes a magazine for potential clients. The address of the Metal Building Contractors & Erectors Association is 28 Lowry Drive, P.O. Box 117, West Milton, OH 45383-0117; his phone number is (800) 866-6722 and his website is www.mbcea.com.

2.5 ASSOCIATION OF NORTH AMERICAN INSULATION MANUFACTURERS (NAIMA) The Association of North American Insulation Manufacturers represents the leading manufacturers of fiberglass, rock wool and slag wool insulation. NAIMA, which has its roots in one of its predecessor organizations founded in 1933, seeks to spread information about the proper use, performance and safety of insulation products. Like other similar trade groups, NAIMA conducts technical education and promotional affairs. Because the group's interests extend far beyond metallic building systems, it is the NAIMA Metallic Construction Committee that sets performance standards and establishes testing programs for insulation products used in prefabricated buildings. Some of the most valuable NAIMA Steel Structures publications are: ● ● ●

Understanding Insulation for Metal Buildings ASHRAE 90.1 Compliance for Metal Buildings NAIMA Standard 202

The North American Insulation Manufacturers Association is located at 44 Canal Center Plaza, Suite 310, Alexandria, VA 22314; his phone number is (703) 684-0084 and his website is www.naima.com.

2.6 METAL CONSTRUCTION ASSOCIATION (MCA) Established in 1983, the MCA was primarily formed to promote the wider use of metal in construction.3 The MCA's best-known contribution to this goal is the annual Metalcon International, a major trade show that brings together the entire Metal construction industry represented around the world. MCA has its own Merit Awards program that recognizes projects it deems notable, publishes a newsletter and conducts market research. MCA's market research activities include collecting and disseminating information on emerging and growing market segments and promising new applications for metallic components. The Group's annual Metal Roofing and Wall Panel Survey tracks metal panel usage by installed weight and square footage. To discuss specific areas of interest for just a select few of its members, MCA sponsors its Industry Councils - Lightweight Framing, Building Finishes and Architectural Products/Metal Roofing & Cladding. Membership is open to any person or company involved in the manufacture, engineering, sale or installation of metal components. The Metal Construction Association is located at 4700 W. Lake Avenue, Glenview, IL, 60025; his phone number is (847) 375-4718 and his website is www.metalconstruction.org.

INDUSTRY GROUPS, PUBLICATIONS AND WEBSITES 18

CHAPTER TWO

2.7 NATIONAL ROOFING CONTRACTORS ASSOCIATION (NRCA) The members of this centuries-old organization are mainly roofers, but also roof manufacturers, suppliers, consultants and designers. The NRCA offers a variety of educational programs, tests and evaluations of new and existing roofing materials, and disseminates technical information to its members. Rather than developing its own design standards or performance requirements, the NRCA prefers to support other standards-writing bodies. Of particular interest to designers of metalwork systems is the NRCA Roofing and Waterproofing Manual, which includes a section entitled Architectural Sheet Metal and Metal Roofing. This section provides a wealth of information about metal roofing, from the general to the very specific, including a sheet metal details section. The manual also includes a section on opting out. Another NRCA publication, Residential Steep-Slope Roofing Materials Guide, covers asphalt shingles and clay shingles. The NRCA also publishes a monthly magazine, Professional Roofing. NRCA is located at 10255 W. Higgins Road, Suite 600, Rosemont, IL 60018-5607; his phone number is (708) 299-9070 and his website is www.nrca.net.

2.8 LIGHT GAGE STRUCTURAL INSTITUTE (LGSI) Most manufacturers of metal building systems produce their own cold formed metal components, but there are also independent manufacturers of used roof purlins, eaves and wall joists in prefabricated buildings. For years, these producers felt underrepresented by existing trade organizations. As mentioned, lightweight cold-form construction is subject to the rather complex and frequently changing AISI specification. Following a 1986 revision of the specification, a number of lightweight frame manufacturers felt the need to work together to make major changes to the specification's provisions. In 1989 they founded the Light Gage Structural Institute. The main technical result of the Institute's activities was the publication of its Light Gage Structural Steel Framing System Design Handbook8, which contains tables of design properties and allowable load capacities for typical C and Z steel sections manufactured by LGSI members. This information is very valuable, as we shall see in Chap. 5. In addition to producing technical information, LGSI is committed to promoting the quality of light frame manufacturing. Member company factories receive up to four unannounced annual inspections by LGSI representatives. Inspectors check the thickness and material properties of the steel used by the manufacturer and carry out product measurements for compliance with LGSI guidelines; A special sticker is attached to each tested steel beam. The Light Gage Structural Institute may write to P.O. PO Box 38217, Houston, TX 77238; her phone number is (713) 445-8555 and her website is www.loseke.com/lgsi.html.

2.9 COLD FORMED STEEL STRUCTURES CENTER (CCFSS) The CCFSS was established in 1990 through an initial grant from AISI to provide a coordinated way to address research and education efforts in cold formed steel structures. The purpose of CCFSS is to bring together the technical resources of academia, product manufacturers, consultants and government agencies and advance the theory and practice of designing with cold formed steel. The center is physically located and administered by the faculty of the University of Missouri-Rolla, an institution at the forefront of research in the field. Of primary interest to metalwork system specifiers is the Center's website, which contains helpful links to the Center's sponsors such as AISI (including their specifications and standards), MBMA and MCA. There are other useful links to a list of computer programs for designing lightweight structures, the continuing education and seminar schedule, and research publications on cold formed steel.

INDUSTRY GROUPS, PUBLICATIONS AND WEBSITES INDUSTRY GROUPS, PUBLICATIONS AND WEBSITES

The Cold Formed Steel Structures Center is located at Butler-Carlton Civil Engineering Hall, University of Missouri-Rolla, Rolla, MO 65409-0030. Their website is www.umr.edu/~ccfss/.

2.10 MODERN TRADE COMMUNICATIONS INC. Modern Trade Communications is best known for publishing three magazines targeting different segments of the metal fabrication industry: ● ●

●

Metal Architecture, of interest to architects and other designers of metal building systems Metal Construction News (formerly Metal Building News), the industry's premier tabloid magazine primarily aimed at builders, fabricators and suppliers Metal Home Digest, dedicated to metal building systems for residential buildings

These three publications, Metal Architecture in particular, should be invaluable to anyone interested in keeping up to date with the latest developments in the industry. Modern Trade Communications Inc. is located at 109 Portage Street, Woodville, OH 43469; her phone number is (419) 849-3109 and her website is www.moderntrade.com.

REFERENCES 1st 2nd 3rd 4th 5th

„MBMA: 35 Years Leading the Industry“, eine Sammlung von Artikeln, Metal Construction News, Juli 1991. „MBMA öffnet die Mitgliedschaft für Industrielieferanten“, Metal Construction News, Februar 1996. „Industry Associations Playing an Important Role… “ Metal Architecture, 1994. „MBMA Research Impacts Building Codes and Standards“, Metal Architecture, Mai 1993. Metal Building Systems Manual, ehemals Low Rise Building Systems Manual, Metal Building Manufacturers Association, Inc., Cleveland, OH, 2002. 6 Steel Construction Handbook, zulässig Stress Design, American Institute of Steel Construction, Inc., Chicago, IL, 1989. 7. Specification for the Design of Cold Formed Steel Structural Members, American Institute of Iron and Steel, Washington, DC, 1986, mit Nachtrag von 1989. 8. Light Gage Structural Steel Frame System Design Manual, LGSI, Plano, TX, 1998.

REVIEW QUESTIONS 1. Which professional organization represents manufacturers of metal building systems? 2 When was the MBMA founded? 3 Name the authoritative design specification that covers cold-form frames. 4 List any two areas of MBMA activity. 5 Which MBMA employee serves as the de facto senior technical representative for the industry?

INDUSTRY GROUPS, PUBLICATIONS AND WEBSITES

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 3

THE BASIC

3.1 STRUCTURAL LOADS In this chapter we consider the structural foundations of metallic building systems. We begin with a brief discussion of the structural loads (or loads for short) that systems are typically required to support, methods of combining those loads, and methods of analysis. We then discuss how metal building systems work structurally and what their competition is, and examine the system selection process. Our goal is to show how and when to make an informed judgment about the suitability of prefabricated frames for a specific project.

3.1.1 Permanent loads and collateral loads Permanent loads are the weight of all permanent building materials such as roofs, frames and other structural elements. Because the continuous load is well defined and known in advance, it is given a relatively low factor of safety (load factor) in the final design. Collateral or superimposed permanent load is a specific type of permanent load that includes the weight of a material other than a permanent structure. It can account for the weight of mechanical ducts, plumbing, sprinklers, electrical work, future ceilings and roof renovations. How much do these components weigh? The MBMA manual1 suggests the following typical values: ● ● ● ●

Ceilings: 1 to 3 psf Lighting: 0.1 to 1 psf Heating, Ventilating, Air Conditioning (HVAC) ducts (office/commercial): 1 psf Sprinklers: 1.5 psf for dry systems, 3 psf for wet systems

When you add up the numbers, a commercial or industrial building with sprinklers, lighting, and mechanical piping—but no roof—could be rated for a collateral load of at least 5 psf. In theory, this 5 psf collateral load is sufficient to explain the action of most overhead pipes, lights and even small fans. In practice, however, the weight of these elements is not applied evenly and it may be necessary to specify a larger collateral load. However, manufacturers tend to reject such (in their opinion) artificially high safety loads, as discussed in Chap. 10. The device load, which corresponds to the weight of each specific device supported by the ceiling or floor, must be declared separately. For example, the weight of an HVAC roof unit in excess of 200 pounds is best represented by a concentrated downward force in the support purlin design. The device load can be "calculated" - converted into a uniform collateral load - for the main structure design.

21 Transferred from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

THE BASICS 22

CHAPTER THREE

3.1.2 Overhead Overload Overhead refers to the weight of building occupants, furniture, storage items, portable equipment and partitions (International Building Code2 lists the loads of partitions in the Surviving Loads section). Due to the fact that the payload is relatively short term and not easily predictable or quantifiable, it introduces large safety factors (actually uncertainty) into the final design methods. Other sources of congestion arise during construction, repair or maintenance of buildings and are even more difficult to predict and quantify. To deal with this uncertainty, building codes have enacted conservative values for moving loads – the structure must be designed to withstand loads that may occur only once or twice in the life of the structure, if any. For example, office buildings are typically rated for a 50 psf load, while the actual weight of all people and furniture in a typical office is unlikely to exceed 15 psf. It is very likely that design overload will occur in a relatively small area of the building at some point; The entire floor is much less likely to receive this load. To reflect this reality, building codes establish the rules governing reductions in overhead for structural members that support relatively large floor or roof areas. In single-story steelwork systems, the roof live load, essentially an addition to the roof load during construction and maintenance, is the reduced load. With reduced overhead, greater uniform loads are applied to secondary members supporting limited roof surfaces than to primary structural structures. Reduction formulas are included in the building codes. The magnitude of the live load on the roof is often compared to the snow load and the highest value used in the design.

3.1.3 Snow load The projected snow load represents the most likely weight of snow that can accumulate on the roof. In contrast to the traffic load, the snow load is independent of the building occupancy, but strongly dependent on the location. Building codes and the MBMA handbook traditionally provide ground snow load maps. Both the MBMA Handbook and the International Building Code now follow ASCE 73 for determining snow load on ground. Once determined, the magnitude of the ground snow load is typically reduced to obtain the predicted roof snow load by multiplying the ground snow load by certain coefficients. For example, ASCE 7-98 gives the following formula for determining the snow load on a flat roof: pf ⫽ 0.7 CeCt Ipg, where pf is the snow load on a flat roof, pg is the snow load on the ground, Ce and Ct are the exposure and heat factors, and I is the importance factor. These factors can be found in various tables in ASCE 7. To get the design snow load on a pitched roof, pf is multiplied by the pitch factor Cs. The main reason why the snow load on the roof is usually less than the corresponding snow load on the ground is that some of the snow is often removed from roofs by melting and wind. However, there are situations when the opposite is true: more snow can accumulate on a super-insulated and protected roof than on a warm floor. In one case, the measured weight of snow on the roof of a collapsed freezer building was found to be more than twice the legal amount—and also exceeded the weight of accumulated snow on the ground.4 Two more snowfalls, if any Snow-related factors often turn out to be as critical: snow slides and snowdrifts. Most people living in northern climates have seen snow slide off a pitched roof; this snow can slide onto an adjacent underlying roof and increase the snow load thereon. Snow from the roof falling against the walls and parapets is another familiar sight. The amount of this additional snow load depends on the roof size, wall or parapet height and other factors (Fig. 3.1). (Note that snow on the gabled roof is plotted following its slope, as any snow is bound to do, but the snow load is actually given as the horizontal load acting on the projected area of the roof.) The added weight of the slip and of Accumulated snow is highly concentrated and cannot be averaged

THE BASIC THE BASIC

FIGURE 3.1 Snow loads on buildings.

the whole roof. It follows that some elements of the roof structure have to withstand greater snow loads than others. In fact, roof areas next to high walls and parapets are often designed to withstand up to three times (and sometimes more) snow loads elsewhere. Another design condition to consider is uneven snow on gable roofs. The design requirements of different codes vary in this regard. The unbalanced snow provisions of ASCE 7-98 Section 7.6, referenced in the 2002 edition of the MBMA Handbook, specify load levels as a function of building size, roof pitch, snow load on flat and pitched roofs, and other factors. These determinations are quite complex, but probably represent a more accurate assessment of uneven snow depths that accumulate on gabled and hipped roofs. Unbalanced snow load on the roof is not to be confused with partial load. Partial loads are usually taken into account when designing continuous members such as purlins or multi-span rigid frames. A partial load is when some panels are carrying a reduced payload or snow while other panels are fully loaded. It has long been known that some structural effects, such as positive bending moments, are stronger at part load than at steady full load. Some spans of continuous members may experience stress reversals even under partial loads: flanges that would be compressed under full load may be loaded in tension, and members in less heavily loaded spans may flex upwards rather than downwards. Again, the 2002 edition of the MBMA manual references ASCE 7-98 for load determination. Section 7.5 of ASCE 7 specifies three loading conditions to be considered, applying a fully balanced snow load to some spans and a half load to the remaining spans. Actual snow load accumulation will likely not follow pure part load formulas, but neither will it occur 100% evenly. Snow depth can vary not only along the length of the building, but also from eaves to eaves, and the formulas are a useful approximation of the complex reality. In addition, the roof can be partially loaded during snow removal. Despite typical recommendations to evenly and gradually remove snow from the entire roof, it is very convenient to completely clear some areas at once - and involuntarily create a classic partial load.

3.1.4 Rain and Snow Fees These two fees were rarely used in the past, although some codes included them for a long time. Now they incorporate the International Building Code and ASCE 7 on the same basis as the other more well-known loads.

THE BASICS 24

CHAPTER THREE

The rain-on-snow surcharge, with a maximum of 5 psf, is levied on roofs with a pitch less than 1Ⲑ2 to 12, provided the snow load on the ground does not exceed 20 psf. It is intended to reflect a general condition in northern climates when a blizzard turns to rain. If the roof pitch is low, rainwater cannot drain away quickly and is caught by the snow. (The ability to avoid this load is a good reason to specify a minimum roof pitch of at least 1Ⲑ2 to 12 - and preferably greater, as discussed in Chapter 6 - rather than the very common pitch of 1Ⲑ4 to 12.) The load is in a rainfall set under a different section of the Code than the rain-on-snow surcharge and represents a different phenomenon - the weight of rainwater that can accumulate on the roof if the drainage system is clogged. This load includes the weight of "water rising at its design flow above the entrance to the secondary drainage system."2 Water weight is assumed to be 5.2 psf per inch of depth. As discussed in Chapters 5 and 11, the secondary roof structure on steel frame systems is quite flexible, and the rapid removal of rainwater is critical to its survival in heavy rain. For this reason, prefabricated buildings are typically designed with external gutters rather than internal drains. It is important to understand that the increasingly popular exterior parapets interfere with free drainage and require special measures to prevent roof damage and leakage from under the accumulated weight of water. One such step is to locate an interior gutter behind the railing.

3.1.5 Wind Charge Since we heard about the sad experience of the three little pigs, most of us have come to appreciate the destructive power of wind. Several recent hurricanes, such as Hugo, Andrew (1989) and Iniki (1992), have highlighted our vulnerability to this common natural disaster. Property damage attributed to the wind is enormous. To design windproof structures, engineers need to know how to quantify and distribute wind loads across different building elements. Unfortunately, the effects of wind on buildings are still not fully understood; Ongoing research leads to frequent revisions to building codes. Most modern building codes include maps showing wind speeds in miles per hour for various locations. Design wind speed was previously defined as the fastest wind speed measured at 33 feet above the ground and had an annual probability of return of 0.02. However, the 1995 and later editions of ASCE 7.3 define it as the maximum burst of three seconds, reflecting a new method of data collection by the National Weather Service. Using the formulas provided by the code, it is possible to convert wind speed to an equivalent pressure speed in pounds per square foot. The projected wind pressure on the entire building depending on the height and exposure category can be determined from the velocity pressure, which takes the local ground conditions into account. Hurricane damage studies show that local wall and roof failures occur most commonly near building corners and eaves. Supporting structures and roofing in these areas must be designed for significantly higher wind loads - both indoors and outdoors - than in the rest of the building. The actual formulas for such an increase vary by building code and are not reproduced here, but the basic definition of "corner overhang" areas exposed to higher wind loads is similar. Figure 3.2 illustrates the traditional approach to defining them. Winds can damage buildings in four basic ways: 1. Component damage, when part of the building fails. Some examples are a ripped off roof, ripped off side panels or broken windows. 2. Total collapse when the building collapses like a house of sticks due to lack of rigidity or adequate fixings. 3. Overturning when the building becomes whole and falls due to lack of weight and anchoring of the foundation. 4. Sliding, when the building remains whole but loses its anchorage and slides horizontally. Engineers have long considered wind as a strictly horizontal force and calculated it by multiplying the velocity pressure by the projected area of the building (Fig. 3.3a). like wind search

THE BASIC THE BASIC

FIGURE 3.2 Areas of high localized wind loads for low-rise buildings. (Actual numbers vary from code to code.)

As development progressed, often driven by the metal construction industry, a more complex picture of wind power distribution in gabled buildings was gradually recognized (Fig. 3.3b). In current thinking, wind is applied perpendicularly to all surfaces; both roof and wall pressure and suction are taken into account, as well as internal and external wind pressure. Sorting the various permutations of all these wind load components takes some practice and should be left to experienced professionals.

3.1.6 Earthquake damage Earthquake damage is in the headlines; Though not witnessed directly, the earthquake's devastating effects appear unexpectedly on our living room television screens, accompanied by familiar commentary on the limits of scientific knowledge in the field. As the forces of nature are better understood, building codes mandate more and more sophisticated methods of earthquake analysis. Still, the most basic notions of seismic design don't change, and some of them are worth reviewing. The first classical theory states that most earthquakes occur when two segments of the earth's crust collide or move relative to each other. The movement creates seismic waves in the

THE BASICS 26

CHAPTER THREE

FIGURE 3.3 Wind load on gabled building. (a) Projected area method of wind load application. It is generally obsolete, but a variation of this, placing the roof lift load on the projected area or roof, is used in the Uniform Building Code, 1997 Edition. (b) Wind acts normally on all surfaces.

surrounding soil perceived by humans as earthquakes; the waves decrease with distance from the epicenter of the earthquake. The wave analogy explains why earthquakes are cyclical and repetitive. The second seismic axiom states that earthquake forces, unlike wind, are not external. Rather, these forces are caused by the inertia of the structure trying to resist ground movement. When the earth literally pulls away from the building, it takes the base of the building with it, but inertia holds the rest of the building in place for a short time. According to Newton's first law, motion between two parts of the building creates a force equal to the acceleration of the ground times the mass of the structure. The heavier the building, the greater the seismic force acting on it. Among the factors affecting the magnitude of the seismic forces on the building is the type of soil, since certain soils tend to amplify seismic waves or even take on a liquid consistency (liquefaction phenomenon). The degree of rigidity of the building is also important. In general, the design seismic force is inversely related to the fundamental period; Strength is also affected by the type of lateral support system in the building. The notion of ductility, or the ability to deform without breaking, is central to modern seismic design philosophy. In addition to being desirable, ductility is critical to the process of determining the magnitude of seismic forces. Building codes may not state this explicitly, but some degree of ductility is required for the rules of the codes to be valid. Without ductility, design forces could

THE BASIC THE BASIC

were easily four or five times larger than those currently reported. Systems with ductile properties, e.g. B. Frames that withstand properly detailed moments can be designed for lower seismic forces than those with lower ductility, such as. B. Shearwalls and braced frames. Why? To answer this question, it is necessary to examine the goals of earthquake planning in general. Most building codes agree that structures designed under the terms of seismic codes must withstand minor earthquakes without damage, moderate earthquakes without structural damage but some non-structural damage, and major earthquakes without collapse. Because the magnitude of actual seismic forces is highly unpredictable, the goal of preventing collapse requires the structure to deform but not fracture under repeated large loads. The structure must be able to stretch beyond its elastic range to dissipate the energy generated by the earthquake. To achieve this goal, codes are filled with many prescribed requirements and design constraints; Particular attention is paid to constructive details, since any interruption in the load path destroys the system. It is important to remember that in real life seismic forces are dynamic and not static, although in practice their effects are commonly approximated by the so-called static equivalent force method. This method is used partly for practical reasons, since dynamic analysis methods are too complicated for routine office use, and partly to compare the results to those of wind load analysis and to use the control load to validate a design against rollover, slip and other in to use failure modes described in Section 2. 3.1.4. Current formulas for determining seismic forces vary widely between building codes and even between different editions of the code. In general, these formulas start with the weight of the structure and multiply it by several coefficients that represent all of the factors discussed above. 3.1.7 Crane loading This type of loading is created by cranes, monorails and similar devices. Crane loads are described in detail in Chap. 15. 3.1.8 Temperature Stress This often overlooked and misunderstood stress occurs whenever a steel element with fixed ends undergoes a change in temperature. A 100-foot piece of mild steel that is free to move expands 0.78 inches for every 100°F increase in temperature and similarly contracts as the temperature decreases. If anything impedes this movement, there will be no expansion or contraction, but internal stresses within the "solid" element will increase dramatically. A basic formula for increasing thermal stress in a steel member with fixed ends is unit stress change ⫽ E ⭈ ⑀ ⭈ t where

E ⫽ modulus of elasticity of steel (29,000,000 psi) ⑀ (epsilon) ⫽ coefficient of linear expansion (0.0000065 in/°F) t ⫽ temperature change (°F)

For example, if the temperature increases by 50°F, a 50 foot long steel beam that is restrained from expanding will experience an additional stress of 29,000,000 ⫻ 0.0000065 ⫻ 50 ⫽ 9425 psi, a significant increase. Temperature stresses are rarely a problem in conventional bolted steelwork or in metal building systems. In fact, the author is not aware of any building failures caused by temperature changes alone, although he has studied a metal building system damaged by heat build-up from a fire - a separate issue. Thermal loading is negligible and can often be neglected when designing primary structures for small prefab buildings or buildings located in areas with relatively constant ambient temperatures or air-conditioned buildings. It could be

THE BASICS 28

CHAPTER THREE

more important for unheated, fixed structures with large spans and low eaves height in climate zones with large temperature fluctuations or for cold storage. For large prefabricated buildings hundreds of meters long, the effects of thermal expansion and contraction must be considered. Thermal contraction of steel frames appears to be more dangerous than thermal expansion because during contraction the steel is under tensile stress which, if large enough, can damage connecting bolts or welds. Temperature changes are felt most strongly by exposed metal roofs, as detailed in Chap. 6 and by metal panels, but the trapezoidal design of its metal plates can reduce the expansion and contraction to a certain extent. It is more difficult to absorb thermal stresses in continuous secondary structures, beams and purlins in unheated buildings. Consider the worst case scenario when the structure is fully loaded with snow and the purlins have contracted towards the center of the building. Therefore, the main frames are likely to be displaced laterally from their original vertical position by contracted purlins (and struts if they are continuous) and metal cladding. The result: an unexpected level of frame twist. Some building codes and the MBMA manual are silent on thermal stresses. How much temperature change should be assumed in the design? The answer depends on the climate, building use and level of insulation. When this load is included, thermal stresses due to a rise or fall of at least 50°F (100°F total change) from the temperature likely to occur at the time of assembly must be accounted for.

3.2 MOVEMENT METHODS AND LOAD COMBINATIONS The loads discussed above need not be randomly grouped. For example, it is very unlikely that a single hurricane would coincide with a record snowfall. The likelihood of having a roof payload, an allowance for infrequent roof repair and maintenance purposes, during a major earthquake is also slim. Traditionally, in order to obtain a realistic picture of the combined stresses on the structure, two approaches have been taken, which are reflected in the design methods for maximum and allowable stresses.

3.2.1 Ultimate design method In this method, also known as the strength design method, loads are summed in various combinations using load factor multipliers for each load and the sum is modified by a 'probability factor'. The resulting combined load is then compared to the "ultimate" capacity of the structure. As previously mentioned, load factors reflect a degree of uncertainty and variability in loads. For steel structures, this method is followed in the Strength and Load Factor Design Specification (LRFD) for Structural Steel Buildings published by the American Institute of Steel Construction5, which provides a list of load combinations similar to those in ASCE 7. The LRFD method of structural analysis provides more consistent reliability than allowable stress design discussed below and may become prevalent for steel structures in the future. However, as of this writing, it has not yet gained widespread acceptance in the design community and is therefore not discussed further here. In addition, users of LRFD in metal building systems may actually be at a disadvantage compared to users of the allowable stress design (ASD) method. Also? The LRFD load factors (1.2 for continuous load and 1.6 for continuous load) were set to ensure a level of reliability equivalent to the ASD for a given ratio of traffic loads to continuous loads. Below this ratio, the LRFD generally offers a more economical design; about it does the ASD. It is easy to figure out which payload-to-deadload ratio for the LRFD and ASD methods provides the same level of reliability. In this ratio, the average ("total") LRFD load factor should be 1.5, which is also the implicit safety factor of the ASD method (remember that the allowable bending stress in compact wide flange members is 0.66 Fy, which is 1 .5 results in ). Therefore, for example, for a dead weight of 1.0 psf and a variable load of R times 1.0 psf, the following equation can be set up to find the ratio R:

THE BASIC THE BASIC

1.0 ⫻ 1.2 ⫹ R ⫻ 1.0 ⫻ 1.6 ⫽ (R ⫹ 1.0) ⫻ 1.5 According to this equation, the “break-even” ratio between live charge and permanent charge is R 3, 0 As the reader is well aware, the steady state load in metal structural systems is extremely small (typically 2 to 3 psf) and any realistic design height of payload or snow will exceed the steady state load by a factor of more than 3.0. making the ASD design more cost-effective for this type of construction. 3.2.2 Method for sizing the allowable stresses In this method, some fractions of loads representing perceived probabilities of simultaneous loads occurring are added in various combinations. The total stress level of the loads in each combination is then calculated and compared to the allowable stress value (expressed as a function of the yield strength for steel members). In the case of wind loads or earthquakes, the permissible stress can usually be increased by a third. There is no universal agreement or single best way to combine the loads acting on the building. Designers should follow the provisions of the applicable building code or, if unavailable, a nationally recognized standard such as ASCE 7 modified for design conditions, if required. For single-storey steelwork systems, the following “basic” load combinations were formerly specified: Tot ⫹ snow (or useful roof load) Tot ⫹ wind (or earthquake) Tot ⫹ snow ⫹ earthquake Tot ⫹ 1Ⲑ2 wind ⫹ snow Tot ⫹ wind ⫹ 1Ⲑ2 snow These sensible load combinations can still be found in some state regulations and are included as “alternative base load combinations” in both the Uniform Building Code, 1997.6 Edition and the International Building Code, 2000 Edition. Both add another combination to their list: 0.9 dead ⫹ earthquake/1.4 The earthquake load in another "alternative" combination is also divided by the factor 1.4. ASCE 7 has used a different approach to combining loads since its 1995 edition, simply summing the effects of all loads. For metal building systems subjected only to dead, live, live roof loads, snow, wind, and earthquakes, the critical ASD load combinations are as follows: death ⫹ snow (or live roof load) [⫹ some other loads, such as temperature and ground pressure] Self-weight ⫹ wind (or earthquake) Self-weight ⫹ living ⫹ snow (or live load on the roof) ⫹ wind (or earthquake) If there are two or more loads acting in addition to the permanent load, the sum of these loads (unloaded) is by a factor of 0, 75 can be reduced. The sum must not be less than the effect of the permanent charge plus the largest undiminished charge. No additional stress increases are permitted for these load combinations. Seismic load is excluded from this reduction and there are separately defined load combinations when this load is present. The load combinations in the latest editions of ASCE 7 are more severe than those listed previously because a one-third increase in load is not allowable for wind acting in combination with a permanent load and because extreme snow and wind levels are simply combined . The International Building Code, 2000 ed. (IBC) contains specific provisions for load combinations. There are two sets of allowable stress design method combinations. The first ("simple") combination is similar to the combinations of ASCE 7 (1995 and later editions), but their combinations involve only permanent and lateral loads:

THE BASICS 30

CHAPTER THREE

0.6 death ⫹ wind 0.6 death ⫹ 0.7 earthquake The second (“alternative”) group of IBC load combinations has already been described. A very relevant standard specification for metal construction systems concerns the alternating load combination "death ⫹ wind". In this combination, the code allows the use of only two-thirds of the minimum probable standing load during a design wind event. Permanent charge combinations must contain an additional charge if they increase the overall effect. When determining the height in the combination "death ⫹ wind", the collateral load is negligible, but must be included if the wind is blowing downwards. Thermal loads not included in the "base" combinations above should be considered as appropriate as discussed above. For all combinations of loads involving snow, both balanced and unbalanced (rather partial) snow loads must be considered. Occasionally designs may require some non-standard load combinations to be considered, either based on local regulations or engineering judgement. In this case, planners should make manufacturers aware of this requirement early - in the bidding or negotiation phase - and be prepared to persevere in the face of some resistance to changing routine practice and available computer programs.

3.3 HOW METAL BUILDINGS STRUCTURALLY FUNCTION 3.3.1 Some Anatomical Structures A typical one-story metal building system is supported by main structures that form multiple spans (Fig. 1.2). The case size is the distance between the center lines of the frame, measured along the sidewall. In the vertical direction, the free span of the frame is the free distance between the supports of the frame. At the roof level, the metal roof panels provide a weatherproof envelope and carry structural loads for purlins, the secondary structural members that span between the main structures. Metal building systems can feature a variety of wall materials, the original and still the most popular being metal paneling supported by sidewalls or sidewalls. Headboards are usually framed with headboard pillars which provide support for the chords and are therefore spaced at intervals dictated by the structural capacity of the chords. End wall pillars support roof beams that extend from pillar to pillar, as in post and beam frames. If a future expansion of the building is planned, a normal main frame can be used instead of the end wall frame; The only function of the bulkhead pillars is to support the belt laterally and vertically. Future expansions will remove pillars and add one or more bays.

3.3.2 Lateral stability of steel structures: Typical approach A building without lateral stability against wind loads and earthquakes will not stand for long. The most popular prefabricated structure, the rigid structure, relies on its own moment capacity to support the building laterally (Fig. 3.4). Other frame systems, such as the well-known post and beam construction, do not have their own rigidity and, due to the lack of rigid walls, can collapse like a house of cards when pushed sideways (Fig. 3.5a). So, the second way to maintain the lateral stability of the building is to use braced frames, as shown in Fig. 3.5b. Vertical bracing not only resists lateral loads, but also stiffens the entire building, especially against crane-related loads, minimizes vibration, and aids in building erection. Vertical stiffness can also be provided by shear walls, which will be discussed separately. A typical structural solution for metal building systems is to provide moment fixed frames spanning the short length of the building and braced frames on the exterior walls. The vertical bracing located on the side walls serves primarily to resist the lateral loads acting in the direction parallel to the frames, while the lateral bracing resists the loads in the vertical direction. The roof panel, usually a system of horizontal bracing, distributes the loads to the lateral load-bearing elements.

THE BASIC THE BASIC

FIGURE 3.4 Moment resistance of a rigid structure.

FIGURE 3.5 Post and beam frames. (a) Unlocked (unstable); (b) supported (stable).

3.3.3 The roof panel When using rigid frames in combination with bracing, the roof panel plays a relatively minor role in accommodating wind or seismic loads acting parallel to the frames, as its span is only a gap between the porticoes. However, the roof pane plays a crucial role in buildings with non-rigid frame types such as single span and bar rafter systems, where the roof pane covers the space between the side walls (Fig. 3.6). Although roof reinforcement is the common type of roof skin found in metal building systems, the same result can be achieved by the rigidity of the steel, timber or concrete roof deck. The corrugated metal cover is probably the most commonly used cone in conventional construction; it also has its place in metal building systems. The continuous metal roof works on the same principle as the metal roof, although it has less rigidity due to the thinner gauges of metal. Nevertheless, the typical roof membrane construction on metal building systems consists of diagonal steel rods designed to withstand tensile forces and struts designed to withstand compression. The diaphragm is essentially a horizontal truss that contains the primary framing beams. For ease of construction, the fascia bars are placed below the purlins (Fig. 3.7), although theoretically both bars and struts should be in one plane. The vertical distance between the purlins and the battens must exceed the maximum expected vertical deflection of the purlins under the full weight load. Some manufacturers prefer steel cable over rods. However, cables tend to come loose; even the bars are difficult to tighten and maintain in a stretched condition throughout the life of the building. Loose rods or cables can cause the building to move significantly before it is engaged, causing damage to non-structural members. Rod connection details are discussed below. Membrane struts can consist of additional purlins designed for axial compression, which usually requires them to be locked laterally at close intervals, as in Fig. 3.8. Without lateral bracing, a

THE BASICS 32

CHAPTER THREE

FIGURE 3.6 The roof skin distributes lateral loads to the braced sidewalls.

A typical cold formed purlin has a very limited compressive strength. The one shown in Fig. 3.8 consists of special Z-profiles fixed with screws both to the beam and to the adjacent purlin; other types of third reinforcement are discussed in Chap. 5. Purlin supports must be provided with roll-off protections on the supports, also in chap. 5. Roof racks are normally designed for axial forces only and are not allowed to carry any other loads. This means that no ceiling brackets or hooks of any kind should be attached to the supports and that no tiles should be laid in the locations of the supports. Purlin supports are usually made of thicker steel than regular purlins, but if the compressive capacity of even the thickest purlin is insufficient (or the span exceeds 35 feet, for example), tubular supports can be used. Unlike cold-formed shapes, tubes require little or no side reinforcement to be fully effective. There are two methods of attaching a pipe mount to the primary structure. In the first, a web connection is established (Fig. 3.9); in the second, the tube is bolted to the top flange of the frame (Fig. 3.10). In both cases, at least two heavy-duty screws are required. The arrangement of the pressure supports can usually be found in the manufacturer's roof construction plan. Fastening details must be carefully coordinated with the location and details of the purlin reinforcement. For example, if the purlin reinforcement consists of two rows of angles at the top and bottom of the purlins, it may be possible to place the pipe bracket in between and Fig 3.9 would get the nod. The fixation of Fig. 3.10 can present more difficulties in this regard unless the pipe size is small enough not to disturb the top row of purlin reinforcement.

3.3.4 Wall Bracing Rigid frames offer little or no lateral resistance perpendicular to their plane unless they are attached to the base - a rare and often undesirable solution. Instead, stability in this direction is usually provided by sidewall reinforcement, as shown in Fig. 3.11. A typical sidewall bracing panel consists of steel bars or diagonal cables, eaves braces and posts on each side. Some manufacturers place the struts in the end panels of the sidewalls, others avoid the end panels and start in the first inner panels, as in Fig. 3.11. The first approach helps stabilize corner areas most vulnerable to hurricane damage; The latter only affects the frames with the highest continuous loads, reducing the lifting forces on the frame anchor bolts. Of course, in small buildings consisting of only two fields, the wall reinforcement can be placed in any field. Manufacturers tend to avoid using standard wall bracing in adjacent bays to avoid complicated detailing and assembly. Shear loads are transmitted along the wall from brace to brace by overhang braces. Eaves supports are designed for axial compression or for combined axial compression and biaxial deflection.

THE BASIC THE BASIC

FIGURE 3.7 Typical roof skin details. (a) With shackle, used with 7⁄8-in. or larger; (b) without hitch, with 3 ⁄4 inch or smaller bars. (Star building systems.)

THE BASICS 34

CHAPTER THREE

FIGURE 3.8 Purlin support resting on the side of the purlin. (Nucor Building Systems.)

FIGURE 3.9 Detail of the connection of the pipe support to the column or rafter. The manufacturer recommends using two A 325 1-in. In diameter. (Nucor Building Systems.)

THE BASIC THE BASIC

FIGURE 3.10 Alternative detail for attaching pipe bracket to column or rafter. The manufacturer recommends using two A 325 1-in. diameter for 6 inches. and three A 325 1-in. Diameter for 8 inches. (Nucor Building Systems.)

How many bays with sidewall bracing are required? In public works, contract drawings are usually made showing all doors and windows before the manufacturer is chosen and the designer has to make an educated guess. Besides the basic orientation of Fig. 3.11, which suggests a maximum of five unlocked slots between locked slots, it might be helpful to ask a few manufacturers. One source (Nucor Building Systems7) recommends using Table 3.1 with the following notes: 1. The building must have the minimum total number of spans for the required number of spans in Table 3.1: The required spans

Minimum total bays

1 2 3 4 5

2 5 7 9 11

2. The table is based on Occupation Category II as defined in the MBMA Handbook. (This category includes most buildings; it excludes essential facilities and those that pose a significant risk to human life in the event of failure.) 3. The letter B or C refers to the wind exposure category. The table should not be used for structures located on the shore of a hurricane. 4. Relatively long buildings may require additional reinforcement. In addition, at least one bracing field must be provided on each side of the expansion joints. 5. Contact the manufacturer for further explanation of the table and the conditions not covered.

THE BASICS 36

CHAPTER THREE

FIGURE 3.11 Typical bracing locations. (Star building systems.)

The end walls of the building must also be braced unless a rigid frame is provided for future expansion or other reasons. The standard positions of the struts on the bulkhead are between the first and second inner pillars (Fig 3.12), although they can be located anywhere along the bulkhead as permitted by the designer.7

3.3.5 Common Wall Bracing Details at the Column Base The most common details of the diagonal bar and cable bracing connection to the column are in Fig. 3.13. Essentially, concentrated strut loads are transferred directly to the strut webs through shoulder washers. The thrust washer (Fig. 3.14) is a cast circular element with a vertical slot that allows for variable angles of rod insertion. A corresponding vertical slot is drilled in the column web. The best washer designs have a protrusion on the back that fits into a matching hole in the web and prevents the washer from riding up under load. Despite the wide distribution, these details could be improved. Thin webs of unreinforced structures are rarely tested for localized bending due to concentrated loads applied by bracing and may not survive actual load application. The author saw that.

Font: Nucor Building Systems.

1 2

90 km/h, c

160 km/h, c

20′

80 km/h, C 100 km/h, B

ⱕ16′ 1

eave height

ⱕ80′

70 mph, B oder C 80 mph, B 90 mph, B

wind speed

building width

TABLE 3.1 Minimum number of braced panels.

24′

30′

ⱕ16′

20′

24′

⬎80′ ⱕ160′

30′

ⱕ16′

20′

24′

⬎160′ ⱕ200′

30′

ⱕ16′

20′

24′

⬎200′ ⱕ240′

30′

THE BASIC

THE BASICS 38

CHAPTER THREE

FIGURE 3.12 Typical locations for bulkhead bracing. Buildings no more than 100 feet wide may only need a single set of braces; If the width is between 100 and 240 feet, two sets are required as shown. (Nucor Building Systems.)

FIGURE 3.13 Typical rod and cable tie details. (a) detail of bar mount on frame; (b) Cable tie to frame detail. (Metal construction systems.)

THE BASIC THE BASIC

FIGURE 3.14 Thrust Washer.

Failure of the bank wash system to connect to the grid was reported by Miller,8 who was investigating damage to several metal buildings caused by the February 17, 1994 Northridge earthquake in California. He reported that five out of six lines in one building were down. The failure mechanism involved rupture of thrust washers (Fig. 3.15) and in some cases subsequent pull on the rods. The missing washer in Figure 3.15 was more than 15 feet from the column. Surprisingly, the building did not collapse. Why? Miller attributed the positive result to its light weight - meaning little seismic loading was generated - and the redundancy of the frame. In this context, redundancy refers to beneficial effects not normally considered in design, such as partial fixation of column bases and even some help from deburring sheet metal. Sinno9 attempted to conduct a definitive study on the ultimate connection behavior. Their lab testing identified five possible failure modes, including stem fracture and four failure modes for the column material and welds. Surprisingly, the breakage of escarpment discs documented by Miller was not one of them. In any case, it seems that the widespread use of standard swashplates attached directly to thin screens should be reevaluated. Fortunately, the problem has been recognized and there is now an alternative. The proprietary line of washers was developed by Triangle Fastener Corp. developed. from Cleveland, Ohio (allegedly inspired by a discussion of this topic in the first edition of the book); a heavy product is in Fig. 3.16. The washer shown appears solid enough to prevent it from breaking under load. However, the thin post web can still be damaged and we recommend placing a steel backing plate under the washer. The plate has to be mounted between the column flanges and welded to them (Picture 3.17). The panel thickness can be determined by calculation.

THE BASICS 40

CHAPTER THREE

FIGURE 3.15 This is the location of an escarpment disc before the Northridge earthquake. (J.R. Miller & Associates.)

FIGURE 3.16 A unique heavy-duty replacement for hillside scrubbers.

THE BASIC THE BASIC

3.3.6 Other wall mount details on the underside of the uprights Some manufacturers offer rod-to-frame connection details that do not involve shoulder washers attached to the frame web. Alternatively, the stiffening rods can be connected to the column tables with bolted brackets. In Fig. 3.18a, the bar is connected to a straight column at its inner flange, a solution that avoids burn holes in embedded beam webs that are usually connected to straight columns. In Fig. 3.18b the rod is bolted to the outer flange of a conical support. The rod is fully contained within the depth of the column to avoid interference with the shunt straps commonly used on tapered columns. (See the discussion in Chapter 5 on the types of chords used in prefabricated buildings.) While these details avoid the problem of attaching directly to an unreinforced web, they introduce torsion into the column near the base and into the anchor rods. A radically different solution to the challenge of connecting support rods to the base is shown in Fig. 3.19. The rod is not attached to the column, but to a separate foundation clip that is screwed directly to the foundation. The obvious advantages of this approach are offset by the need to provide additional anchor rods. In addition, widened pillars or foundation walls are required at the clamping points, as described in Chap. 12.

3.3.7 Wall mount details at the top of frames So far we have described common bar-to-column connections at the bottom of columns. Similar detailing is used for the top bar-to-column connections and in the horizontal membrane mount attachments to the frame. With the obvious exception of the foundation clip, the details available are the same as those used at the base of the column. In Fig. 3.20, the choices frequently offered are: (a) direct fixation to the knee structure; (b) connection to a straight column at the inner flange; and (c) bolted to the outer flange of a tapered column. The details of Fig. 3.20 invite the same comments made for the versions below the column, except that these arguments can be made even stronger here. In fact, in Fig. 3.20a, the thin web of the column must not only withstand the forces applied by the two bars, but also transfer the load from the horizontal roof disk to the vertical bracing of the wall. The author saw this kind of

FIGURE 3.17 The added plate is used for web reinforcement.

THE BASICS 42

CHAPTER THREE

FIGURE 3.18 Details of wall stiffener attachment to column flanges. (a) Mounted in a straight column at the inner flange; (b) Attachment to the conical column at the outer flange. (Star building systems.)

THE BASIC THE BASIC

FIGURE 3.19 The use of foundation clips avoids the difficulty of attaching the locking bars to the column. (Star building systems.)

Connection error under heavy load. In Fig. 3.21, which shows an error in fastening the horizontal roof skin to the beam web, the stressed rod pulled the web, breaking it and even locally bending the beam. In Figures 3.20b and 3.20c, the torsion resulting from the eccentricity of the forces on the bars with respect to the centers of the piers must be contained by the frame itself, as there are no tie bars to oppose the drag. A method of improving the simple connection of Fig. 3.20a is to introduce a combination of reinforced washers and gussets as described above for the lower connection with a thrust washer. Even better details are shown in Fig. 3.22. The horizontal and vertical support rods are connected to a support bolted to a shim on the web, so that the force transmission takes place within the robust support and not within the web of the structure. One final note on the rod-to-column connection: it is important to keep the tie rods taut to avoid excessive noise and building vibration. However, tightening the rods with just a nut behind the thrust washer is difficult, especially considering the tightening is against a thin web plate. A more reliable detail is the provision of a turnbuckle for tightening and securing the bars directly to the columns. A possible solution is shown in Fig. 3.23; could be further improved by providing a plate or weft augmentation angle as discussed above.

3.3.8 Untypical wall bracing systems In some cases the standard rigid frames and bracing systems described above cannot be used and other solutions must be sought. For example, in a very tall building, the proportions of a standard wall brace may exceed the limits of standard connection details. In this case, layered braces can offer the solution. In stepped bracing, an intermediate compression element - a reinforced chord or brace similar to those used in roofing membranes - is introduced to keep the bracing ratio reasonable (Fig. 3.24). In another common scenario, one of the side walls is completely filled with hanging doors or windows, leaving no room for wall bracing. There are three possible design solutions to this situation. The first is to provide reinforcement on only one sidewall and both endwalls in combination with a relatively rigid roof membrane that can effectively distribute the torsional load between the three sides. This solution and the conditions that must be met for it to be feasible are shown in Fig. 3.25. This scheme is more suitable for smaller buildings.

THE BASICS 44

CHAPTER THREE

EDGE STRUCTURE

3/" OR LESS 4-ARM BAR WITHOUT FORK

BOUNDARY STRUCTURE (a)

STRAIGHT STRUCTURAL COLUMN

EDGE STRUCTURE

7/8" OR LARGER ROD MOUNT WITH HOSE

(b)

8" OR LARGER ROD MOUNT WITH TAPERED FORK STRUCTURE COLUMN (c) FIGURE 3.20 Various details of top wall mount attachment. (Star Building Systems.)

THE BASIC THE BASIC

FIGURE 3.21 Fracture of beam web at horizontal bar attachment point. Notice that the force applied by the bracing has also displaced the rafter section laterally.

FIGURE 3.22 A column for load transfer between roof and wall stiffener. (Butler Manufacturing Co.)

THE BASICS 46

CHAPTER THREE

FIGURE 3.23 Detail of rod anchor with welded plates and turnbuckles. (Nucor Building Systems.)

The second solution is to design each pillar like a flagpole as a cantilever attached to the ground. Foundations and columns designed in this way are usually quite expensive. However, a special version of the fixed-base column design called a wind mast—an outdoor, fixed-base column—is relatively common. Wind poles can be used where the wall mount cannot be. In some cases it is possible to place wind masts only on the end walls. As with all fixed-base columns, wind towers create bending moments in the foundations; They should be used with caution and only when the foundations for them can be designed in advance. Some other limitations of the scheme based on wind poles placed only on the end walls are shown in Fig. 3.26. The third solution is to use scaffolding, small rigid rectangular scaffolding that fits between and is fixed between the main columns of the building (Fig. 3.27). Portal frames are discussed further in the next section. Alternatively, concrete or masonry shear walls, which are more rigid than struts, can be used to provide lateral stability (Fig. 3.28). Although expensive to build specifically for the purpose of bracing, shear walls cost very little in buildings with masonry or prefabricated facades. Multiple types of wall stiffeners should not be combined on the same wall unless a detailed relative stiffness analysis is first performed.

3.3.9 Portal Frame As just mentioned, the portal frame is a rigid structure that fits between the main columns of the building. Portal frames are usually placed on the side walls - perpendicular to the span of the main frames. A portal frame can be integrated into the metal building in two ways. the frame

THE BASIC THE BASIC

FIGURE 3.24 Detail of layered bar support. (Nucor Building Systems.)

can be as in Fig. 3.29, the supports of which reach the foundation and are attached to it with anchor rods. At the top, the portal frame is bolted to the main frame posts using small angles (Fig. 3.30). Alternatively, the portal frame supports can end just short of the foundation. To do this, columns must be attached to the primary structure at the top and bottom. A major benefit of not extending the portico columns to the ground is that the pillars do not need to be widened, which can be appreciated by the foundation designer who may not know the exact position of the porticoes.

THE BASICS 48

CHAPTER THREE

FIGURE 3.25 Wall bracing on one side wall only. One manufacturer recommends that this "twist support bay" design be used only when all of the following conditions are met: eave height ⱕ 20 feet; the width ⱕ70 feet; roof pitch ⱕ1:12; there is no crane; Seismic loading does not affect the design. (Nucor Building Systems.)

FIGURE 3.26 Fixed base bulkhead columns. One manufacturer recommends using this bracing option only when all of the following conditions are met: eave height ⱕ 18 feet; the width ⱕ160 feet; the roof pitch ⱕ1:12. (Nucor Building Systems.)

in advance. (The whole subject of constructing foundations before selecting the fabrication fabricator is covered in Chapter 12.) The downside of not extending the portal frame column all the way down is that the underside of the main building column would now provide the level of strength and rigidity comparable with that of the portal frame. This goal can be difficult to achieve because the most important

THE BASIC THE BASIC

FIGURE 3.27 Portal side frame. One manufacturer recommends using it only when all of the following conditions are met: eaves height ⱕ 30 feet; the width ⱕ240 feet. (Nucor Building Systems.)

FIGURE 3.28 Shear wall.

the column is oriented in the weak direction relative to the portal frame. Manufacturers tend not to like this detail and prefer the former. A simpler attachment of the portal frame to the primary frame column can be done by a single bracket, as shown in Fig. 3.30. Unfortunately, this detail suffers from two shortcomings. The first: an elbow placed eccentric to the plane of the portal frame is likely to introduce torsion in it. A better detail is to align the bracket with the plane of the portal frame, or at least use a reinforced angle, as in Fig. 3.31. The second problem is that the portal frame column is not secured against rotation under load. The solution is again in Fig. 3.31: The inner flange of the portal frame can be braced by a pair of horizontal full depth gussets or by flange bracing. For buildings with low eaves, there must be room above the top of the opening for a portal frame. On the other hand, in tall buildings, some space is left between the top of the portal frame and the eaves support. If this space is large, a half-height cross brace can be fitted over the portal frame (Fig. 3.32). The X-strut allows the transfer of lateral forces from the

THE BASICS 50

CHAPTER THREE

FIGURE 3.29 Portal structure in a side wall. (Star building systems.)

Eaves braces in the portal frame without introducing a weak axle bend in the primary frame supports. To maintain reasonable bracing proportions, the distance from the bottom of the portal frame to the top of the eaves support (distance X in Figure 3.32) should be at least 6.5 to 7 feet. Otherwise, a full-height gantry frame is usually supplied.7 The safest way to determine the dimensions and spacing of a gantry frame is to contact the manufacturer who supplies the frames. When dimensions need to be known prior to manufacturer selection, such as B. in public tenders, the following approach (according to Ref. 7) is proposed. First, determine the horizontal loads of the structure by independent analysis or by following the procedures in some manufacturers' catalogs (e.g. Ref. 7). Independent calculations are fairly easy; The projected wind pressure on the wall is calculated and multiplied by the secondary area of the end wall. The resultant force is divided by the number of frames in the wall to get the horizontal frame load, V. Next it is possible to determine the approximate clearances - the maximum clear height or the minimum clear width - that standard frames offer depending on the frame dimensions, span and load V. If a certain headroom H is to be provided, bear the span and the load V in Table 3.2 to determine the minimum headroom W available with standard frames. The maximum clear height of the H-frame for a given overhang height is given in Table 3.3. The numbers in these tables - and most other reference data in this book - should only be used as a rough guide as each manufacturer may have their own standards. In addition, special versions can be supplied at any time with or without a surcharge. 3.3.10 Load Path In a properly functioning building, structural load is transferred between different building elements, like a ball in a soccer game, until it is absorbed by the ground or otherwise dissipated. This load transfer system is called the load path. In order to clarify its function, we trace the course of a wind load acting on the roof of a prefabricated building. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

THE BASIC THE BASIC

FIGURE 3.30 Details of connection of portal structure to building structure. (Nucor Building Systems.)

FIGURE 3.31 The two support wings of the portal frame are braced laterally to the building structure. (Nucor Building Systems.)

THE BASICS 52

CHAPTER THREE

FIGURE 3.32 Nomenclature of side frames. (Nucor Building Systems.)

TABLE 3.2 The minimum headroom provided by standard gantry cranes as a function of enclosure size, strength and desired headroom.

release, h

5K width free, W (minimum)

10K width free, W (minimum)

15K width free, W (minimum)

20′

12′-0″ 14′-0″ 16′-0″ 18′-0″ 20′-0″

16′-10″ 16′-6″ 16′-2″ 16′-2″ 15′-11″

25′

12′-0″ 14′-0″ 16′-0″ 18′-0″ 20′-0″

21′-10″ 21′-6″ 21′-2″ 21′-2″ 20′-11″

30′

12′-0″ 14′-0″ 16′-0″ 18′-0″ 20′-0″

26′-10″ 26′-6″ 26′-2″ 26′-2″ 25′-11″

bay

Font: Nucor Building Systems.

THE BASIC THE BASIC

TABLE 3.3 Maximum free height of frames depending on compartment size, eaves height and strength. loading, (tilts)

eave height

20' shelf spacing, H (max)

25' chassis clearance, H (maximum)

30' shelf spacing, H (max)

12' 16' 20' 24' 30'

9′-10″ 13′-8″ 17′-6″ 21′-6″ 27′-4″

9′-6″ 13′-6″ 17′-6″ 20′-6″ 27′-4″

9′-2″ 13′-2″ 17′-2″ 21′-2″ 27′-2″

12' 16' 20' 24' 30'

9′-6″ 13′-6″ 17′-4″ 21′-4″ 27′-2″

9′-4″ 13′-2″ 17′-2″ 20′-8″ 26′-8″

9′-2″ 12′-8″ 17′-0″ 20′-6″ 27′-2″

fifteen

12' 16' 20' 24' 30'

9′-2″ 13′-0″ 17′-0″ 21′-10″ 26′-8″

8′-8″ 12′-8″ 16′-8″ 20′-8″ 27′-9″

8′-8″ 12′-8″ 17′-0″ 20′-6″ 27′-2″

Font: Nucor Building Systems.

As in Fig. 3.3b, the wind acts perpendicular to the roof, either towards the surface (pressure) or away from it (height or suction). Under wind pressure, the roof panels, the building's first line of defense, are pressed against the purlins and transfer the load by rolling. With the wind lift, the panels are pulled off the roof; The fasteners that hold them in place can fail if constructed incorrectly and send the roof flying away. When the fasteners hold, the purlins flex and transfer the load to the primary structures. Again, the connections must be adequate or the entire roof and purlin assembly will be hanging in the air. The primary structures, in turn, resist bending stress and can also fail if their strength or connections are poor. If the structures resist and the lifting force is not overcome by the weight of the structure, the force shifts to the anchor bolts that secure the structures to the foundations. And finally, when the anchor bolts resist, the wind load is transferred to the foundation, which is expected to have enough weight to counteract the wind uplift. Otherwise, the entire building can be lifted like a giant tree with shallow roots. The final load transfer takes place between the anchor bolt and the foundation and is usually not the responsibility of the metal construction manufacturer. This leaves it to the external engineer to complete the final load path connection and to design the foundations for the most critical load action, a task discussed in Chap. 12.

3.3.11 Stiffening for pressure flange stability The stiffening discussed so far served the lateral strength of buildings. Every bending member - purlin, frame, truss or beam - must be stable even under load. It is a well-known phenomenon that the compression flange of components, when flexed, tends to deflect laterally and this must be prevented by appropriate bracing. Compression flange reinforcement for primary structural members is usually provided by purlins, while purlins in turn rely on purlin reinforcement or by fixed roofs. To be effective, this type of strut must be attached at or near the compression flange. While upper support purlins are certainly the right place for this task, the usual purlin support, which consists of tilting rods or tilting brackets, is usually attached to the purlin net at a distance from the purlin.

THE BASICS 54

CHAPTER THREE

compression flange. Whether this is a problem or not depends on rod size, stress level and network thickness. A 5 foot deep moment frame section can easily tolerate a 4 inch eccentricity while an 8 inch deep purlin cannot. The reinforcement must be designed for a compressive force required to prevent deflection of the compressive flange. The force that the brace must withstand is generally assumed to be 2% of the flange compression force in single span panels and is sometimes increased to 4% in continuous panels. The stiffness of the orthosis must be carefully evaluated since a curved orthosis is essentially useless. The strengthening of primary structures and thirds will be dealt with again in the next two chapters.

3.4 THE COMPETITION OF METAL BUILDING SYSTEMS Our examination of metal building systems is not complete without a cursory overview of the competition. Even die-hard enthusiasts of prefabricated buildings can benefit from an objective comparison with other framing systems, as there is no one most cost-effective framing solution for all circumstances. 3.4.1 Open web steel girders One of the most economical framing systems consists of open web steel girders supporting a galvanized metal deck and supported by joist girders or wide flange steel girders (Fig. 3.33). Open lattice steel girders, popularly known as bar girders, are typically constructed from double angle chords (upper and lower horizontal members) and angled round or diagonal bars. Beams are designed and built by their fabricators to Steel Joist Institute10 specifications, often using proprietary steel design software. Using high-strength steel, the open-mesh beams offer an exceptional strength-to-weight ratio. The beam system is ideal for roof structures that support evenly distributed loads, false ceilings and mechanical ducts. The open design saves space by allowing tubes, ducts and even small HVAC ducts to pass through. However, point loads are a problem and should only be applied at the plate points where the diagonals cross the chords, as the chord sections are generally quite weak in terms of local deflection and the beam's bearing capacity may not be sufficient. Heavy point loads may require a special support structure. Beams are unstable during assembly and SJI specifications require multiple rows of bridges for lateral bracing. Once the brace is properly attached and the roof panel is secured, the system will be stable. Together with the surrounding steel girders, the steel ceiling forms a horizontal roof membrane that fulfills the same function as the horizontal roof reinforcement in metal construction systems. The distance between the beams depends on the load-bearing capacity of the roof surface.

FIGURE 3.33 Open joist system and steel deck.

THE BASIC THE BASIC

Historically, joist rafters have been used on relatively flat roofs because large slopes present some difficulties in manufacturing sloped seats and require careful structural analysis. Many engineers avoid this system when the roof pitch exceeds 45°. The open joist system is very economical for spans from 25 to 50 feet. Long-span steel beams are economical for even longer spans from 60 to 144 feet. In a rectangular column layout (a 30' x 40' grid is considered by many to be the most economical cabinet size for an office building), the beams typically run longitudinally (40 feet in a 30' x 40' cabinet). Open web beams can be supported by hot rolled wide flange beams or girder beams. Joist trusses work on the same principle as joist trusses—like a mini truss—but are generally made from heavier sections at all angles. The joist support plate points correspond to the bar support positions. Joist joists are often preferred over wide flange sections, especially on longer spans and on larger projects where more than a few are required. The system is generally supported on wide flange or tubular columns and requires wall bracing for lateral stability.

3.4.2 Hot rolled wide flange beams The system of wide flange beams and beams supporting the steel roof should be familiar to most of those involved in building construction. The simplest and most versatile of the framing systems, it easily adapts to any roof pitch and easily accommodates suspended, concentrated and axial loads. Beams can be cantilevered and arranged in complex configurations for non-perpendicular planes, modified or reinforced for localized loads. Of course, flexibility has its price. Unless some of these complications actually exist, steel beams are likely to be more expensive than joist beams. Beams tend to get excessively deep and heavy when spans exceed about 40 feet. Civil engineers identified two ways to increase the efficiency of this system. The first is a continuous beam principle. Three simply supported beams (Figure 3.34a) have maximum bending moments and larger vertical deflections than a continuous beam (Figure 3.34b) spanning the same three spans. Therefore, a three-span continuous beam is more efficient and requires less metal than three single-span beams. Continuous framing has its limitations. Since it is statically indeterminate, it does not tolerate differential settlements of the supports, which can lead to large secondary stresses that threaten its integrity. Another problem is the potential for large thermal stresses to build up in a single long piece of metal. The registered design professional should carefully investigate the potential for problems caused by these two factors before specifying the ongoing framework. A second way to increase the efficiency of the system is the cantilevered beam scheme. Instead of a continuous beam, this truss consists of alternating cantilevered and simply supported beams (Fig. 3.35). The beam connections form hinges designed not to transmit bending moments. The length of the overhangs is chosen to produce approximately equal negative moments on the overhangs and positive moments on simply supported beams. This system is statically determinate and is less influenced by different settlements or temperature stresses. The success of the project

Figure 3.34 The efficiency of continuity.

THE BASICS 56

CHAPTER THREE

FIGURE 3.35 Cantilever beam frame.

depends, so to speak, on the correct design of the joint. Connections that are too rigid tend to convert this system into a continuous beam with all its limitations. Other efficiency paths can be taken with the LRFD method (see section 3.2.1) and the specification of high-strength steels.

3.4.3 Steel Truss Long span steel trusses have been used in bridge and building construction for many decades. Before the advent of metal building systems, trusses were the structure of choice for industrial, storage and commercial applications. Roof trusses were, and still are, very economical for spans of about 40 to 140 feet. For example, a portal truss was designed to span approximately 132 feet over an existing building11; Some recent projects have used open trusses to span nearly 200 feet. Steel lattice girders can be designed with a single or double slope of the top chord. The double drop leads to a structurally advantageous lower section in the middle of the span. In the past, trusses supported wide flanged steel purlins and truss spacing was limited by the length the purlins could span. Today's Tuesday options include not only the still popular profiles, but also open lattice girders and even extra deep steel roof decks. Optimum truss spacing, dictated by purlin capacity, is generally between 20 and 30 feet, just like prefabricated buildings. As with the two previous support systems, lateral stability is ensured by horizontal roof panels made of sheet steel or diagonal bars in combination with vertical wall reinforcements. In the past, trusses were usually laterally supported by knee braces up to the first plate point (Fig. 3.36a). This solution sacrificed some internal headroom and lost popularity in favor of a truss design with some depth in the supports, with the column integrated into the truss (Fig. 3.36b). In the latter case, the column is erected and braced first, followed by relatively simple truss-to-column connections. Trusses are partially assembled in the workshop to the maximum allowable shipping width of approximately 12 feet and fully assembled on site.

3.4.4 Hot Rolled Steel Rigid Frames A rigid frame, also known as a moment bearing frame, consists of column and beam sections rigidly connected to each other by moment bearing connections. The resulting unified structure is stable and does not require bracing in its own plane. We have already mentioned rigid frames as the most popular frames for metal building systems, but frames can be built by others as well. In fact, low-rise buildings have traditionally used gabled structures of regular wide flanged members or tapered sections. The taper was achieved by bevel cutting a wide flange beam web, turning a section and welding the webs together, or simply welding the sheet steel construction. As the benefits of prefabricated construction become apparent and labor costs increase, custom manufacturing of rigid structures has fallen out of favor.

THE BASIC THE BASIC

FIGURE 3.36 Roof truss. (a) finch truss; (b) Warren truss.

However, one report12 claimed that an efficient steel fabricator, in conjunction with an innovative structural engineer, produced gable structures that were priced below the lowest bids from prefabricated building manufacturers. Among other things, the engineer was able to take into account the fixing of the support to the side wall when planning the foundation, since both the foundation and the structure were designed in-house. This would have been difficult if the bases and structures had been designed by different parties (see discussion in Chapter 12). It remains to be seen whether this experiment can be successfully replicated. Rigid frames offer many advantages over other types of frames, such as: B. More effective use of steel as a single beam, easier maintenance and cleaning compared to trusses, and the ability to withstand heavy, concentrated loads. The disadvantages include a relatively high unit cost of the material and susceptibility to different settlements and temperature stresses. The structures create horizontal reactions in the foundations, an additional design complication. And as with any fixed web structure, tubing and piping must be placed below the bottom flange unless expensive web openings are provided. 3.4.5 Heavy-Duty (Hybrid) Structures In the case of large industrial and multi-storey buildings with flexible and variable loads, extremely large spans or heavy cranes, the advantages of the conventional structural systems listed above become clear. Instead of leaving this market open to competition, some visionary steel fabricators have created “heavy duty” divisions. Essentially, these hybrid structures use conventional structural steel beams or rigid frames for the primary frame, but have cold-formed secondary members. Buildings are generally clad with metal roofing and cladding. Carter13 attributes the growing popularity of heavy construction to the fact that the relationship between labor and material costs has changed, and the high manufacturing costs of metal building systems often outweigh the material savings, the traditional benefit of prefabricated buildings. In addition, the new steelworks became very efficient in the production of steel profiles. As a result, the structure with the lowest weight is no longer necessarily the most economical. Additionally, structural steel beams have an advantage in buildings where structural deflections must be tightly controlled, such as aircraft hangars, theaters, and precision factories. Trusses can be bent with precision, while rigid structures are difficult to bend. Other reasons for using hybrid structures include corrosive environments and fatigue-promoting applications where welded structures are at a disadvantage, Carter said. 3.4.6 Other structural systems There are many other types of structural systems that may be more appropriate for the project in question than metallic structural systems. Lack of space precludes a discussion of all, not even in detail. Among the most popular systems we can mention the following:

THE BASICS 58

CHAPTER THREE ●

●

Laminated wooden bows. Functionally similar to rigid steel frames, these structures share the same design concepts as their steel brethren. The great advantage of wooden arches is their charming and cozy look as opposed to the cold and utilitarian look of steel frames. Wooden arches in combination with wooden roofs and masonry walls are used for churches, community buildings and luxury homes. Laminated wood arches are economical for spans from 30 to 70 feet. Roof pitches range from 3:12 to 14:12. This system offers a unique combination of beauty, strength, ease of installation and economy. Disadvantages include that unlike steel, wood can rot and be attacked by termites. precast concrete construction. Precast concrete is heavier and generally more expensive than metal structural systems, but in some circumstances factors such as fire resistance and soundproofing are more important than cost. Concrete offers both. Precast hollow core prestressed planks are generally available in depths of 6 to 12 inches and can span distances of 20 to 50 feet. I-T panels ranging in depth from 12 to 32 inches can span distances from 12 to 100 feet. Roof panels are typically supported on precast concrete structures or masonry walls and rely on shear walls for lateral strength. Building structures made entirely of precast concrete offer some of the same advantages as metal building systems: speed of erection (some projects were erected in 21Ⲑ2 weeks – in winter), shell construction responsibility from a single source and even expansion flexibility. special construction. Some truly exceptional structural systems have been developed for applications that require bold appearance, very long spans and other unusual criteria. Suspension systems that use external steel cables for roof support are most commonly used for bridge applications, but can also occasionally be used in civil engineering. Airborne fabric structures, such as those used at Denver International Airport, offer an impressive opportunity to cover large areas. Special structures can be used for spans greater than 1000 feet. As the name suggests, specialists should be sought to help with this type of project.

3.5 DECISION TIME How and when do architects and engineers decide whether or not to specify metallic building systems for the project? How and when do the systems compare to other frame types available? These are not idle questions. Determined by enthusiastic designers too early, before design requirements are fully established, metal construction systems can end up being stretched beyond their ideal field of application. Systems defined too late in the design process can prove incompatible with design elements already selected. Let's take a quick look at the milestones of a typical construction project, which begins when the planners realize the client has a problem to solve. The problem can be anything from a lack of space to the need for new equipment. In the first phase, programming, the problem is examined and analyzed. The program report summarizes the planners' recommendations for the space actually required and defines the basic requirements for the planned building. At this stage it is too early to discuss structural systems unless the only solution is already apparent. During conceptual design and preliminary design, the program requirements are translated into a proposed building layout, size and mass; various aspects of building regulations are examined; and a preliminary cost estimate will be prepared. This is the best time to involve civil engineers. Unfortunately, engineers are often not hired until the preliminary design is complete, and the opportunity to influence design decisions related to the building's shape and span is lost. Some major customers also prefer to do schematic design in-house. When asked about the lessons he learned, Eli Cohen, one of the most respected engineers of our time, responded, “You need to spend more time on concept development because the first 10% of your time can save you 25% of the time. the cost of the building.”14

THE BASIC THE BASIC

At this point the project can go in one of three directions: 1. Conventional delivery. The building is designed by an external architect-engineer and then built by a general contractor selected through a public tender or negotiated procedure. 2. Project construction. The building is designed and built by a single entity that includes designers and builders. 3. Prefab construction sold directly. A local contractor, acting as a concessionaire for the fabrication system manufacturer, contracts directly with the owner, who may be assisted by an architect. Of course, choosing the third method means that a metal construction system has already been selected for the task. However, if one of the first two delivery methods is chosen, the decision on the use of metal construction systems and their construction is made in the next design phase, the project development. At this point, structural engineers, armed with information about the building from preliminary drawings and after researching building codes, will determine the design loads on the structure and evaluate various structural alternatives.

3.6 STRUCTURAL SYSTEM SELECTION CRITERIA Having discussed the issues of structural loads, design philosophies, available framing systems and design methods, we can finally consider some of the system selection criteria.

3.6.1 Architectural requirements The selected system must meet architectural and structural requirements; The relative importance of each must be determined during schematic design. It is wise to put Louis Sullivan's timeless words into practice: "Form follows function." In most manufacturing buildings and other "useful" occupations such as factories, warehouses and department stores, the harmony between function and structural form is evident. For other uses, such as churches, community and commerce, architectural expression is likely to be of dominant importance and may override considerations of sheer structural efficiency.

3.6.2 Fire resistance The fire safety requirements set out in local building codes often determine the choice of construction. Prefabricated buildings of light steel construction, trusses and support systems are difficult to spray with fireproof spray; these must be carefully determined where fire protection is required. Fortunately, for single-storey buildings that meet the non-combustible, unprotected classification, this is rarely necessary. However, there are circumstances where the building structure must have a fire rating of one or two hours. The metal construction industry has developed some fire-resistant systems that rely on multiple layers of plasterboard supported by hat channels running between wall joists. The roofs of metal buildings are usually high enough not to require fire protection, but plasterboard can also be used if required. The main problem with these projects, of course, is cost: the multi-layer plasterboard plus the metal siding and wall framing can cost as much or more than cinder block or precast concrete walls. Also, as in chap. 11, attaching gypsum board to the metal structure of the building drastically increases the lateral rigidity requirements of the building, further increasing its cost. It is often more economical to use 'hard' walls, either as part of metal construction systems or their competitors when fire protection is required.

THE BASICS 60

CHAPTER THREE

3.6.3 Cost Efficiency Throughout the discussion in this chapter we have mentioned optimal spans, advantages and disadvantages of different systems. Where structural efficiency and cost are of paramount importance, these guidelines are intended to assist an experienced professional in narrowing down system options. However, the design team must choose a framing system that results in the lowest overall cost of construction, not just the most cost-effective structure. When a structural system puts other building systems at a disadvantage, it may not be the deal it seems.

3.6.4 Flexibility of use and expansion The design team should carefully assess the owner's requirements for the building's free span, height and floor plan and compare this information to similar newly designed buildings. With the information and technology revolution in full swing, it is unlikely that a factory designed today will contain the same production processes 20 years from now. On the other hand, the floor plan of a church must not change. There is an obvious trade-off between cost efficiency and scheduling flexibility. For maximum flexibility, the framing should be easy to remove, change or reinforce to meet future needs; All of this is easier to achieve with a single-span frame. Of all the frame materials, hot-rolled steel beams are still the most conformable. On the other hand, the most economical building systems use continuity, multi-span cantilever beams, or post-tensioning. The owner and his team have to decide whether it makes sense to spend a little more now in order to have full planning freedom later.

3.6.5 Construction time The owner often attaches great importance to shortening the construction time. This is understandable: time is money. Several months of schedule delays can mean real home loan savings that may be greater than the differences between competing structural plans. Frames with shorter assembly times (e.g. metallic building systems) receive some additional points with this article.

3.6.6 Soil data Preliminary designs and project developments are often undertaken without adequate geotechnical information; Engineers are expected to recommend a support system without ground data. This is unfortunate, since the soil properties are decisive for the system selection. If the soil is good, economical foundations are possible; Poor soil may require deep and expensive foundations. Much of the remaining land near major cities is likely to have poor soils considered unsuitable for earlier development. Late realization of the need for expensive piles can ruin a tight-budget project. In such circumstances, choosing a lightweight and flexible construction system capable of tolerating some differential settlement can mean the difference between going ahead with the project or not.

3.6.7 Local Practices Prevailing local practices can have a major impact on system selection and should never be ignored. For example, on the island of Guam, most buildings are made of concrete; The specification of a metal construction system, while seemingly appropriate for a new building, may raise many eyebrows. The abundance of local contractors specializing in a particular build means there are always qualified people interested in submitting a proposal with few eventualities. It also likely means the materials needed are plentiful and cheap.

THE BASIC THE BASIC

3.6.8 The choice There are many other, often conflicting, factors that can influence the choice of structural system; most relate only vaguely to structural issues. People who can make such decisions realize that choosing a structural system is more an art than a science. Some members of the design team may disagree at first; In many cases, studies of alternative systems are developed along with cost estimates. Most importantly, the decision-making process allows all stakeholders to make their first and second choices of systems, which are then thoroughly analyzed and discussed. The best solution is not always the most obvious. Swensson and Robinson15 discuss the selection of a structural scheme for a large sports facility. Four final schemes were considered: a basic gabled metal building, gabled truss, flat truss, and arch. The designers eliminated the flat binder because it required a much larger volume of conditioned space than others. Arches were discarded because they offered less workable space than pediments. And finally, despite the higher cost, the gable truss was preferred to the gable metal building. Why? Because "the design quality and flexibility that [this system] offers more than offsets the additional cost of about 10%..." Given such notions, which are still prevalent among engineers, it can be difficult to select a pre-engineered - - Construction system for a high visibility project. As the metal fabrication industry continues to prove itself in non-traditional applications and its level of engineering sophistication continues to grow, the quality of prefabricated structures is likely to rival rod structures. This book is just a small attempt at that effort.

REFERENZEN 1. Handbuch für Metallbausysteme, MBMA, Cleveland, OH, 2002. 2. International Building Code, International Code Council, Falls Church, VA, 2000. 3. Minimale Bemessungslasten für Gebäude und andere Strukturen, ASCE Standard 7, American Society of Civil Engineers, New York, NY, 1998. 4. David B. Peraza, „Lessons from Recent Collapses of Metal Buildings“, Proceedings, 15th International Specialty Conference on Cold-Formed Steel Structures, St. Louis, MO, 19.–20. Oktober 2000. 5. Festigkeits- und Belastungsfaktor-Konstruktionsspezifikation für Stahlkonstruktionen, American Institute of Steel Construction, Inc., Chicago, IL, 1986. 6. Uniform Code of Construction, International Conference of Building Officials, Whittier, CA, 1997. 7. Engineering and Product Handbook, Nucor Building Systems, Waterloo, IN, 2001. 8. James R. Miller, „Performance of Pre-Engineered Buildings in the CA Earthquake“, Metal Construction News, Juli 1994. 9. Ralph Sinno, „X-Bracing Anchorage Connection“, Journal of Structural Engineering, vol. 119, Nr. 11, November 1993. 10. Standard Specifications for Steel Joists Open Web, K-Series, Steel Joist Institute, Myrtle Beach, SC, 1994. 11. Alexander Newman, „Boston Edison No. 514“, Moderner Stahlbau, Nr. 3, 1989. 12. Thomas Sputo, „Innovative Design of Gable Frame Buildings“, Modern Steel Construction, Mai 1994. 13. Malcolm A. Carter, „Expanding the Market: Using Heavy Structures“, Metal Construction News, Mai 1997, S . . 107. 14. Cindi Crane, „Designing Buildings That Work“, Modern Steel Construction, Oktober 1994. 15. Kurt D. Swensson und Douglas W. Robinson, „Field of Dreams“, Modern Steel Construction, März 1995.

REVIEW QUESTIONS 1 List three structural systems that compete directly with metal buildings. 2 Explain the common method of attaching a vertical rod or cable tie to metal building columns. What improvements could be made to this type of accessory?

THE BASICS 62

CHAPTER THREE

3 At what point in the planning process should the decision to use metallic building systems be made? 4 Why is collateral encumbrance required? What collateral value is typically used in metal construction systems? 5 Name at least three types of wind damage to buildings. 6 Which part of a metal building system is most affected by temperature changes? 7 Is the load combination (Tot ⫹ 1⁄2 Wind ⫹ Snow) included in the latest edition of ASCE 7? 8. List at least three methods of resisting lateral loads used in metal building systems.

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 4

PRIMARY FRAMEWORK

4.1 INTRODUCTION This chapter examines a range of primary structural systems used in prefabricated buildings. As discussed in the previous chapter, the complex process of choosing a framework involves much more than just structural considerations. If a metal construction system is selected for the project in question, the next step is to choose from the available types of prefabricated primary structures. Proper selection of the primary structure, the backbone of metal buildings, contributes significantly to the successful implementation of the following design steps. Factors influencing main frame choice include: ● ● ● ● ●

Building dimensions: width, length and height Roof pitch Free requirements without columns Building occupancy and acceptance of exposed steel columns Proposed roof and wall materials

When all of these factors are considered, the most appropriate type of primary structure system often becomes apparent.

4.2 THE AVAILABLE SYSTEMS Manufacturers refer to their frame systems by many different names, often distilled into a soup of letters of abbreviation. However, only five basic types of steel structures for buildings are currently on the market: ● ● ● ● ●

Cone Beam Rigid Single Span Gun Multi-Span Starr Track Continuous Single Span Traversen Lean-to

Each type can be supplied with a single or double pitched roof. The most common primary frame systems are shown in Fig. 4.1. The primary structure is usually made of high strength steel conforming to ASTM A 992 with a minimum yield strength of 50,000 psi or, rarely today, ASTM A 36 steel. Every system has an optimum range of free spans, as described below, but before this discussion we must first define the terms, relating to the measurement of metallic buildings. The width of the structure is measured between the outside faces of the joists and the eaves, while the free span is the distance

63 Transferred from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

PRIMARY FRAME 64

CHAPTER FOUR

FIGURE 4.1 Primary frame types. (a) Conical beam; (b) single-span rigid frame, low profile; (c) rigid single-span frame, medium profile; (d) rigid frame with multiple spans; (e) single-span trusses; (f) continuous truss; (g) porch; (h) posts and beams with an incline. (Adapted from the Star Building Systems Design Guide.)

between the insides of the posts.* The eaves height is measured between the bottom of the post base and the top of the eaves brace; The free height is the distance between the floor and the lowest point of the structure, usually the rafter (see fig. 1.2 in chap. 1). How do I dimension metallic buildings in architectural drawings? Manufacturers expect the width and length of the building to be shown as distances between the exterior faces of the walls (this plane is called the sidewall structure line), not between the centerlines of the exterior columns. *The term free span is sometimes misunderstood, despite the name, as some people measure it at the base and others at the widest part of the spine, such as the crotch. B. the knee.

PRIMARY FRAME PRIMARY FRAME

The misunderstanding of this convention leads to disputes between designers and manufacturers, and buildings being delivered in sizes slightly smaller than the designers envisioned.

4.3 TAPERED BEAM The tapered beam, also known as wedge beam or inclined beam, is a logical extension of conventional post and beam construction in metal building systems. What really distinguishes this system from a plate girder built on two wide flange columns is the variability of the girder depth and the partial stiffness of the girder-column connections. Most often, the beam is tapered by sloping the top flange for water drainage and keeping the bottom flange horizontal for ceiling applications (Fig 4.1a). A less common version, resembling a scissor truss, surrounds the beam with both angled wings. This configuration can be particularly useful for the steeply pitched roof used in combination with a low-pitched cathedral ceiling. Splices usually occur in the middle of the span. The cone blast system is suitable if: ● ●

●

The width of the structure is between 30 and 60 feet, and the height of the eaves does not exceed 20 feet. Straight columns are desired (an important consideration for office and commercial buildings with drywall). The roofing material can tolerate a relatively small roof pitch.

Tapered beams lose their appeal at spans greater than 60 feet. Also, the standard hot rolled frame can be less expensive if the frame width is less than 30 feet. Tapered beam structures are typically specified for offices and small commercial and retail uses with moderate spacing requirements. Tapered girders are sometimes preferred for crane buildings as their lower flanges, which are horizontal, facilitate anchorages and local reinforcements. The system is detailed in Fig. 4.2. Typical frame dimensions for different spans and moving roof loads are shown in Fig. 4.3. The design of tapered straps involves an often-overlooked nuance. Manufacturers sometimes assume that the beams in this system are connected to supports with "wind connections" that are rigid enough to withstand lateral loads but flexible enough to behave under single-span payloads. The question is: How realistic is this assumption? In steel construction according to AISC Manual of Steel Construction, vol. II, Connections1 indicates that these semi-rigid connections must meet certain criteria. One of these requirements is that "the fastener material must have sufficient inelastic rotational capacity" to prevent failure under combined gravity and wind loads. The semi-rigid connection chosen in the AISC manual consists of a pair of flexible brackets that attach the upper and lower beam flanges to the column (see Fig. 4.4). A flexible behavior of this compound has been demonstrated experimentally. On the other hand, in metal construction systems, the bars are usually connected by bolted end plates, as shown in Fig. 4.2. This type of connection is quite rigid and does not go beyond the flexibility required for "wind connections". When a joint runs out of rotational capacity, it is considered nearly rigid and therefore able to transmit bending moments across the interface. The ends of the beam connected to the columns in this way have some bending moments which are partially transmitted to the columns. If a one-span assumption is made by the manufacturer, the supports are not designed for this bending moment. This dangerous simplification can lead to overloading of the columns by combined axial and bending stresses. Whenever a cone blast system is proposed, it is prudent to investigate the manufacturer's approach to the problem. There are a few steps manufacturers can take to increase connectivity. A solution may be to introduce compressible deflection pads, as in Fig. 4.14 (in the context of another system), which absorb some movement from the beam ends. The designers could of course circumvent the whole problem by simply adding a note to the contract documents that all beam-to-column connections in the conical beam system are to be considered rigid for design purposes, but this approach could lead to heavier connections in column sections.

PRIMARY FRAME 66

CHAPTER FOUR

FIGURE 4.2 Details of cone blasting system. (Metal construction systems.)

4.4 SINGLE-SPAN FIXED STRUCTURE If a conical truss system is a legacy of conventional construction, then the single-span rigid gable structure (Fig. 4.1b and c) is a prefabricated product par excellence. In fact, one reason for the success of the metal fabrication industry is its rigid structure. In contrast to the conical beam system, the single-span rigid frame is designed in such a way that it fully utilizes the rigidity of the connection: the frame bars are tapered according to the bending moment diagram. The deepest part of the structure is the knee, a joint between the beam and the column. In a double-hinged frame, the most common version of the system, the frame section is flattest about midway between the knee and crest (Fig. 4.5); In a less common three-hinged frame, the flattest section occurs at the ridge (Fig. 4.1b). Splices are made at the knee, at the ridge, and possibly elsewhere on the wearer depending on the width of the frame. Splices are typically made using bolted end plate connections. Knee correction can be done in three different places: On the vertical side of the spine (Fig. 4.1 and Fig. 4.6b); diagonally across the knee (Figs. 4.5 and 4.6a); or horizontally below the bar (Fig. 4.8). A typical endplate beam joint is shown in Fig. 4.7. The main reason for the popularity of the rigid gable framing system is its cost effectiveness - it requires less metal than most other structural systems of the same span and eaves height. As McGuire2 demonstrated and can easily verify, a rigid two-hinged gable structure with a span of 60 feet and an eaves height of 14 feet plus a gable height of 10 feet is 19% more efficient than a similar rigid roof structure. , and an incredible 53% more efficient than a statically determined structure based on the single-span principle. This frame system is suitable if:

PRIMARY FRAME PRIMARY FRAME

FIGURE 4.3 Typical dimensions of the tapered beam system. (American construction company)

●

● ● ●

The frame width is between 60 and 120 feet. Both smaller and larger spans are always less economical. The eaves height is between 10 and 24 feet. Tapered columns are acceptable. The clear height on external walls is not critical.

Single-span rigid frames can be categorized as high-profile (4:12 pitch), medium-profile (2:12 pitch), and low-profile (1/4:12 pitch to 1:12 pitch). The high profile frames are particularly recommended for roofs that require a large roof pitch and for applications that require large clear heights near mid-span. Inwardly tapered columns are the norm, but some other column configurations are possible for special conditions (Fig. 4.8). The single span rigid frame system is used wherever an unobstructed work area is desired. It is suitable for various applications such as lecture halls, gymnasiums, aircraft hangars, showrooms, churches, leisure facilities and industrial halls (Fig. 4.9). While frame widths are best kept between 60 and 120 feet, single tenter frames can be built in excess of 200 feet wide for cases where design flexibility is key. Tables showing typical dimensions of single span rigid frames are given in Figures 4.10 and 4.11.

PRIMARY FRAME 68

CHAPTER FOUR

FIGURE 4.4 Traditional semi-rigid connection.

FIGURE 4.5 Details of the single span rigid frame. (Metal construction systems.)

4.5 MULTI-LANE FIXED STRUCTURE The multi-span rigid structure, also known as a continuous beam, mullion-transom or modular structure (Fig. 4.1d), uses the same design principles as the single-span rigid structure. The variety of spans allows for unlimited building size in theory, although in reality thermal stress build-up requires that expansion joints be used for buildings over 300 feet wide. Rigid frames with multiple spans can have straight or conical supports, the latter mostly external. Rays are usually conical. Construction details are similar to single span rigid frames except

PRIMARY FRAME PRIMARY FRAME

FIGURE 4.6 Two methods of connecting the column to the beam at the elbow of a rigid frame. (a) diagonals; (b) vertical. (Steelox Systems Inc.)

FIGURE 4.7 Splicing End Plate Beams. (Steelox Systems Inc.)

for the additional interior speakers. Some typical frame dimensions are shown in Fig. 4.12. Fastenings between internal supports and beams are generally considered as pin connections rather than full moment connections, and supports are designed as components with purely axial loads. Relatively high shear stresses in beams above columns usually require web reinforcement (Figure 4.13). Rigid structures with multiple spans are often the only solution for the largest buildings such as warehouses, distribution centers, factories and resource reclamation facilities. Rigid multi-span frames use continuous frames and are typically more economical than their single-span cousins. One of the disadvantages of continuous construction is the susceptibility to different settlements of columns, as discussed in Chap. 3. Site soil conditions must be carefully evaluated before specifying this system. In addition, it will be difficult to change the positions of the internal speakers in the future should this become necessary due to a new device layout.

4.6 Single-span and continuous trusses Single-span trusses (Fig. 4.1e) and continuous trusses (Fig. 4.1f) are similar in their function to single-span and multi-span rigid frames. The decisive difference between frame and half-timbering lies in the construction of the rafter web - open in the case of half-timbering and solid in the case of half-timbering. An open web allows for the passage of pipes and ducts and therefore allows for a lower eaves height in a timber framed building, resulting in a lower eaves height

PRIMARY FRAME 70

CHAPTER FOUR

FIGURE 4.8 Different column profiles. (VP building.)

FIGURE 4.9 This open-fronted industrial plant is ideal for constructing a rigid single-span frame. (Photo: Maguire Group Inc.)

PRIMARY FRAME PRIMARY FRAME

FIGURE 4.10 Typical dimensions of a low profile rigid single span frame. (Metal construction systems.)

PRIMARY FRAME 72

CHAPTER FOUR

FIGURE 4.11 Typical dimensions of a high profile single span rigid frame. (American construction company)

PRIMARY FRAME PRIMARY FRAME

FIGURE 4.12 Typical dimensions of multi-span rigid frames. (Constructive appendix systems.)

(Video) THE FLASH: SuperHero Kids Classics Compilation!

PRIMARY FRAME 74

CHAPTER FOUR

FIGURE 4.13 Connection of internal columns to beams in multi-span rigid frames. (Steelox Systems Inc.)

a smaller building volume to be heated or cooled and thus lower energy costs. Therefore, trusses are best suited for pipe and utility-heavy applications, such as factories and distribution centers. An example of a simply supported truss structure is the Landmark Structural System by Butler Manufacturing Company. Figure 4.14 shows the details of the 3 and 4 foot deep trusses that are common in this system. Note the deflection pad between the lower truss chord and the post, which is designed to allow some rotation of the element under gravity loading without inducing bending moments in the post. In fact, this is a good "wind connection" discussed above for a tapered truss system that provides lateral strength in the plane of the truss. At the Landmark, lateral strength along the length of the building is provided by solid-based pilasters, an approach with some pitfalls, as discussed later in this chapter.

4.7 LEAN-TO FRAME Lean-to-framing, also known as a wing unit (Fig. 4.1g), is not really a support system in its own right, but rather an addition to another building system. Cone-shaped beams and straight columns are common in this type of construction (Fig. 4.15). For optimal efficiency, the system is best specified for distances of 15 to 30 feet. The inclined structure is typically used for building additions, equipment rooms, storage and a range of other smaller outbuildings. The structural details are similar to a tapered beam system, except that a single slope is usually provided on the top surface and the beam's taper prevents the bottom surface from being horizontal.

4.8 OTHER FRAME SYSTEMS In addition to the frameworks described above, there are several proprietary systems on the market that are truly unique, as well as those that only pretend to be different by assuming an unknown name and some unusual details. Some of the “substantial others” are mentioned below.

PRIMARY FRAME PRIMARY FRAME

FIGURA 4.14 Details von Butler's Landmark Building. (Butler-Manufacturing Co.)

PRIMARY FRAME 76

CHAPTER FOUR

FIGURE 4.15 Inclined structure. (Metal construction systems.)

4.8.1 Single Pitch Structures Any of the five basic prefabricated framing systems can be fabricated in a single pitch configuration rather than a gable configuration. The single slope characteristic does not significantly affect the structural behavior of the structure, its free span ability, or typical details. To confuse the terminology, some companies refer to their single pitch rigid frame products (Fig. 4.1h) as “slanted rigid frames”. In such cases, a picture really says more than a thousand words. Frames with a pitch are often used for office complexes and shopping malls where rainwater needs to be drained from parking lots or adjacent buildings. 4.8.2 Truss Structures Coronis Building Systems, Inc. began producing its own line of frames in 1956 and has since developed over 8,000 variations of the frame. A typical truss frame* resembles a tapered beam, except that its web consists of truss-like rather than solid elements (Fig. 4.16). Other types of truss frames include multi-span cantilever shelters, canopies, and porches. A major advantage of this system, as stated by the manufacturer, is the absence of horizontal reactions in the supports under the action of gravity, a natural property of the conical beam frame (or any other simply supported straight beam, by the way). *Trussframe is a trademark of Coronis Building Systems, Inc.

PRIMARY FRAME PRIMARY FRAME

FIGURE 4.16 Lattice structures. (Coronis building systems.)

4.8.3 Delta beam system Delta beams* are probably not to be confused with other constructions. We would have included this system in the next chapter if these triangulated, three-dimensional beams designed by Butler Manufacturing Co. were not designed and sold as a complete roof support system, but as purlins only. The joists, which are available in 1-foot increments, have a consistent depth of 251Ⲑ4 inches regardless of loading. Its upper and lower strings are made of hot-rolled steel angles; the diagonals consist of round bars. The beams have a very desirable property - lateral stability - which makes them really different as it avoids the need for purlin bracing and perhaps even the traditional horizontal roof membranes. The Delta-Joist system is best suited for buildings with load-bearing masonry or prefabricated walls where external columns and wall bracing are not required; can be adapted for non-structural bulkheads when using an optional steel frame. The system typically features a 1Ⲑ4:12 roof pitch and requires the building width to be a multiple of 4 feet. Joists can be as long as 60 feet, a distance not typically achievable with the secondary members traditionally used in metal building systems. The Delta-Joist system was designed to support Butler's proprietary roof panels with permanent seams; Beam top chords are pre-drilled for the attachment of roof clips. 4.8.4 Pole-type and Fixed-Base Column Systems Occasionally a manufacturer will propose a system that is not significantly different from competitors but costs less. Savings can sometimes be explained by the fact that all columns in the building or only the external columns are firmly founded. In this design, *Delta Joist is a registered trademark of Butler Manufacturing Co.

PRIMARY FRAME 78

CHAPTER FOUR

The action of the rigid frame is partially or fully transmitted from the knee of the frame to the base of the spine. With the first approach, the knees of the frame remain rigid, but in addition, the columns are attached to the base. The second approach relies solely on column strength; In essence, each solid column becomes a mast-like structure. The flagpole approach allows the manufacturer to not only dispense with expensive rigid structures, but also with wall and ceiling braces - all at a more affordable price. Why doesn't everyone design like this? The answer is simple: this type of design, while allowing some savings for the manufacturer, penalizes the foundations by subjecting them to high bending moments that would be absent in a fixed base scenario. Whether the fixed-base design results in a higher or lower overall construction cost depends on the specifics of the project. If planned in advance, this can be a viable option if foundation capacity can be increased cost-effectively. However, damage can occur when such a fixed-base design is proposed for a building where the foundations have already been designed - or worse, built - based on the assumption of a fixed base, as discussed in Chap. 10. Another problem with fixed base pillars: the fixation of the base is easy to assume, not easy to achieve. For example, a common industrial practice of not providing grouted leveling plates under column bases and not tightening anchor bolts, often supports the base plates at only one edge and allows for some "gap" between steel and concrete. Furthermore, the details used by steelwork fabricators to achieve a base attachment rarely follow the details used in heavy structural steelwork such as that shown in Fig. 4.17. In Fig. 4.17, the anchor bolts transmit their forces through rigid supports (also known as bolt boxes or boots) to the support tables, bypassing the baseplate itself. The simple baseplate in Fig. 4.18a will also develop some momentum, but it requires a very large one Plate thickness to prevent rotation under load. The one shown in Fig. 4.18a would leave the base plate much closer to the pinned state than in the solid state, many experienced engineers agree.3 A typical manufacturer detail for the ones shown in Chap. 12 uses eight anchor bolts and a relatively thin base plate. This design can reduce, but not completely eliminate, plate deflection under load (Fig. 4.18b) and is likely to provide less than full column fixation. Another problem that makes column fixation difficult to achieve is the fact that even the best detailed solid-based columns support foundations that tend to move under load. As in chap. 12 these foundations usually consist of shallow foundations. Large as they are, these moment-resisting or tethered shoes require some ground deformation—and therefore twisting under load—to be effective. Geoengineers understand this all too well.4 Each of the factors discussed above potentially result in less base rotation. Taken together, however, they could provide a sufficiently significant degree of rotation to render a fixed base assumption simply unrealistic in many cases.

4.9 A ROLE OF FRAME STRUCTURE All of the structural systems we have considered, except perhaps Butler's Delta Beam, require lateral bracing of the beam compression flange for full load bearing capacity. With downward loads (dead, live and snow), the top chord of the main beams is mainly loaded in compression. Fortunately, this flange supports purlins that provide the necessary bracing. However, under wind lift, it is the bottom flange that is primarily in compression. Without the help of secondary girders, the lower chord is to be stabilized against buckling by chord struts, usually made of screwed angle pieces (Figure 4.19). Similar bracing is required on inner flanges of rigid frame supports, which are normally in compression under downward loads. Bracing connects the inner flange to the wall studs (Fig. 4.20). Flange support positions are determined by the fabricator of the metal parts and need not concern the designers. However, the lack of any flange support warrants further investigation.

PRIMARY FRAME PRIMARY FRAME

FIGURE 4.17 Typical detail for column attachment used in heavy steel structure buildings.

4.10 CHOICE, CHOICE Any major manufacturer can supply most of the large frame systems discussed above. Specify which? Hopefully the information in this chapter as well as in chap. 3, will be helpful. (The dimensions and details shown in the figures are to be considered preliminary as each manufacturer has its own specifics.) Also, Figure 4.21, adapted from Means Building Construction Cost Data 1995.5, provides material and total costs per square meter for different building types.

PRIMARY FRAME 80

CHAPTER FOUR

FIGURE 4.18 Details of column bases offering questionable attachment: (a) four screws; (b) eight screws.

PRIMARY FRAME PRIMARY FRAME

FIGURE 4.19 Typical flange mount on the roof. (VP building.)

prefabricated structure. The prices, while a bit dated, are useful for comparing frame types. Generally, in a large facility that can tolerate internal columns and is unlikely to undergo drastic layout changes, a multi-span rigid frame construction should be tried first; usually offers the lowest cost. When internal supports are in question, a clear span system such as a rigid structure with a span. Smaller buildings can be economically framed with tapered beams or even canopies. Proprietary frameworks are difficult to specify for tenders without severely restricting competition and are best suited for privately negotiated work. This helps to understand what is behind the customer's approval requirements, as discussed in the previous chapter. Most customers want a speakerless plan that gives them unlimited flexibility; only when the costs of this flexibility become clear do the budgets begin to be coordinated. It is somewhat discouraging to see a building with large free spans - paid dearly - promptly subdivided by the owner's erected partitions. Whether it's noise barriers or privacy walls separating different activities, each partition could include a pillar, saving some costs. In addition, the specification of buildings with unusually long spans limits the list of bidders to the largest manufacturers capable of producing heavy structures. The architect must decide in consultation with the client whether straight, conical or other supports are suitable for the design. While a utility or manufacturing building can easily tolerate tapered columns of rigid structure, a plasterboard-clad library or retail store probably wouldn't. Attempting to encase a tapered column in plaster is usually not worth the hassle and expense; A straight speaker system might be a better fit, albeit a bit more expensive.

PRIMARY FRAME 82

CHAPTER FOUR

FIGURE 4.20 Typical column flange arm. (VP building.)

Example 4.1 Selecting Frame Type and Overhang Height Select the basic frame type and overhang height for an industrial facility with a plan area of approximately 80 feet by 250 feet without cranes. The building will have a roof pitch of 1:12 and must withstand a roof snow load of 30 psf. Multiple devices measuring 30 feet by 30 feet by 12 feet high must fit anywhere in the building. Solution A rigid frame single span steelwork system provides the best solution because: ● ● ● ●

The span of 80 feet is in the ideal range for this system. Tapered columns are acceptable. The eaves height is between 10 and 24 feet. This system offers maximum planning flexibility for device placement.

See Fig. 4.10; Look for the payload of 30 psf (actually snow) in the column and the width of 80 feet in the row. The first distance below the knee, distance G, exceeding 12 feet corresponds to an overhang height of 16 feet. However, the clearance provided at this eaves height would be very small - only 6 inches. Also, the data in the table represents only a single manufacturer and others may provide a deeper framework than shown. A conservative choice would be to move to the nearest eaves height - 20 feet.

4.11 DASHBOARD STRUCTURE The previous discussion dealt only with internal structures. What about the design of the front wall? While each manufacturer has a slightly different approach and detail to the final wall frame, the basic design is essentially the same. The function of the end wall frame is to withstand all loads applied to the end walls of the building and to support the wall beams. In buildings with flared side walls, a regular internal frame is provided at the top of the walls. This frame will withstand all vertical loads as well as lateral loads applied to the side walls and the bulkhead frame is only needed to support the wall joists. In the

PRIMARY FRAME PRIMARY FRAME

FIGURE 4.21 Typical cost per square meter of prefabricated buildings. (From Means Building Construction Cost Date 1995. Copyright R.S. Means Co., Inc., Kingston, MA, 800-334-3509. All rights reserved.)

Buildings where the end walls are not extendable, the end wall structure also supports vertical loads and contains wall bracing (or fixed base wind bars). The bulkhead frame consists of columns (posts), roof beams and corner posts, all with base plates and other accessories. Sidewall pillars are often fabricated from cold formed single or double channels of at least 14 gauge metal thickness as shown in Figs. 4.2, 4.5 and 4.22. Alternatively, headboard mullions can be manufactured from hot rolled or engineered wide flange sections. End beams are usually made from cold-formed channels, unless a regular rigid frame is used for future expansion. Girders are designed as single-span girders and are spliced at each support; A reinforcement channel may be required at each joint (Fig. 4.22).

PRIMARY FRAME 84

CHAPTER FOUR

FIGURE 4.22 Side wall frames seen from inside the building. (Star building systems.)

End wall columns are typically spaced 20 feet apart on center, a distance determined primarily by belt span capabilities. The pillar layout can start from a central pillar on the ridgeline. Where no center column is provided, two front wall columns span the ridge line. In the case of non-extensible end walls, a connection between the end wall column and the beam can be made simply by bolting the column tab to the beam web (Fig. 4.2) and connecting the purlin to the beam.

PRIMARY FRAME PRIMARY FRAME

Beam - not for the column - with a suitable angle (Fig. 4.23a). Or the support can be attached to the beam by small bulkhead connecting channels (Fig. 4.22). In both cases, a rake angle is required at the top of the purlins to support the wall cladding. With extendable sidewalls, the strut-to-beam connection requires an additional support or mounting bracket between the strut and frame strut (Fig. 4.24a) and between the sidewall and frame strut (Fig. 4.24b). Headboard straps can have a flush or bypass insert (these terms are explained and described in more detail in the next chapter). The embedded beams are designed as single-span beams that fit into the webs of the end wall pillars (Figures 4.22 and 4.23). Bypass straps are designed as continuous links; in the corners they can be connected to the supports or the crossbeams using special beams (Fig. 4.24). In some buildings with masonry, glass, or concrete walls, the curtain wall structure can extend right from the foundation to the roof. There the end wall frame may consist of a rigid free span structure only, similar to the case with extensible end walls, but without end wall beams and supports.

4.12 CERTAIN DEVELOPMENT AND MANUFACTURING ISSUES

4.12.1 One-sided welding As already mentioned, the primary structures in metallic building systems usually consist of welded plates and bars. Welding between the flanges and the web is usually done by automatic welders and only on one side (some manufacturers even use intermittent welding). Typically, fillet welds are used, except that thicker sheet metal (1 to 1.5 inches) may require partial penetration welds. There are engineers who consider such single-side welds to be structurally flawed.6 This suspicion is perhaps understandable, but there is no overt ban on single-side welds in AISC or AWS

FIGURE 4.23 Details of bulkhead structure for non-extendable bulkheads. (a) connection between rear wall support and purlin; (b) flat at the corner. (Metal construction systems.)

PRIMARY FRAME 86

CHAPTER FOUR

FIGURE 4.24 Headboard frame details for extendable headboards. (a) Connection between the column of the back wall and the rafter of the portico; (b) flat at the corner. (Metal construction systems.)

specifications. According to a testing program by Prof. Thomas M. Murray, single-side welds do not reduce the ultimate load-carrying capacity of primary frames, except at end-plate joints where seismic loads are encountered. The simulation of the cyclic seismic forces in the test program repeatedly generated local buckling, which resulted in the fracture of the one-sided welds on the frame members near the end plates. Some have found that single-sided welding may be acceptable for static loads, but not for structures subjected to lateral forces, concentrated loading, or fatigue where double-sided welds must be used.7 Of course, most rigid structures must withstand both gravity and charges withstand gravity. 4.12.2 Fabrication and Assembly Tolerances Normal fabrication and assembly tolerances for metal structure systems are given in the MBMA Handbook Section 9. It shows tolerances for cold formed shapes, engineered structural members and crane runway beams. Allowable tolerances in the MBMA manual are generally looser than those used by the AISC for fabricating and assembling structural steel. Why should these tolerances be of interest to system planners in metal construction? The main reason: Components in prefabricated frames are designed with very little margin for error. Unlike mast designs that use a limited range of frame sizes, metal frame components can be designed with near 100% efficiency. If tolerance-related eccentricities are not taken into account in the design, the frames can be overloaded under the full design load. For example, the allowable sweep magnitude MBMA (a deviation from the theoretical web position measured in the weak element direction) and curvature (a deviation from the theoretical flange position measured in the strong direction) for structural elements other than track beams is: (1⁄4 inches) L ᎏ 10

PRIMARY FRAME PRIMARY FRAME

where L is the length of the pole in feet. Thus, a 20 foot tall column has an allowable sweep of 1⁄4 ⫻ 10/20 ⫽ 0.5 inches; An 80 foot framing beam has an allowable deflection of 2.0 inches. Presumably the prop must be designed for this 0.5 inch weak axis eccentricity and the torsion in the structure caused by the 2.0 inch weak axis eccentricity must be similarly accounted for in order to to avoid overstressing under the design load. Although torsion in the beam can be relieved by kicker angles connecting the bottom flange of the beam to purlins (see Fig. 4.19), a sweep internal column cannot be simply stiffened; it must rely on its own strength to resist the eccentricity of the resulting weak axis. It should be noted that such "accidental" eccentricities are believed to be included in the AISC equations for structural steelwork and do not need to be checked for mild steel. However, as we have just said, the AISC tolerances are tighter than those of the MBMA. 4.12.3 Torsion due to member eccentricity Consideration of many frequently used details in Chap. 3 and 4 and other chapters suggest that these details sometimes seem to overlook torsional stresses. Torsion can be introduced by methods of joining structural members and their asymmetric shapes. The problem of torsion caused by design errors in intersecting elements has already been discussed in Chap. 3. Torsion is present when the sidewall supports are attached laterally to the primary frames (Figures 4.23 and 4.24), when the external masonry walls or jambs are attached to the lower wings of the eaves braces (see Figures in Chapters 7 and 10) and in many other similar cases . Unfortunately, so-called open cold-formed steel sections - those that do not form welded tubing or tubing - have inherently low torsional strength. Consequently, it is often desirable to provide diagonal flange bracing ("kickers") in the situations just described. Two examples of the proper use of end frame flange clamps are shown in Figures 2 and 3. 4.25 and 4.26. When such flange reinforcement is impractical and the torsion generating detail cannot be altered, the use of "closed" tubular sections rather than cold formed original parts should be considered.

REFERENCES 1. Steel Structures Handbook, vol. II, Connections, American Institute of Steel Construction, Chicago, IL, 1992. 2. William McGuire, Steel Structures, Prentice-Hall, Englewood Cliffs, NJ, 1968. 3. Charles J. Carter, "Fixed vs. Pinned, the Answer in Technical Questions and Answers,” Structure, June 2001, p. 50. 4. Eric E. Coustry, Answer to a reader's question about spinal fixation, Structural Engineering Forum, May-June 1996, p. 10. 5. Means Building Construction Cost Data 1995, R.S. Means Co., Kingston, MA, 1995. 6. Jeffrey S. Nawrocki, "How Fabricators Can Combat Metal Buildings", Modern Steel Construction, May 1997, p. 7. Letters to the Editor for Ref. 6, Modern Steel Construction, November 1997, pp. 29-31.

REVIEW QUESTIONS 1. Name at least three common exterior column profiles. 2 Select an eave height for the building with a 50 foot wide, single span rigid frame supporting a 40 psf live load on the roof and a 4:12 roof pitch. Minimum knee spacing required is 15 feet.

PRIMARY FRAME 88

CHAPTER FOUR

FIGURE 4.25 Rigid structure bulkhead frame with Z purlins and embedded beams. (Nucor Building Systems.)

3 Why is flange reinforcement required on bottom chords of frame beams? 4 What factors prevent the abutment from being fully fixed in the base? 5 How are the bulkhead posts attached to the end studs? 6 Which primary frame system always uses straight supports? 7 What advantages do trusses offer? 8 How can a bulkhead-beam connection be relieved of torsion?

PRIMARY FRAME PRIMARY FRAME

FIGURE 4.26 Rigid structure bulkhead frame with open I-beams and down-conductor straps. (Nucor Building Systems.)

PRIMARY FRAMEWORK

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 5

SECONDARY FRAME: CIRCLES AND SPA

5.1 INTRODUCTION Secondary structures bridge the gap between the primary frames of metallic building systems. They play a complex role that goes beyond supporting roofing and walls and transferring external loads to larger structures. Secondary structures, as these members are sometimes called, can act as flange stiffeners for the primary structure and function as part of the building's lateral load bearing system. Secondary roof elements, so-called purlins, often form a significant part of the horizontal roof panes; Secondary wall members, known as chords, are commonly found in wall bracing assemblies. A third type of secondary truss, known as eaves braces, eaves purlins or eaves flanges, acts as a partial purlin and partial flange - its top flange supports the roof panels, its core the wall cladding (Fig. 5.1). Belts, purlins and eaves braces have a similar load-bearing capacity. Since most of the secondary components commonly found in structural steel systems are made of cold formed steel, our discussion begins with some aspects relevant to the design of cold formed steel structures.

5.2 COLD-FORMED FRAME CONSTRUCTION As described in chap. 2, the primary design standard for cold-formed structures, is the American Iron and Steel Institute (AISI) specification for the design of cold-formed steel structures.1 The specification, comments, design examples, and other information make up the AISI Handbook. 2 The first edition of the specification appeared in 1946, with subsequent editions in 1960, 1968, 1980, 1986, 1989 (by Addendum), 1996, 1999 and 2000 (the last two by Supplement). The LRFD-based specification was first published in 1991.3 In 2002, the title was changed to "North American Specification for the Design of Cold-Formed Structural Members"4 to reflect the fact that many provisions of the specification are not unique to apply to the United States, but also to Canada and Mexico. Provisions common to the three countries are included in the body of the document; country-specific points are listed in the appendix. Users of the 2002 specification can choose between ASD, LRFD and LSD (limit state design) formats. (The LSD design approach is widespread outside the US.) As one can imagine, the Combined Specification doesn't look any simpler than its notoriously complex predecessors. The changes between the different editions are significant, which is reflected in the ongoing research in this field of steel construction. Because the specification's provisions are so fluid, aircraft manufacturers are challenged to meet the latest requirements. Unfortunately, some lag behind and still use previous editions. Anyone who has ever attempted to design a lightweight element within the terms of the specification has probably noticed how lengthy and complex the process was. 91 Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

SECONDARY FRAME: CIRCLES AND SPA 92

CHAPTER FIVE

Cold formed steel structures are rarely designed in most building construction offices. When such a structure is needed, engineers typically do one of two things: they uncritically rely on vendor literature, or they simply avoid any cold-formed design by specifying hot-rolled steel members and waiting for a contractor to fabricate and replace the replacement come up with the results of necessary calculations. In this chapter we limit our immersion to actual specification formulas, which may be slightly out of date by the time you read this book. Instead, we just point out a few key concepts. What makes cold formed steel construction so time consuming? First, materials suitable for cold forming are generally quite thin and therefore susceptible to localized deformation under load. (Remember how easy it is to dent a can?) This type of failure is much less of a concern when designing thicker hot rolled parts. These local deformations can take two forms: local and distorted buckling. The nature of twisting buckling (Fig. 5.2a) is not well understood, at least not as well as local buckling (Fig. 5.2b). In local buckling, part of the compressed chord and web flexes when the stresses reach a certain limit; this part then no longer carries its share of the load. During deformation buckling, the compression flange and adjacent reinforcing lip move away from their original position as a unit, which also weakens the section. Research on strain buckling continues, with some important work being done by Bambach et al.5 and Schafer and Pecoz6, among others. Second, it cannot be assumed that the chords of lightweight profiles are subject to a uniform stress distribution, as might be the case with chords of an I-beam (shear deceleration phenomenon). To account for both local buckling and shear warping, the specification uses a concept of “effective design width”, where only certain parts of the cross-section are considered effective to withstand compressive stresses (Fig. 5.3). This concept is fundamental to stress analysis and deflection calculations performed on cold-formed components.

FIGURE 5.1 Typical eaves brace.

FIGURE 5.2 Local strains of cold-formed Z-sections during bending, with top flange in compression: (a) twisting buckling; (b) local buckling. (After references 5 and 6.)

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

The effective design width depends on the stress in the bar, which of course can only be calculated once some cross-sectional properties are assumed. This "vicious circle" requires some design iterations. A common simplified but conservative method of calculating the effective width assumes that the stress level is the maximum allowable. Another complication caused by uneven stress distribution in thin, often asymmetric, sections is the lack of torsional stability. Lightweight compression and flexure elements can fail in torsional flexural buckling mode by simultaneous torsion and flexure, a failure that can occur at relatively low stress levels. In the top view, the side-bent purlins are shifted from their original position, as shown in Fig. 5.4. The maximum lateral displacement typically occurs at the center

FIGURE 5.3 Effective width concept for C and Z profiles (shaded areas are considered ineffective).

FIGURE 5.4 Purlin movement due to lateral buckling. (LGSI.)

SECONDARY FRAME: CIRCLES AND SPA 94

CHAPTER FIVE

the gap. Torsional-flexural-buckling can be avoided by keeping the compressive stresses very low or by tightening them as described later in this chapter. The complexity of designing lightweight construction elements does not end with bending and compression calculations. Tedious shear calculations are often accompanied by even more cumbersome mesh damage tests. Although web weakening fractures also occur in hot-rolled steel components, light profiles are much more susceptible. Web deformation errors, as shown in Fig. 5.5, are more likely to occur at supports where the shear stresses are greatest. Network deactivating stresses add to bending stresses and a combination of both needs to be investigated. If the web weakening stresses are too high, support reinforcements are required at the supports, in which case it is usually assumed that all reaction force is transmitted directly through the reinforcement to the primary skeleton, neglecting any structural contribution of the bar web. A small gap can be left even under the edge of a belt or purlin. Reinforcements are usually made from clip angles, metal sheets or U-pieces. In Fig. 5.6, the load is transferred from the web of a Z purlin via screws or bolts to the angle stiffener of the coupler and then from the stiffener to the beam. Some other clip designs that not only help the purlin to withstand web deformation stresses but also stabilize it laterally are described later in this chapter (Section 5.5.5). The specification recognizes that analytical methods for determining the resistance of some cold-formed structures may not always be available or practicable, and allows for the determination of resistance by loading tests for such cases. The test procedure is

FIGURE 5.5 Web destruction.

FIGURE 5.6 Angle of bearing clamp acting as web reinforcement.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

described in the Specification section titled Special Case Tests. In the 1986 edition, the test criteria were relatively clear. In particular, the item or assembly to be tested must be capable of withstanding twice the payload plus 1.5 times the continuous load (endurance test) and must not be excessively below 1.5 times the payload plus 1.0 permanent load (the deflection test). The effective section modulus and moment of inertia values were determined based on deflection and deflection measurements. Test results apply only to the sample tested. When the test was applied to the full class of sections as usual, material properties such as yield strength were measured and the test results adjusted for the ratio between the steel's nominal and actual strength. For example, if the nominal yield strength of steel is 55 ksi, but the actual is measured at 60 ksi, the test results will be reduced by 55/60 ⫽ 0.917. Otherwise they would overstate the capacity of similar elements made from steel with a yield strength of less than 60 ksi. The 1996 and later editions derive the design allowable strength of the component or assembly as the average of all test results divided by a safety factor. The latter is equal to 1.6 divided by the drag factor, which requires some calculations to determine.

5.3 COLD FORMED STEEL PURINS 5.3.1 Available sizes and shapes Cold formed C and Z purlins are the workhorses of the industry. The configurations for these elements have their origins in the press brake—they represent the two basic ways to bend sheet metal into a section with one web and two flanges. Light purlins 8 to 12 inches deep can span 25 to 30 feet and even more depending on loading, material thickness and deflection criteria. The distance between the beams is determined by the load-bearing capacity of the ceiling panels; a spacing of 5 feet is common. Appendix B contains section properties for purlin sizes offered by some manufacturers. Cold formed purlins are typically made from high strength steel. Uncoated cold formed members, still the majority, generally conform to ASTM A 570 or A 607. Galvanized purlins are occasionally supplied. The old designation for galvanized components, ASTM A 446, has been replaced by a new ASTM standard specification A 653.7. The new standard includes the zinc coating designations G60 and G90, which were formerly part of a separate ASTM A 525 standard. (The latter has been superseded by ASTM A 924, which now covers all types of metallic coatings applied by a hot dip process.) Three grades – 33, 40 and 80 – are available for structural quality (SQ) products, which correspond to the former ASTM A 446 Class A, C and E. For example, ASTM A 653 SQ Class 40 with G60 coating designation replaces the former ASTM A 446 Class C with G60 coating. The minimum yield strength for 16-gauge and heavier steel sections is typically listed as 55,000 psi, although the Light Gage Structural Institute (LGSI) bases its load charts8 on a minimum yield strength of 57,000 psi. How is it possible that LGSI can use a higher strength steel than most manufacturers for the same material specification? ASTM specifications define the minimum yield strength for steel, but the actual yield strength is often higher. It may be possible to justify the use of 57 ksi instead of 55 ksi yield strength if a reliable material inspection and testing program is maintained and only steel with a minimum actual strength of 57 ksi can be used. This is what LGSI does, although this practice is not followed in structural steelwork. Likewise, LGSI member companies have adopted slightly different profile properties for their cold-formed profiles than most manufacturers of metal building systems (Fig. 5.7). LGSI products attempt to optimize flange and lip sizes. Cold-formed purlins can be designed as single or continuous elements. The positive effects, but also the disadvantages of continuous framing are discussed in Chap. 3. The concept of continuous Z profiles was invented by Stran-Steel Corp. in 1961. introduced, already in Chap. 1 as a pioneer for prefabricated buildings. (Prior to this invention, manufacturers used cold-formed single span sections or joist beams.) Cold-formed purlins can be made by lap and continuous

SECONDARY FRAME: CIRCLES AND SPA 96

CHAPTER FIVE

FIGURE 5.7 Typical cross-sections of C- and Z-beams and purlins: (a) used by large steel fabricators; (b) offered by LGSI members.

fixation. Light cesareans can easily be lapped back-to-back; Theoretically, the Z-profiles can be nested in one another. In reality, however, it can be difficult to mount thicker traditional Z-profiles with the same flange. Zamecnik9, in examining a bearing with visibly distorted z-purlins, finds that it is "impossible to nest [the] two sections ... the purlins on the supports (see Fig. 5.8). It is worth noting that the LGSI Z-profiles have slightly unequal width flanges to facilitate splicing and provide a better fit.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

FIGURE 5.8 When Z-purlins of the same size are forced into each other at the joint, they rotate at the support. (After Zamecnik, Ref. 9.)

A reminder to designers: LGSI sections should not be imposed or specified indiscriminately on manufacturers of metallic building systems since manufacturers have aligned their production lines to their own standard members. Please inquire about availability first. Additionally, local steel erectors may not be familiar with LGSI sections and thus may not be aware of the need to turn all purlins upside down as needed to take advantage of the uneven flange design. Installers need to be educated on the benefits of using uneven flange sections and their installation techniques.

5.3.2 Design for Continuity To achieve a degree of continuity, cold formed sections are overlapped and bolted at least 2 feet apart; that is, each link protrudes at least 30 cm beyond the support (Fig. 5.9). The degree of continuity can be increased with a longer overlap distance, but at the cost of additional material used in the loop. Some research10 suggests that Z-purlin bearing capacity continues to increase until the loop length approaches midspan, while other research11 suggests that the limit is much lower.

SECONDARY FRAME: CIRCLES AND SPA 98

CHAPTER FIVE

FIGURE 5.9 Examples of (a) coils with one panel and (b) purlins with continuous panels. (Star building systems.)

A purlin can be attached to rafters in a number of ways, depending on the magnitude of the crippling stress on the purlin web. Simple bolting through the component flanges is acceptable if web weakening stress is not critical; otherwise, retaining clips are required to act as mesh reinforcement. Continuous framing requires careful consideration, although it offers significant material savings. The effects of possible problems caused by temperature changes and different settlements have already been discussed. In addition, continuous purlins in different spans are exposed to variable bending moments, even under uniform loads: The most critical bending stresses in a continuous beam occur at the ends of the spans. It follows that the last thirds of the span must have thicker sections than the inner ones. Alternatively, some manufacturers prefer to use the same purlins throughout the building and provide additional joint lengths for the purlins of the last bays. Any approach is good; A possible red flag may only be raised if the workshop drawings indicate the same purlin sections and back spacings at all points, although this could simply mean that some cost-efficiency has been lost and all purlins are kept to the size controlled by the final spans will. Another economical solution is to make the outer (final) spans smaller than the inner ones. For example, if the inside spans are 25 feet, the end spans may be 23 or 24 feet. The opposite construction, where the end spans are longer than the interior space, should be avoided, although there are circumstances where this is required. Then additional single span purlins can be added in the end spans between the lines of continuous purlins. In Fig. 5.10, two additional one-span thirds had to be placed in this way to support the load in the unusually long final spans.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

FIGURE 5.10 Two additional single-span thirds are placed between continuous thirds in the last spans of this building.

5.3.3 General design methods for purlins Not so long ago, continuous C and Z purlins were designed by lengthy - very lengthy - calculations. The increasing complexity of your project today requires the use of computers. Larger manufacturers often use proprietary design software; its smaller competitors and independent designers often use off-the-shelf computer programs. Some of these programs are listed on the CCFSS website, referred to in Ch. 2. Design input for computer and hand calculations includes detailed information on purlin size and dimensions, loading, design steel strength, length and number of spans, roof pitch, length of seams, width of girder supporting flanges and shape of lateral bracing. To save space and avoid design formulas that change from one edition of the AISI specification to the next, we refer the reader to a comprehensive design example for a four-bay continuous purlin in the Cold Formed Steel Design Handbook AISI .2 The analysis procedure follows approximately that of any continuous beam, with some specifics given below. Continuity is ensured by properly designed bolted joints.

5.3.4 Prismatic Versus Non-Prismatic Analysis In contrast to continuous steel structures with compact welded or bolted joints, cold-formed continuous purlins use overlapping members bolted over the columns (see Fig. 5.9). As a result, the purlin stiffness at the support points is twice as stiff as elsewhere. How does this additional rigidity affect the purlin construction? There are two opposing approaches to this dilemma: the first takes into account the increased stiffness of the third, the second does not. The first approach, which considers the real (double) cross-section of the purlin at the supports, is called the non-prismatic or total stiffness analysis model. the second approach

SECONDARY FRAME: CIRCLES AND SPA 100

CHAPTER FIVE

assumes that the purlin cross-section is constant and is referred to as a prismatic or reduced stiffness analysis model. The AISI specification does not prescribe which analysis model to use, but leaves it up to the designer. The prevailing non-prismatic method better reflects real-world conditions, while prismatic analysis is simpler. Most current design programs follow the non-prismatic analysis model as well as the design example in the AISI manual. The two analysis models provide similar but not identical results. For the same structure, the maximum negative bending moments of the non-prismatic analysis exceed those of the prismatic analysis, and the opposite is true for the maximum positive moments (Fig. 5.11). Note that any reduction in the maximum negative bending moment in a prismatic analysis model also reduces the design moment at the splice. The moment at the end of rotation calculated by the prismatic analysis method is smaller than that calculated by the non-prismatic method. Consequently, the purlin, designed as a “prismatic” bar, can in some cases be overstressed by a combination of moment loading and shear at the end of the bend. In fact, the end of the loop is a common critical location for purlin construction (the others include the supports and the point of maximum positive moment). Epstein et al.11 recommend using prismatic analysis only when the design is considered safe even under the non-prismatic analysis model.

FIGURE 5.11 Cold formed continuous purlins under uniform gravity loading. Notice the difference in moment diagrams for prismatic and non-prismatic analysis methods.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

101

5.3.5 Can the turning point be considered as a bracket point? The turning point is where the moment diagram changes sign, i.e. the moment is zero. This is where the compression flange, which requires lateral bracing as described in Section 5.4, ceases to be uncompressed. The adjacent part of the flange is in tension and does not require lateral reinforcement. It has been argued that the inflection point acts as a virtual purlin support, so the length of the laterally unbraced purlin can be measured from this point rather than from the end of the joint. Measuring from the end of the splice is a more conservative approach, as shown in Fig. 5.11. Measuring the unbraked length from the turning point generally reduces the length of the unsupported purlin and potentially results in a more economical design. However, the inflection point is imaginary and may change as the load changes. For example, with partial loading (Fig. 5.12) the inflection point is much closer to the support than with full uniform loading (Fig. 5.11). (In addition, the positive design bending moment is greatest under partial load.) Another argument against using the buckling point as a support point is shown in Fig. 5.13. As can be seen here, the inflection point of the bottom flap prevents a third from having

FIGURE 5.12 Cold formed continuous purlins under partial gravity loading.

SECONDARY FRAME: CIRCLES AND SPA 102

CHAPTER FIVE

Figure 5.13 A buckling point does not prevent the translation of the bottom chord of a continuous beam with laterally supported top chord under lateral torsional buckling. (Adapted from Ref. 12.)

The upper flange is continuously laterally supported against lateral movement in failure mode by lateral torsional buckling. Yura13 concludes that "not only is it incorrect to assume that an inflection point is a bracing point, but also that the bracing requirements for beams with inflection points are greater than [for] simple curvature cases".

5.3.6 Some other design assumptions for purlins In addition to the main design assumptions discussed above, some others should be mentioned. First a relatively small point: if the length of the unbraced purlin is measured from the end of the joint, where exactly is this point obtained? It is possible to consider the end of the joint as the point where the bolts are located and the purlins are physically connected. A more common approach is to place the splice end at the true end of the lap purlin, adding about 1.5 inches. more on each side to the length of the joint and correspondingly decrease the length of the unbraced purlin.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

103

Another common design assumption is to consider the seam area between the column and the end of the bend as fully laterally stiffened (as indicated in the design example in the AISI manual, among others). Despite its widespread use, this assumption only makes sense if both purlin flanges are effectively locked against rotation and translation under load in the overlapping area. Such restraint can be provided by heavy duty anti-roll clips as described in Section 5.5.5. Alternatively, the top flanges of the purlin should be reinforced laterally by the purlin cover or reinforcement. The bottom flange can be considered clamped if it is connected directly to the bracket. In reality, however, the purlins that support vertical seam roofs are not always so restrained. Too often Z-purlins are simply bolted to the corbels - and if you force them into the joint they tend to twist, as in Fig 5.8 - and are constrained at the top by nothing more than vertical seam roofs with sliding clips . Analyzing this issue, Epstein et al.11 conclude: "The currently accepted assumption that the lapped area is laterally braced appears unjustified and may significantly overestimate the calculated strength". A related assumption treats the negative moment region between the end of the loop and the inflection point as a cantilever with an unlocked free end. Of course, if one questions the stability of the faceted region itself, this assumption can also be called into question.

5.4 TIMBER STRESS: AVAILABLE SYSTEMS 5.4.1 Why purlin and chord reinforcement is necessary As civil engineers have long known, an unrestrained compression chord of a single web bending beam, even when perfectly symmetrically loaded through its core, tends to buck laterally under vertical loading . A cold-formed symmetrical (C-section) or point-symmetrical (Z-section) purlin is even more prone to buckling because its center of shear is at a different point from the point of application of the load, which is usually higher at the center of the flange. Also, the major axes of a Z-profile are inclined towards the web, and any downward loading creates a lateral component. Because of these factors, unbraced C and Z profiles tend to twist and become unstable, even under gravity loading on a perfectly horizontal roof. In pitched roofs, the purlin web is inclined from vertical, further complicating the torsion problem. The weight load acting on an inclined C or Z purlin can be broken down into components parallel and perpendicular to the roof, both of which tend to rotate the purlin, albeit in different directions, when the purlins are properly aligned, as shown in Fig. 5.14. A calculation based on the beam geometry quickly finds that the two components are equal when the slope is equal to the ratio of the purlin flange dimensions to their depth. For example for an 8.5 inch. deep with a 2.5-inch. wide, this slope is about 3.5:12. The load on torsion (torsion) increases with decreasing roof pitch and reaches its maximum with a completely flat roof, because then the force component perpendicular to the roof predominates. For a purlin with a 4:12 pitch (Fig. 5.14a), the total torsional loading effect of the two force components is quite small, but as the pitch decreases to 1Ⲑ4:12, the torsion becomes significant (Fig. 5.14b). On roofs with appreciable pitches (above 1Ⲑ2 to 12), the correct purlin orientation is upwards, as shown in Fig. 5.15. In the case of almost flat and solid roofs, the purlins are often arranged in a staggered manner (Fig. 5.16), a construction which relies on the roof acting as a pressure support between the two facing purlins. This design does not apply to vertically connected roofs because, as discussed below, vertically connected roofs may not be suitable as side bracing for purlins. Opposite purlins are sometimes used in buildings with a slope, where placing all the purlins in the same direction without a counterbalance from the opposite slope would create large strut forces. To summarize our discussion in this section and elsewhere, effective purlin and flange reinforcement must achieve the three main goals listed below. The first two criteria are taken from section D3 of the AISI1,4 specification and the third from the comment on section D3.2.1. The trusses must be designed and spaced in such a way that local injuries at the attachment points are avoided.

SECONDARY FRAME: CIRCLES AND SPA 104

CHAPTER FIVE

FIGURE 5.14 The weight load applied to the top chord of the Z-purlin can be broken down into components parallel and perpendicular to the roof: (a) steeply pitched roof (4:12); (b) flat roof (1⁄4:12).

1. To support the side flange. Depending on the direction of loading, either the inner or outer flange of the component may be in compression and lateral reinforcement may be required for both flanges. The tighter the strut spacing, the shorter the free length of the section in the weak direction. 2. To restrict or gird purlin rotation and relieve torsion. Rod rotation essentially occurs under any type of loading: gravity, wind, true vertical or tilted, as shown in Fig. 5.14. In addition, as in Chap. 3, pipes, ducts, lines and similar objects are often hung on purlins. Unfortunately, they are usually attached to the bottom chords of purlins with C-cleats or eyebolts, which puts additional torsional stress on the purlins.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

105

FIGURE 5.15 Typical purlin alignment on medium pitched roofs.

FIGURE 5.16 Possible purlin orientations on roofs with a pitch less than 1:12.

The purlin brace is designed to help relieve this sprain. (Nevertheless, it is better to secure the suspended items to the purlin net rather than the flanges. Another option is to hang them from a light steel channel placed between two adjacent purlins. Not only would the channel allow some flexibility of hook position , but also provide additional reinforcement for both purlins.) 3. To prevent lateral displacement of the entire set of purlins and roofing. Even when each panel is adequately laterally and torsionally locked, the entire single pitch roof assembly with purlins aligned in the same direction tends to move upward as a unit. The bracing system must therefore be anchored at the ends - and strong enough to quench the accumulated bracing forces. On gable roofs, this is usually achieved by channels or strong ridge angles. Alternatively, an effective roof membrane can be provided to carry and transmit all bracing forces to properly designed primary structures capable of withstanding those forces. Not every purlin bracing system in use today meets these three objectives.

5.4.2 Types of purlin and flange bracing What types of bracing are used for secondary bars? First, continuous lateral bracing can be provided by some types of metal roofs, mainly the roof types. To qualify, panels must be of the correct thickness and configuration, with fasteners that provide a continuous load path. Metal roofs with vertical seams can only provide a limited degree of purlin stiffening as discussed in Section 5.5.2. Many engineers consider this type of roof to be completely devoid of any bracing capacity. Even full fastener covers may only serve the first purpose of purlin reinforcement - lateral flange containment. The cover cannot provide torsional stability for purlins, and its strength and membrane rigidity cannot be sufficient to prevent the entire set of purlins and covers from moving laterally. Therefore, the metal roof must be supplemented with other purlin reinforcement to ensure the two remaining objectives are met.

SECONDARY FRAME: CIRCLES AND SPA 106

CHAPTER FIVE

The second type of purlin bracing is provided by discrete bracing, the spacing of which is determined by analysis. An additional purlin strut is usually provided for each point load. Perhaps the most effective discrete purlin reinforcement system is provided by spaced parallel rows of channel sections bolted between the purlins (Fig. 5.17). Channels are similar to the solid block used in timber construction. They represent a superior method of stabilizing purlins against rotation, although this type of orthosis can be more labor intensive than other systems. A less effective, but also less expensive, discrete purlin reinforcement can be provided by steel angles or braces running from eaves to eaves perpendicular to the purlins. These brackets attach to the eaves braces at each third and at the ends. Columns can be placed parallel to the roof or diagonally, running from the top chord of one purlin to the bottom chord of the next. Some of the many variations of discrete third orthotics are discussed immediately below.

5.4.3 Purlin Rafters Parallel to Roof Pitch Purlin rafters running parallel to the roof pitch from eave to eave are perhaps the most common. The simplest and cheapest to assemble are flat bracings, which are connected to purlin flanges with bolts (Fig. 5.18). However, purlin rebar must be stretched to work properly, but flat strips and round bars tend to sag and are almost useless in this condition. Also, unlike pre-cut angle profiles, flat strapping does not facilitate the alignment of purlins and can even block purlins in temporarily offset positions. Finally, since chords can only work under tension, parallel chord lines cannot fulfill the last two functions of purlin reinforcement: providing torsional stability and restraining the entire set of purlins and covers against lateral displacement under load. For these reasons, purlin reinforcement with flat chords is not particularly effective. Some manufacturers try to overcome the disadvantages of using slings by crossing the slings at regular intervals. In Fig. 5.19, the stripes are crossed at every third field of the third and at

FIGURE 5.17 Perhaps the most effective system of purlin reinforcement is rails bolted tightly together.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

FIGURE 5.18 Lateral bracing of both purlin flanges with straps. (LGSI.)

FIGURE 5.19 Cross chords. (A&S Building Systems.)

107

SECONDARY FRAME: CIRCLES AND SPA 108

CHAPTER FIVE

the summit. This design could be used for relatively narrow buildings where bracing forces are lower. On wider structures, the brace sizes required may make the straps too heavy to bend and cross easily in the field. But even on narrow roofs, the system will only work if the eave supports are suitable for purlin anchoring - a big "if" as discussed below. In contrast to flat bracing, the angle bracing can be delivered in sections tailored to the purlin spacing. Angle brackets are attached to the purlins by inserting pre-cut tabs into pre-drilled slots in purlin netting and bending the tabs with a hammer (Fig. 5.20). Simplicity and speed of assembly are the main reasons for the popularity of this design. However, it suffers from at least two major disadvantages. First, it is difficult to predict the intended anchorage capacity of a bracket connected in this manner, particularly for a flap that is not bent a full 90°. Second, the crossbars must necessarily be offset to allow field flexing of the copied legs, rather than being placed and locked in a straight line as they should be. Some manufacturers offset adjacent camber elbows by as much as 12 inches. Because of the staggering, strut-to-strut force transmission occurs through localized web flexing, an undesirable situation. A refinement of the copied leg design to allow for alignment and connection of the clamps is shown in Fig. 5.21. Instead of bending, the leg copied from the strut angle is inserted into the purlin groove and attached to the next angle piece. Some manufacturers offer horizontal slots (Fig. 5.21a), others vertical slots (Fig. 5.21b). The clamps in Fig. 5.21 are connected with two self-drilling screws. A stronger fixation can be achieved with screws (Fig. 5.22). Another, and perhaps better, approach is to forego slotted purlins altogether and attach the angles to the top and bottom of the purlin with self-tapping screws, small rivets, or screws (Fig. 5.23). 5.4.4 Anchoring the purlin braces to the eaves and ridge A simple connection by parallel waling lines or slope angles will not prevent the purlins from bending sideways as a group (as in Fig. 5.4). It also cannot prevent the entire set of purlins and covers from shifting laterally under load. Effective reinforcement requires anchoring at their ends - the ridge and eaves. At the ridge, each purlin reinforcement line should be anchored to a rigid and strong ridge rail or ridge angle (Fig. 5.24). This element is designed to resist the accumulated bracing forces from both roof slopes when compressed. It is often not enough to simply provide a different slope angle at the ridge.

FIGURE 5.20 Purlin bracing by sag angles installed in pre-drilled vertical grooves.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

FIGURE 5.21 Third side bracing with brackets connected in series. Typically, such bracing at the top and bottom flanges requires: (a) use of a horizontal purlin groove; (b) with vertical purlin groove. (LGSI.)

FIGURE 5.22 Bolted connection between tilt angles. (Modified from a drawing by Star Building Systems. The company no longer uses this detail.)

109

SECONDARY FRAME: CIRCLES AND SPA 110

CHAPTER FIVE

FIGURE 5.23 Bracing of purlins by angles attached to purlins with self-drilling screws.

FIGURE 5.24 Edge Angle. (Butler Manufacturing Co.)

At the eaves, parallel gusset plates are usually attached directly to the beams or via the purlin bearings (Figure 5.25). Some manufacturers use special adjustable cambers between the eaves brace and the first purlin (Fig. 5.26) to facilitate purlin alignment. Adjustability is provided at the end of the purlin, where the bend angle becomes a threaded rod with two nuts. Simply moving the purlin reinforcement towards the bottom of the eaves brace (Fig. 5.25a) is not very effective for purlin stability, as the torsional strength of the eaves brace can vary widely, as explained in the next section. Crossing the rail (Fig. 5.25b) has better chances

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

111

FIGURE 5.25 Anchoring parallel bracing to eaves: (a) sliding bracing at bottom chord of eave tie bar; (b) Crossing on the eaves.

successful, but only if one of the cross members can act as a strut. Because of this, crossing can be effective when using bend angles, but ineffective on flat belts. An even better construction is to place a solid block (like the gutter in Fig. 5.17) between the eaves brace and the first Z purlin. The interlock provides superior resistance to twisting and lateral buckling of both members. For wide buildings, traversing at the eaves and fastening at the ridge may not be sufficient, and angles in some inner spans may also need to be traversed to keep the purlins stable and reduce bracing forces (Fig. 5.27).

SECONDARY FRAME: CIRCLES AND SPA 112

CHAPTER FIVE

FIGURE 5.26 Adjustable lintel angle between eaves brace and first purlin. (Butler Manufacturing Co.)

5.4.5 The immovable eaves brace Eaves braces are often thought of as the anchor point for the side purlin brace - a stationary point in the horizontal direction. How valid is this assumption? The lateral stiffness of the eaves section alone is comparable to or less than that of a typical Z purlin. For example, the moment of inertia in the weak direction (Iy) of an 8 inch purlin. 14GA with 3.375 inches. on 14 ga Z purlins with 2.5 in., 1.289 in.4 flanges (Appendix B, Table B.7). In comparison, the Iy is an 8-inch. and 14 ga per LGSI is 2.475 in.4 (Appendix B, Table B.8), so its lateral stiffness is between the two-thirds Z. It's true that the eave support is attached to the siding, but how much lateral resistance can a cantilevered metal cladding section 6 to 7 feet of perimeter provided below? The closer the brace is to the eaves brace, the more horizontal resistance it offers, up to a point where the brace is directly against the eaves. Despite this, the rigidity of a single lattice wall may not be sufficient for lateral reinforcement of the entire roof structure: the lattice girder is far from having the required strength and rigidity to withstand the accumulated forces of the rafters without significant horizontal deflection. In general, it makes more sense to assume that the eaves support moves with the rest of the roof skin and purlins, rather than acting as true lateral support for them. Does the eaves brace provide torsion support for adjacent purlins? It depends on the construction details. Most of the torsional strength of the eaves brace comes from the connection of the siding to its web. Depending on how this connection is made, the torsional capacity of the eaves brace can vary from significant to minimal. When the siding is fastened with spaced fasteners near the center web depth of a trough-shaped eaves section, a drag force couple between the fasteners and the strut flanges can develop significant initial torsional drag. The initial load may gradually decrease as the leverage caused by repeated purlin movements loosens the bolts. Of course, if the eaves support is not fixed to the wall, its ability to withstand torsion is negligible.

5.4.6 Diagonal purlin braces Another purlin brace system uses steel angles placed diagonally between the top chord of one purlin and the bottom chord of the next (Figure 5.28). Here each third is part of a rigid triangle consisting of the third grid, the diagonal and the cover. The principle just works

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

113

FIGURE 5.27 Typical detail of criss-cross camber angles. (VP building.)

if the cover is frictionally connected to the purlins and can withstand compressive forces. For this reason, the diagonal bracing system can generally only be used on buildings with clip-on roofs, not on slide-clamp connection roofs. Angle brackets can be placed in pre-drilled holes in purlin networks, as in parallel gussets, or (less commonly) connected to purlins by bolts or nuts. The diagonal brackets are usually attached to the eaves brace either directly (Figure 5.28a) or via the purlin supports (Figure 5.28b). The detail of the anchoring in the eaves, used by some manufacturers, is shown in Fig. 5.29. Our earlier discussion of the lack of lateral strength of the eaves brace and its torsional capacity applies here, and the eaves detail with transom (Fig. 5.28b) should be more effective than the detail in Fig. 5.28a. With negligible torsional strength of the eaves bracing, it is easy to see how all of the inner purlins in Fig. 5.28a can rotate clockwise under gravity and the eaves brace counterclockwise (due to the tensile force exerted by the brace). As with the parallel purlin reinforcement, the diagonal reinforcement system relies on strong channels or ridge angles to withstand the cumulative reinforcement forces from both roof slopes (Fig. 5.30).

5.4.7 Diagonal strips at roof level In addition to the systems described above, purlins can be laterally stiffened by diagonal steel strips located above or below the purlins. Unlike the parallel, inclined struts that run from eaves to eaves, these struts run at an angle to the purlins and are anchored to the top chords of the main girders. Fastening to purlins and rafters is done by screws or welding. Diagonal straps can be used in combination with other types of purlin orthoses (Fig. 5.31).

SECONDARY FRAME: CIRCLES AND SPA 114

CHAPTER FIVE

FIGURE 5.28 Third diagonal brace.

The belts must be correctly positioned for tensioning. The orientation of the strips in Figure 5.31 works when the strips are above the top purlins, just below the decking. The mirror image orientation, with the strips at an angle from the frames to the bottom of the arch in Fig. 5.31, works if the strips are placed under the lower purlins. (Imagine the top flanges of the purlins moving toward the ridge and the bottom flanges moving away from the ridge, and correct alignment of the braces becomes easier to understand.) This bracing system has its origins in heavy industrial buildings with purlins made of mild steel channel. The channels were enclosed laterally by round sagging bars, sometimes arranged diagonally, similar to Fig. 5.31.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

FIGURE 5.29 Details of the diagonal stressing anchor used by some manufacturers. (A&S Building Systems.)

FIGURE 5.30 Diagonal third brace. Note the rugged channels of the ridge.

115

SECONDARY FRAME: CIRCLES AND SPA 116

CHAPTER FIVE

FIGURE 5.31 Purlins braced by diagonal steel braces placed over the top flange of the purlins.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

117

As shown in the inset, the PL force—the component of force parallel to the slope that attempts to bring the purlin down by moving its top edge toward the ridge—is opposed by a combination of tension in the strip, T, and compression of the purlin, C. When combined as vectors, T and C equal PL. The number of strips per field depends on the required distance between the purlins. The layout of Fig. 5.31 provides for a purlin brace at a quarter point of the purlin span. Note that each diagonal strip supports two-thirds laterally and is designed for a tensile force equal to twice PL. Depending on the direction of loading, either the inner or outer flange of the component may be in compression and lateral reinforcement may be required for both flanges. The tighter the strut spacing, the shorter the free length of the section in the weak direction.

8 Recommended purlin shoring system In theory, all of the commonly used purlin shoring systems discussed above can be effective if properly designed and anchored. In practice, available design details and design practices make most of them less than ideal for meeting the three required parameters listed in Section 1. 5.4.1. To reiterate, these are 1. Provide side flange bracing 2. Restrict panel rotation and relieve torsion 3. Prevent lateral shifting of the entire set of purlins and caps It is important to remember that both panel flanges must be laterally stabilized. The AISI specification,1.4, Section D3.2.2 states: When supports are supplied they shall be fixed in such a way that the cross-section is effectively limited against lateral deflection of both flanges at the ends and at all intermediate support points.

Now consider what happens when there is only a single row of parallel drafts near the top of the purlin flange that is under the vertical seam roof with slip joints, a very common design (Fig. 5.25). Even when anchored properly, this single brace line is usually placed too low to prevent rotation of the purlin under load (Fig. 5.32a) and is therefore of little use in securing the profile against rotation. The AISI specification,4 Section D3.2.1 recognizes the importance of placing the purlin as close as possible to the flange to be supported. However, according to some manufacturers, the gusset is 3 inches below the top flange - much closer to the neutral (mid-depth) axis of the purlin than the top flange, which is said to be reinforced. This construction does little to restrict the anti-rotation or lateral deflection portion of either flange. Properly anchored diagonal braces, on the other hand, offer better purlin stability, even if they are placed at some distance from the flanges (Figure 5.32b). However, we have already established that the diagonal bracing system should only be used on buildings with a continuous metal roof. These roofs are less and less specified due to the superior performance of the vertical seam roof with slip joints. For vertical seam roofs, the diagonal bracing can be supplemented by lines of parallel bracing angles that run close to the top chord. It is important to realize that purlin reinforcement is usually required even when full fixation coverage is used. As previously mentioned, the roof can provide third lateral, but not torsional, bracing. In addition, the hood membrane alone may not be strong enough to prevent the entire hood and purlin assembly from shifting laterally as a unit under load. Our recommended purlin support system consists of bolted rail locks (Fig 5.17) closely spaced. (Recommended clearances are discussed in the next section.) The rails better support both purlin flanges against displacement and rotation, and their bolts provide significantly greater connection capacity than bolts or bent tabs. To reduce deflection under load, the use of heavy-duty mounting brackets, preferably hot-rolled steel, is essential.

SECONDARY FRAME: CIRCLES AND SPA 118

CHAPTER FIVE

FIGURE 5.32 A single horizontal brace line (a) does not prevent rotation of the purlin under load, unlike properly anchored diagonal braces (b).

Alternatively, in smaller buildings, two rows of nodes fastened directly to the top and bottom purlin flanges can be used (Figure 5.23) if their angles and connection options are sufficient. The angles are crossed at the eaves and at other places in between if necessary. It is also possible to combine both approaches by bolting two rows of gusset plates in the middle panels of the purlins and rails near the eaves instead of cross braces. At the ridge, a strong channel or angle is essential to any project. The size of the struts is determined by calculation. Design forces in braces can be calculated using the formulas given in the AISI specification,1,4 Section D3.2.1 or D3.2.2. Formulas are subject to change in the future and are therefore not reproduced here, but it is important to understand that forces on large buildings are measured in thousands of pounds. (See for example the above

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

119

"Four Span Continuous Z Purlin Design Example - ASD" in the AISI manual.2) For a typical purlin span of 5 feet, the required angular section may be 2 ⫻ 2 or 2.5 ⫻ 2.5 inches. The single angle compression function sections are available from a variety of sources and computer programs. According to Walker14, the allowable capacity of a 5-foot class is 36 ∠ 2⫻ 2 ⫻ 1Ⲑ8 1400 lb and ∠ 2.5 ⫻ 2.5 ⫻ 3Ⲑ16 3400 lb. It is clear that the commonly supplied small cold formed angle profiles such as ∠ 1 ⫻ 1 or ∠ 1.5 ⫻ 1.5 will be insufficient for many applications. 5.4.9 Recommended purlin spacing How high should our recommended purlin spacing be? As the AISI stiffening formulas show, the smaller the number of purlin sleepers, the greater the forces on them. If there are too few clamps, excessive forces on them not only lead to heavy angle profiles, but can also lead to damage to purlins. When a light Z purlin is subjected to large lateral forces applied to its flange or to its web near the flange, the purlin section can fail through local buckling or distorted lip buckling (see Fig. 5.2). It can also be difficult to develop large forces through commonly available fasteners. When the purlin reinforcement members are spaced at relatively close intervals, the weak direction length of the unbraced purlin, Ly, is reduced and the strength of the purlin is maximized. However, installing too many supports increases the cost of fieldwork and materials. The optimal strut spacing can be determined through several trial and error analysis runs or through testing. The starting point for both can be the manufacturer's standards or the designer's preferences. For example, Table 5.1 lists the LGSI recommended optimum brace spacing (the lateral support spacing of the purlin).8 To maintain economics, this spacing should not be increased by more than 2 feet according to LGSI. The author prefers to specify the maximum distance between the purlin wedges in the contract documents so that all manufacturers competing for the contract follow the same rules. However, since the actual purlin sizes may not be known until they are designed by the manufacturer, Table 5.1 is of little help to the designer until then. Which purlin side bearing spacing should be specified in this case? The author's practice is to give a maximum unbraced purlin length of 5 to 6 feet or a quarter of the purlin span (Fig. 5.33), whichever is smaller. The quarter span criterion is found in earlier editions (1986 and 1989) of the AISI specification. *Walker assumes that angular loading is applied through a gusset plate, which is absent from purlin supports. This assumption introduces some eccentricity into the design and therefore reduces the allowable angular capacity. TABLE 5.1 Maximum distance between purlin tensioners (support side distances) used by LGSI

section

Lateral support spacing, feet

12 ⫻ 3,5 Z 12 ⫻ 3,0 Z 12 ⫻ 2,5 Z 10 ⫻ 3,5 Z 10 ⫻ 3,0 Z 10 ⫻ 2,5 Z 9 ⫻ 2,5 Z 8 ⫻ 3,5 Z 8 ⫻ 3,0 Z 8 ⫻ 2,5 Z 7 Z ⫻ 5 2,5 Z ⫻ 5 2,5 Z

5,1 4,4 3,4 5,3 4,5 3,7 3,8 5,5 4,7 3,9 4,0 4,2

Those: LGSI.

SECONDARY FRAME: CIRCLES AND SPA 120

CHAPTER FIVE

FIGURE 5.33 Purlin crossings spaced at quarter span points. (LGSI).

Not all manufacturers agree that close spacing between purlin supports is justified. Some would-be low bidders may not want to provide any sheeting at all, or only provide it for purlin alignment purposes, but prudent suppliers generally understand why purlins should be supported at close intervals. The purlin support must not be an area where the corners are cut. From the author's point of view, the purlin joint is one of the best possible investments in the quality of metallic construction systems.

5.5 TIMBER STRESS: OTHER ISSUES 5.5.1 Bracing for high altitude Whenever purlins are stabilized by roofing or top chord bracing, they are considered to be laterally supported only for downward loads that primarily generate compressive stresses in the top chord of the purlin. (Due to continuity effects, some areas of the top chord that are close to the supports will be in tension.) But what about situations where the wind creates upward forces and the bottom chord acts primarily as compression? According to one model15, the maximum compressive stress in lifting occurs at the intersection of the bottom chord of the purlin and its web. There the purlin is unlocked. As Tondelli16 demonstrates, the bottom chord can also be in compression with downward loading in short inside spans of evenly loaded continuous sections of purlin. A similar situation can occur at partial load on the roof (see Fig. 5.12). The behavior of purlins supported only on the tension side is extremely complex and poorly understood. The AISI specification has seen many changes in this regard; the last major change appeared in the 1989 Addendum. The 1989 Addendum design approach described here was retained in the 1996 AISI specification1 and later adapted in the 2002 North American specification for the design of cold-formed steel structures Engineering Library @ McGraw-Hill (www. digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

121

To calculate the nominal moment resistance Mn of the member with a strap attached to the attachment cover, a “reduction factor” method is given in Equation C3.1.3-1: Mn ⫽ RSe Fy where Mn ⫽ nominal moment resistance Se ⫽ RMS modulus of cross-section calculated with compression flange to yield strength Fy ⫽ design yield strength R ⫽ reduction factor. Both the 1989 Addendum and the 1996 specification gave the R-factor as follows: 0.4 for single-extension ducts 0.5 for single-extension Z-sections 0.6 for endless-extension ducts 0.7 for Z-sections Continuous span The North American The 2002 specification retained the same R-values for continuous C and Z profiles, but introduced more complicated provisions for single-span members. R values of 0.4 for C and 0.5 for Z panel sections were retained for structural members ranging in depth from 8.5 to 11.5 inches, but were increased for shallower members. For C and Z sections larger than 6.5 inches but no deeper than 8.5 inches, the R value was set to 0.65; for sections no more than 6.5 inches deep, R was increased to 0.7. In a new provision, the 2002 North American specification stated that the R-values for single span C and Z sections should be reduced by the effects of compressed insulation between the panel and the sections. The reduction ratio was 1% per inch of insulation thickness (uncompacted). No reduction for insulation was specified for continuous C and Z parts. The Mn in this equation is the limiting moment of the cross-section, not the moment caused by the maximum allowable working load, and the R-factors should not be automatically carried over to the allowable load values. However, R-factors are often used to convert the allowable values for uniform loads for fully braced conditions found in manufacturer's tables to design values for purlins braced on one side only, by simply using the tabulated values for "fully braced ' divided by 1.67, the safety factor for bending. The Reduction Factor analysis method is based on an extensive testing program17 performed within very specific limits. It does not apply to situations where any of the conditions listed in the specification are not met. In the event of non-compliance, the user is advised to use the "rational analysis method" or perform a stress test as specified in the specification. The values of the R factors are subject to change in the next editions of the specification and the engineer involved in these issues must be given a copy of the latest specification. Analysis of Eq. C3.1.3-1 and the R-factors it can be concluded that the maximum load-bearing capacity of a continuous Z-beam roof under wind load can be assumed to be approximately 70% of its load-bearing capacity for downward loads. It follows that the roof should theoretically be able to withstand a net wind uplift of approximately 70% of the design snow or live load without the need for additional reinforcement for the (lower) pressure flange. For many areas of the United States, the design wind lift load is actually less than 70% of the design snow or live load. Does this mean that when using top chord slope angles or parallel roof strips, reinforcement is not required for the bottom purlin chord? Certainly not: bottom chord reinforcement is still required to ensure purlin stability and prevent twisting under load - regardless of the load that dictates the purlin design.

5.5.2 Purlin Bracing by Vertical Seam Roofing Unlike roofing that is attached directly to purlins, metal vertical seam roofing (more specifically, vertical seam roofing with hidden brackets) is designed to expand and contract with temperature changes and movement of the supporting structure. Free to move without downloading from the Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

SECONDARY FRAME: CIRCLES AND SPA 122

CHAPTER FIVE

Holding each bag in place by screws or other fasteners is a huge improvement over the old fixed roof design. In chap. 6. In short, the cover of the vertical beams is fixed indirectly to the purlins by means of hidden clamps that allow the cover to slide in relation to the purlins. The clips fitted into the folds of the tiles are connected to the purlins with pins or screws. The maximum glide path is controlled by the clip design. Clips for trapezoidal vertical roof profiles commonly used in industrial and commercial buildings can provide 1 to 1.5 inches. of roof movement in any direction from the "neutral" position. (In the neutral position, the moving part is in the center of the clip, allowing the roof to slide smoothly up and down the slope.) The hidden clips are carefully designed not to impede the movement of the roof. The less they restrict the expansion and contraction of the roof, the better they work. However, this constructive decoupling of the roof and purlins also has a downside: an easily movable shaped sheet opposite the supporting purlins offers them little or no lateral reinforcement. This conclusion is quite obvious, and many structural designers find that the vertical seam roof is inherently incapable of providing lateral bracing for purlins. From your point of view, it may well be that the vertical seam cover causes some reinforcement of the purlin due to the friction between the cover and the purlin. Also, the roof has some membrane stiffness, probably due to friction between adjacent roof tiles. However, this friction is undesirable - and great efforts are made to minimize it structurally - if the free-floating ideal of vertical roofs is to be realized. The research and development departments of manufacturers of metal roofing are undoubtedly looking for ways to reduce, not increase, sliding friction. In fact, the best roofing systems use clip-on clips and have slippery caulk placed in the roof seams. A premium product is the so-called hinge clip, designed to further reduce binding during cover movement. The design of this clip shown in chap. 6, allows its moving part to adapt its inclination to the deflected shape of the ceiling under load. The continual improvement of roofing systems holds little promise for those who rely on roof friction to laterally support purlins.

5.5.3 Changes to AISI specification provisions related to vertical seam caps Once vertical seam caps are identified as poor purlin bracing it is advisable to inform designers of this fact and to take appropriate design safeguards. As modern vertical roofs continue to replace the old fixed-base shingles, the problem is becoming more pressing. It is expected that stringent specification requirements will be enacted forthwith to ensure that adequate purlin reinforcement is maintained when roofs with vertical seams are used. Surprisingly, exactly the opposite has happened: the latest specifications seem to accept the notion that vertical seam cover might be able to provide adequate purlin bracing. In the 1986 edition of the AISI specification (and the 1989 addendum), Section D.3.2.1, "Anchorage of bracing for gravity-loaded roof systems with gravity-loading with upper flange connected to the sheathing", the only specific type of metal cover recognised than "decking or formwork" that can provide purlin support was a fastening cover. The "deck or skin" had to be "attached so directly to the upper flanges [of the purlins] that relative movement between the deck or skin and the purlin flange was effectively prevented..." When this "face" was absent, the purlins became considered laterally unbraced and Section D.3.2.2 “No flange connected to formwork” applied instead. According to Section D.3.2.2 of the 1986–1989 editions of the specification, purlins were to be stabilized with discrete purlin supports attached to the top and bottom flanges of Sections C and Z “at the ends and at intervals not exceeding one quarter of the span to avoid tipping over at the ends and lateral deflection of a flange in any direction in intermediate braces". Additional reinforcement was required in places of concentrated stress. Accommodating these arrangements required at least three rows of purlin reinforcement at the top and bottom flanges, as well as some heavy-duty brackets that could provide translational and rotational purlin reinforcement at the supports.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

123

In contrast, Section D.3.2.1 of the 1996 Edition and the 2002 North American Specification included roofs with vertical seams in the "deck or siding" category that can provide purlin bracing. The minimum membrane stiffness requirement for this "ceiling or paneling" has been removed. The previous requirement for purlin bracing at quarter points and supports has disappeared from Section D.3.2.2. Certainly an important requirement was retained in Section D.3.2.1 that purlins connected to the “ceiling or skin” must be constrained so that “the maximum lateral displacements of the top chord relative to the purlin reaction points do not exceed the span length divided by 360.” Applying this arrangement to a typical purlin span of 25 feet, it is found that the maximum displacement of the top flange should not exceed 0.83 inches. more than 0.83 inches - it's clear that coverage alone can't provide the lateral restriction to qualify in this section. Therefore, the roof still has to be supplemented with subtle reinforcements. A new Section C3.1.4, “Beam with a Flange Attached to a Vertical Splice Roofing System” was introduced in the 1996 edition of Appendix. The new Clause C3.1.4 stated that the nominal bending strength of C and Z purlins supporting roofs with vertical seams can be assessed using one of two approaches: 1. Using discrete bracing of purlins and the provisions of Clause C3. 1.2.1, which contain rather complex formulas for determining lateral-torsional buckling resistance; or 2. Using the formula Mn ⫽ RSeFy, where Mn, Se and Fy are as described above (in Section 5.5.1) and R is a reduction factor reflecting the degree of bracing capability provided by the given roof type beam. (Not to be confused with another R-Factor used in Equation C3.1.3-1.) The R-Factor is determined by a special test called "Basic Test Method for Purlins Supporting a Vertical Seam Roof System".

5.5.4 Basic test The basic test procedure is described in the AISI Cold-Formed Steel Design Manual,2 Part VIII. The purpose of the test is to evaluate the degree of reduction in the final load-bearing capacity of purlins attached to different types of vertical seam roofs, with or without individual purlin braces, in relation to the load-bearing capacity of the same purlins with full lateral bracing. As previously mentioned, full lateral bracing can be provided by fixed roofs or by properly designed and spaced discrete bracing. The basic test is designed to apply a gravity load instead of a lifting load. The basic test uses two simply supported purlins to approximate a very complex behavior of the whole building with continuous purlins supporting a vertical seam roof. Purlins are placed on steel girders inside the test chamber. They support the vertical seam roof, clips, fasteners, roof insulation and discrete purlin bracing used in actual construction. Purlins are oriented as they would be in use - pointing in the same or opposite directions (see Figures 5.15 and 5.16). Manufacturers, of course, want the test results to apply to the full range of purlin thicknesses they manufacture, not just the thicknesses tested. This requires at least three tests for the thinnest gauge bars and three for the thickest gauge bars made on the purlin line for each purlin size. To prevent separation of the roof seams at the ends of the shingles, a small angle (1 ⫻ 1 ⫻ 1Ⲑ8 in) is placed under the shingles. For rectified purlins, a 3 ⫻ 3 ⫻ 1Ⲑ4 continuous angle can be used instead of the smaller angle to approximate the effects of the eaves brace. There must be "sufficient clearance" at the edges of the lid to allow the assembly to be moved sideways under load. The basic test setup is shown in Fig. 5.34. The assembly is loaded incrementally and the corresponding deformations are recorded. The test ends when the maximum load is reached. The final tested resistance is then compared to the flexural resistance of the fully stiffened purlins (as calculated or determined by testing) and the reduction

SECONDARY FRAME: CIRCLES AND SPA 124

CHAPTER FIVE

FIGURE 5.34 Schematic structure of the baseline test.

established factor. For a given third, the reduction factor depends on several variables; These include, above all, the type of cover and the number and type of braces. The main advantage of the basic test is its simplicity: it is much easier to test two single-span purlins than multiple spans of continuous purlins - or the entire building. However, the simplicity of the test raises some questions about its ability to predict the complex behavior of continuous purlins supporting roofs with vertical seams. Relatively minor configuration changes can produce significantly different results. For example, the continuous angle 3 ⫻ 3 ⫻ 1Ⲑ4 in Fig. 5.34 must be anchored at the ends to prevent horizontal (but not vertical) displacement, as the basic test allows? If the angle is anchored in this way, the strength of the system may be overestimated by the test and your results may not be conservative, particularly for vertical seam roofs with low membrane strength and stiffness.18 On the other hand, if the angle is not anchored, the results may be overly conservative.19 Another difficult situation arises when purlins rotate under load and their flanges go beyond 'right space' and catch on the walls of the test chamber. This probably allowed the purlins, which were prevented from turning any further, to carry some additional loads, although in real life the situation was very different. These and other issues can certainly be addressed in the future as the testing process improves. For more information on socket testing and the broader subject of purlin stability, see A Guide for Designing with Standing Seam Roof Panels, published by the American Iron and Steel Institute.19

5.5.5 Supporting purlins on supports So far we have mainly dealt with issues of supporting purlins between the supports - the frame girders and the sidewall construction - although we have recognized that the purlin stability must also be ensured on the supports. We mentioned earlier in this chapter that lightweight C and Z purlins are usually bolted directly to the struts or attached to them with purlin clips. Conventional purlin bearing clips are used to avoid mesh weakening errors and not necessarily to keep the purlin stable against rotation. In fact, the fine weave of a typical purlin clip (as in Fig. 5.6) may not laterally stabilize the purlin. The mesh reinforcements also do not prevent the purlin from twisting in the supports. Special devices designed for this are called anti-roll clips.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

125

Unlike regular stock clips, anti-roll devices are designed with some sort of diagonal legs or gussets that can resist side-to-side reactions from purlins. Some bearing and anti-roll clamps available from the manufacturer are shown in Fig. 5.35. Of these, clamps (b) and (d) are anti-roll, although the welded plate in (a) can also act in this function if thick enough and properly welded to the beam. Another anti-roll clip design is shown in Fig. 5.36. How often should anti-roll devices be spaced? Some manufacturers don't use them, probably hoping for the best; Some only use them for flat slopes. Many manufacturers supply them about every five rows of thirds, a practice that relies on coverage to stabilize intermediate thirds. Indeed, many types of fixed roofs can be considered adequate bracing, capable of transferring lateral forces from purlins stabilized by anti-roll clips to their neighbors. When this cover is used, the actual spacing of the anti-roll clips is best determined by an analysis involving a comparison of the total force on the clip versus its lateral capacity. The force in the clip is equal to the bracing force on a single purlin multiplied by the number of purlin spans between clips. The situation is different with purlins that support vertical seam roofs. As previously mentioned, this type of roofing can slip in relation to purlins and provide questionable bracing for them. This shows that roll-off protection must be provided on each row of purlins and on each support.

FIGURE 5.35 Different methods of joining purlin to beam: (a) welded purlin clamp; (b) welded clip with reinforcement; (c) angle clip; (d) Angle clip with reinforcement. (Steelox Systems, Inc.)

SECONDARY FRAME: CIRCLES AND SPA 126

CHAPTER FIVE

FIGURE 5.36 Anti-roll clip. (Star building system.)

5.5.6 Failures due to improper orthoses Why do we pay so much attention to purlin and belt orthoses? The answer is simple: Insufficiently reinforced purlins and straps are likely to fail in a real hurricane or heavy snow accumulation. Tondelli16 reports: "The most common failures in purlins and chords are due to wind uplift on walls and roofs, usually in end stalls". Additionally, a study published by the American Society of Civil Engineers of damage from the 1970 Lubbock storm and Hurricane Celia found that wind uplift “caused purlin and joist buckling in many metal structures of building systems. In many cases, buckling of the internal, unsupported, lateral flanges of purlins and beams was the initial type of damage sustained.”20 By comparison, prefabricated buildings designed in Canada have a very good loss experience, according to Factory's Canadian offices Mutual Insurance Co. Canada's National Building Code requires construction at a quarter point Tuesday. Without proper support, a theoretical possibility of a progressive collapse could become a reality, with devastating consequences. As we did in Chap. 10 Among some other examples of design flaws, a large metal building collapsed from end to end within 10 seconds. The author's own experience includes investigations of large and newly constructed prefabricated buildings that lacked purlin support and adequate anti-roll clips - and which collapsed under heavy snow accumulation. Interestingly, many in the metal fabrication industry consider some lateral flexibility of the purlins a plus (and of course too much bracing to be a vise). It is generally accepted that most fabricators of metal buildings rely on purlin roll - rotation or horizontal displacement of the upper purlin flange - to reduce notching of the attached metal roof.21 On roofs with vertical seams, purlin roll can occur when the actual expansion of the coverage increases or contraction exceeds the clip's ability to move. In essence, the cover is directly connected to the thirds in the direction of motion. As the cover moves further, it pulls the top purlin tabs with it and the purlins begin to "roll". If the direction of roof movement is reversed, the purlins can return to their original position if they have not been damaged by previous movements. If the purlins are tightly locked and do not roll easily, expansion and contraction forces tend to damage the cover or its attachments to the purlins. Consequently, some manufacturers and even some design authorities consider the purlin roller to be a positive phenomenon. Although the purlin roll can actually help preserve the integrity of the roof, it does affect the carrying capacity of the purlins. A purlin that has rotated more than a few degrees from its original position will be severely weakened and its ability to support even the original design load will be compromised. Displaced purlins with gravity loads are additionally subjected to torsional stress, which makes the situation even worse. In the author's experience, both factors tend to be neglected by some manufacturers.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

127

In addition to strength issues, the moment of inertia of rotated purlins is reduced; They tend to deflect more than one third of their original position - and rotate even more. At a certain point there will be too much spin and the purlins will just go flat and fail. As in chap. The February 10 outage could result in a rapid collapse of the entire building. To date, some steel fabricators ignore the need for proper purlin construction. If purlin reinforcement is not specified on shop drawings, it is advisable to examine the manufacturer's design approach to determine if it is conservative. The best way to avoid this situation is to specify minimum bracing requirements in the contract documents as described in Section 5.4.8. Example 5.1: Pre-selection of purlins. Choose a preliminary purlin size for a 400 ⫻ 200 foot warehouse; Use Sections LGSI Z. From the current building code load combinations (allowable loads), the maximum combined downward load, including permanent and lateral loads, is 37 psf and the combined upward load is 15 psf. The spacing of the primary rigid structures is 25 feet and the roof pitch is 1:12. There are no false ceilings and the space below is unfinished. Select a preliminary purlin shoring scheme and notes to be placed on contract drawings suitable for open bidding. Solution. Because of the size and pitch of the roof, choose a trapezoidal metal roof with a vertical seam, unless it contributes to purlin reinforcement. Use continuous purlins with full lateral bracing of both flanges; In this case, the downward load drives the project. Assuming a purlin spacing of 5 feet, the combined downward load is 37 psf ⫻ 5 ft ⫽ 185 lb/ft Use Table B.27 in Appendix B for a span of 25 feet and select 9 inch deep purlins with 2.5 inch 12 ga metal flanges (purlin designation 9 ⫻ 2.5 Z 12 G), good for 199 lb/ft, which exceeds 185 lb/ft. The maximum allowable deflection for purlins not supporting ceilings or spanning finished spaces is L/150 (see discussion in Chapter 11). The maximum tabulated deflection of the purlin is given in Table B.27 as 1.37 inches for a 199 lb/ft load. The maximum fractional deflection for 185 lb/ft is (1.37)(185) ᎏᎏ ⫽ 1.27 in 199 or 1.27 L L ᎏᎏ ⫽ ᎏᎏ ⬍ ᎏᎏ 25 ⫻ 12 235 150

(OK)

Check if there is another cheaper mother route. From Table B.33 in Appendix B, for a 25 foot span, 8 inch deep purlins with 3 inch flanges of 12 ga metal (purlin designation 8 ⫻ 3.0 Z 12 G) are good for 211 lb/ft , which is also 185 lb/ft, with a maximum deflection under that load of 1.72 inches. The design deflection of Section 8 ⫻ 3.0 Z 12 G under a load of 185 lb/ft is (1.72)(185) ᎏᎏ ⫽ 1.51 in 211 or 1.51 L L ᎏᎏ ⫽ ᎏᎏ ⬍ ᎏᎏ 25 ⫻ 12 199 150

(still alright)

Check which Tuesday segment is more economical (weighs less). According to Table B.6 in Appendix B, both sections 9 ⫻ 2.5 Z 12 G and 8 ⫻ 3.0 Z 12 G weigh the same - 5.333 lb/ft, so either section can be chosen. The deeper section has a lower carrying capacity, but a

SECONDARY FRAMEWORK: CIRCLES AND SPA 128

CHAPTER FIVE

greater rigidity may therefore be preferable when many overhead elements are expected that tend to locally distort purlins. The notes to the referenced tables state that to carry the design load, both purlin sections must be laterally braced a maximum of Ly ⫽ 75 inches or 6.25 feet apart and that purlin support braces are required at the corbels. Contract drawings must therefore state: ● ●

● ●

Design Loads and Load Combinations Maximum purlin gusset spacing at 6.25 feet on center (at quarter span points) Anti-roll clips (or at least very strong purlin support clips) on each beam Vertical deflection criteria (in this example L/150 , but see discussion in Chapter 11 for other cases) Contract drawings may include a proposed bracing scheme similar to Fig. 5.23.

5.6 OTHER TYPES OF THERMAL METAL BUILDING SYSTEMS 5.6.1 Hot Rolled Steel Beams Hot rolled steel purlins are decades ahead of modern metal building systems. A variety of industrial buildings constructed since the early 20th century have used hot rolled channels and I-beam purlins spanning the spacing between roof trusses, a prevalent style of primary roof construction at the time. Beams are still popular with many engineers for heavy industrial applications and can also be used in prefabricated metal buildings. The main advantage of hot-rolled steel beams is their higher load capacity compared to light profiles. Beams can be useful for spans in excess of 30 feet, a ceiling on the commercial use of cold-formed structures. In addition, hot rolled purlins are very suitable for heavy hanging or concentrated loads. Its main disadvantage is a relatively high cost. Hot rolled shapes used as purlins include channel sections and wide flanges. Both can be supported on primary frame supports or framed flush. The upper bearing design is generally more economical as it avoids expensive flange covers. Hot rolled purlins are often used in combination with steel decks, which can span greater distances than screwed roofs and provide better bracing. The purlin spacing depends on the load-bearing capacity of the covering. Hot-rolled purlins on pitched roofs do not escape the roof-parallel component of gravity loading (Fig. 5.37a). This component can be supported by a properly fastened and continuous roof pane (Fig. 5.37b) or by lowering rods (Fig. 5.37c), the spacing of which is determined by calculation. A typical sag bar support arrangement is shown in Fig. 5.38. The closer to the ridge, the greater the stress in the sagging bars, as the top bars carry the loads of all the thirds below. In fact, the most heavily loaded bar is the handlebar on the ridge. Because of its critical function, the tie bar is often made of plates or structural shapes rather than round bars. If slack bars are used to support the top purlin flange, it is advantageous to place them 2 to 3 inches below the top of the steel. This reduces the torsional moment Mz compared to that in Fig. 5.37c, but still allows a comfortable installation. Unlike cold-formed C and Z frames, hot-rolled steel purlins can easily be designed for lifting, whether supported between supports or not.

5.6.2 Open mesh steel girders Open mesh steel girders, also known as bar girders, can span greater distances than cold-formed and hot-rolled purlins. Open I-joists are discussed in Chap. 3 (section 3.4.1) as one of the most economical

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

FIGURE 5.37 Forces acting on wide flange purlins (U purlins are subject to an additional torsional component due to asymmetry): (a) initial force; (b) force resolution when the roof supports the upper flange, force Px resisting the deck membrane; (c) Force resolution when cover provides no support, force Px resisted by falling sticks.

FIGURE 5.38 Typical sag bar details for hot rolled purlins.

129

SECONDARY FRAME: CIRCLES AND SPA 130

CHAPTER FIVE

Contemporary framing systems for buildings with purlin spans greater than 30 feet. Pole trusses are commonly found in warehouses, factories, mail processing facilities and similar buildings that require long spans. For example, it has become quite common to design warehouses with a 40ft ⫻ 40ft column grid to accommodate a popular storage rack layout. In this section, the focus is on the challenges of integrating open lattice girders into metal construction systems. Many large structural steel fabricators produce their own lattice girders, while others order the girders from specialist suppliers. Rod supports are typically field welded to the support beams (Fig 5.39), although some bolting may be required per OSHA regulations. In regular construction (“stick-built”), stick-beams are used on essentially flat roofs and can be braced laterally by a metal deck, which also provides good membrane. In metal building systems there are pitched roofs covered with vertical seam metal roofing and the membrane action is provided by a horizontal bar or cable clamp. The design differences present a unique set of design issues, the most obvious being the need to tilt the beams from the vertical position. The slope introduces torsion into the beams, as with cold-formed purlins. The vertical load can be resolved in the directions parallel and perpendicular to the web of the beam (Figure 5.40a). Unfortunately, truss beams cannot withstand appreciable torsion because they do not have fixed webs to transfer torsional stresses, and other ways must be found to resist the force component normal to the web. If a metal platform is provided with adequate membrane stiffness, it can withstand forces perpendicular to the web. The inclined platform extends as a near-horizontal girder between the primary portal frames and no additional girder bridges are typically required beyond those required for assembly per the Steel Joist Institute (SJI)22 specification. The situation is quite different when using a vertical seam metal cover with concealed clips. We have already indicated that this type of roof is seldom able to reliably laterally stiffen cold-formed purlins, although this is very controversial. But there isn't

FIGURE 5.39 Open joist attached to the rafter of the structure. (Constructive Nucor system.)

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

131

FIGURE 5.40 Open joists on pitched roofs: (a) torsion introduced by the pitch; (b) Transverse bridge to ensure lateral stability.

Controversy Regarding Beam Manufacturers: They generally do not recognize Vertical Seam Metal Roofing (SSR) as side bracing for open joist beams, as clearly stated in their catalogs.23,24 As one of them states, “Industry standards should assume SSR reinforces systems the top chord of rafters NOT sufficient” (Triple Emphasis in the original).24 Two different construction approaches can be chosen if bar rafters are to support roofs with vertical seams. The first is to use a metal deck, as in Fig. 5.40a, and add slight hat channels that run perpendicular to your flutes across the top of the deck. Rain gutters in the hat allow the metal cover to run in the same direction as the deck.

SECONDARY FRAME: CIRCLES AND SPA 132

CHAPTER FIVE

FIGURE 5.40 (continued) (c) Providing HSS collector elements on the supports; (d) Use of tapered shims to prevent beam tilt.

The second approach is to avoid metal decks altogether and instead use spaced transverse bridges, as shown in Fig. 5.40b. The bridge stabilizes the rafters at close intervals, with the top chord angles - rather than the entire rafter section - resisting the force component parallel to the roof across the bridge. Criteria for constructing rafters in this situation are given in specification SJI25, Section 5.8(g), which also states that some roofing systems using vertical seams "cannot be relied upon to provide lateral stability to the rafters". . Designers do not need to be deeply involved in the beam construction other than to advise the beam manufacturer that there will be a metal cover with a vertical seam. It may be useful to add a note to the contract documents warning against relying on vertical seam roofs to laterally brace batten rafters. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

133

Another element to consider related to metal decks: the deck membrane is attached to the joists, not the frame joists, and side reactions build up at the joist seats before transitioning into the frame. Beam seats must be specially designed to withstand these significant lateral reactions attempting to tip them over. To draw the beam designer's attention to this important point, a corresponding note should be included in the contract documents. Alternatively, small tubular collector sections, equal in height to the depth of the beam seats, can be welded between the seats (Fig. 5.40c) to relieve them from tilting forces. Collector elements are routinely used in bar excavation. On very rare occasions, a beam manufacturer will simply refuse to go behind the inclined beam structure and may insist on using beams with vertically oriented webs. Then it is possible to add continuous tapered bars in the top chord and tapered washers in the supports to the regular (non-sloped) beams (Figure 5.40d). This design is obviously expensive but can be suitable when attaching heavy overhead items to the lower flanges of the rafters. A look at Fig. 5.40 should make it clear that inclined beams are not suitable for this, except at cross bridge locations. In this case, the design decisions boil down to having all suspended loads occur at the cross bridge locations, requiring an unusual level of coordination between multiple contractors; using the bars with extra heavy strings, an expensive proposition; or using the beams with vertical webs. Beams with vertically aligned webs can easily support hanging loads placed at plate points - and even some loads between plate points if the beam webs are modified by the added angle brackets that extend from the top chord plate points to the hooks. What happens to the rafters and especially the rafter bridge at the ridge and eaves? The ridge beams have the spacing specified by the manufacturer's standard (Fig. 5.41) and can be connected to one another using crossbars. At the eaves, the bridge is attached to the eaves braces; One of these details is shown in Fig. 5.42.

FIGURE 5.41 Bond detail at the edge of a rigid structure, with default offset distances for Z purlins and open joists. Note the custom depth (31⁄2 inches) of the open mesh beam seats offered by this manufacturer. (Nucor Building Systems.)

SECONDARY FRAME: CIRCLES AND SPA 134

CHAPTER FIVE

FIGURE 5.42 Attachment of open joist bridge to eaves brace. (Nucor Building Systems.)

5.7 COLD FORMED STEEL BEAM Cold formed C and Z beams are similar to cold formed purlins in many ways, except of course the beams are used in walls and not roofs. The discussion of available cross-sections, basic design principles, continuity effects, and bracing requirements for purlins broadly applies to cold-formed flanges as well. These and some other notable differences are summarized below.

5.7.1 Girt Inset Unlike cold-formed purlins that run across building structures to take advantage of continuity effects, luminaries can be positioned relative to columns in three different ways, called insets. In bypass use, the belts lie completely outside the supports (Fig. 5.43a). The cross straps can simply be bolted to the outer column flange if weakening of the web is not a problem, or otherwise connected to bearing clips. The semi-flush insertion requires a girdle and allows part of the girdle section to extend beyond the column (Fig. 5.43b). Bearing clamps are usually required for attachment, which are screwed to the outer flange of the support. (This design is not available from some manufacturers.)

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

135

FIGURE 5.43 Belt inserts: (a) bypass; (b) semi-flush; (c) rinse. (After Ref. 26.)

The beams can also be positioned flush with the outside of the columns (Fig. 5.43c) by bolting them to the column web with fitting angles. In fact, straps typically extend about 1 inch past the spine to accommodate erection tolerances. A close-up of the bypass belt assembly is in Fig 5.44 and the flush mount is in Fig 5.45. Note the position of the eaves support in each illustration: it simply bolts to the top of the post for the flush mount, but requires a special support bracket for the bypass configuration. Any type of belt reinforcement can easily be installed in the end walls, since there are no eaves braces; the purlins simply overhang the end jambs (Fig. 4.23a), and the angle of inclination between the purlins catches the top of the side panel. Connection details can get a bit tricky in the corners. This information is dependent on whether the corner post is a bulkhead stud or part of an extendable structure and whether the sidewall and bulkhead internal bracing are the same. An example set of details for an extendable portal corner column and baffle beams is shown in Fig. 5.46. Some further details are given in Chap. 4, figs. 4.23 and 4.24. While the staggered insertion allows for continuity, there may be a compelling reason to prefer the flush design when it comes to straight exterior columns. The wall panels are held in place by fasteners on the outside of the joists and corner bracket or similar structure fixed to the foundation wall (Fig. 5.47). Butt straps require the foundation to extend to the inside of the wall panel and the resulting space between the inside faces of the straps and studs is often unusable. The cost of this space could easily outweigh any savings that result from continuing the belt. From this point of view, embedded beams combined with straight supports can provide uniform and reasonable overall wall thickness. On the other hand, there can be important structural reasons for using branch or semi-flush beams in combination with conical or even straight supports. As in chap. 12. Column anchor bolts require a certain minimum distance to the edge of the foundation to be fully effective. This means anchor bolts cannot be placed too close to the outside edge of the foundation. Assuming the minimum number of column anchors required by OSHA regulations is four, a recessed beam system may require an excessively deep column section to accommodate the bolts. Anchor bolts are easier to place correctly when the column is away from the edge of the foundation - as is possible with sling straps.

5.7.2 Horizontal versus vertical chords The main function of chords is to transfer wind loads from the wall materials to the primary structure. More commonly, the chords are positioned horizontally to span between the columns of the structure. In this arrangement, the metal sheeting is oriented vertically and is secured to each joist, angle bracket or similar member and the eaves beam. The spacing of the straps depends on the load-bearing properties of the wall panels; it is often between 6 and 8 feet for typical single-ply fairings. Figure 5.48 shows the standard belt spacing for a manufacturer. The first brace is positioned to provide clearance for the doors.

SECONDARY FRAME: CIRCLES AND SPA 136

CHAPTER FIVE

FIGURE 5.44 Bypass Belt Unit. (Star building systems.)

At the base the panels can be supported in different ways (see chapter 7). For all details, the base coat must be used to separate the slab from the foundation concrete. A lesser known alternative is to run the joists vertically from the foundation to the eaves panel and use extra deep wall panels that extend horizontally. This design solution can go beyond the conventional and create an interesting wall treatment using traditional materials. Vertical chords, similar to widely spaced wall studs, are framed in the eaves members, which act as beams that extend between the columns and resist the wind reactions of the chords. Standard cold formed eaves are eaves

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

FIGURE 5.45 Recessed Belt Assembly. (Star building systems.)

137

SECONDARY FRAME: CIRCLES AND SPA 138

CHAPTER FIVE

FIGURE 5.46 Attaching the beams to the corners of the full frame column: (a) contouring the side beams; (b) semi-flush sidewall straps; (c) flush sidewall straps. Deviation boundary conditions are shown for all cases. (Steelox Systems Inc.)

in all likelihood not strong enough for this function unless laterally braced and hot rolled beams may be required. Care must be taken in vertical joist positioning where the building eaves exceed 30 feet, a practical span limit for cold formed frames above which a horizontal intermediate frame is likely to be required. Another issue to think about is how to copy the column flange from the Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com). Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

139

Fig. 5.47 With staggered contours, some space is lost.

FIGURE 5.48 Typical horizontal joist spacing for different eaves heights. (Star building systems.)

The stiffening is taken into account because conventional flange brackets (see Chap. 4, Fig. 4.20) cannot be used with vertical beams. Of course it is possible to make the main frames so heavy that flange reinforcement is not required, but another solution is often cheaper. The tapes remain in a horizontal position with a system of sub-carriers extending vertically between them and supporting the horizontal housing. Bottom chords can be made from cold-formed hat channels, steel pins, or similar sections.

5.7.3 Wind Columns Where primary structure columns are more than about 30 feet apart, wind columns may be provided to reduce belt spacing. Wind columns are essentially vertical intermediate chords that run from the foundation to the eaves. (These items should not be confused with wind poles, which are discussed in Chapter 3. Wind poles are attached to the ground and are used to provide lateral stability to buildings.) Wind columns are usually indicated in buildings with purlins made of steel beams with open webs. for example, the beams can span 50 feet, while cold-formed beams only need to span half that distance (Fig. 5.49). The main problem with wind columns relates to their lateral connection to the eaves. Obviously there are no building frames at this point and the typical eaves support is generally not suitable. All rights reserved. All use is subject to the terms of use provided on the website.

SECONDARY FRAME: CIRCLES AND SPA 140

CHAPTER FIVE

FIGURE 5.49 Crosswind Column. (Nucor Building Systems.)

to withstand the significant lateral reaction imposed by a wind column. Instead, a diagonal bracing system should be used to transfer the lateral reaction to the adjacent primary framework, as shown in Fig. 5.50. The strut layout here consists of round bar horizontal X-struts combined with compression struts, allowing the system to withstand both internal and external wind reactions. Further system details are in Fig. 5.51.

5.7.4 Lateral Bracing of Straps Like purlins, cold-formed straps require lateral reinforcement for strength and maximum effectiveness. Unlike purlins, flanges are usually supported at their outer flanges by a metal housing that is keyed to the foundation. Therefore, outer chords can be considered to be laterally tightened under the action of wind. Another thing is the lateral reinforcement of the inner flanges. As with purlins, external wind pressure (suction) subjects internal flanges to compressive stresses. Internal belted flanges can be reinforced by a variety of means. Office buildings often have a plasterboard interior on steel girders or hat channel bottom chords. These limbs are generally useful as lateral braces for harnesses. The interiors of more functional looking buildings can be fitted with metal cladding panels that hide otherwise exposed struts and insulation. The ceiling panels, as shown in Chap. 7, typically consist of thin corrugated or ribbed sheets fastened to chords and anchored to the foundation. If the inner flanges of the rafters are not given an architectural finish, they can be supported by arch braces (or sometimes angles) attached to the foundation and eave brace. As in Fig. 5.52, fall belts can be bent to engage the inner flanges of the belt without additional anchoring in the concrete. Fastenings - and the elements to which connections are made, such as B. the base end in Fig. 5.52 - must of course be sufficient to withstand the bracing forces. They require engineering attention to complete the load path to the foundation concrete. Example 5.2: Pre-selection of wall beams. Select a tentative size for flush insert wall studs to transport metal siding in an intermediate compartment of a large warehouse. Use LGSI Z profiles. Design wind load is 18 psf. The spacing of the primary rigid structures is 25 feet. There are no finishes or dividers on the interior walls. Solution. Due to the flush insert, the chords are designed as single-span beams. Assuming a spacing of 7 feet, the wind load on a flange is 7 feet ⫻ 18 psf ⫽ 126 lb/ft Some of the acceptable cross sections in the tables in Appendix B are: 10 ⫻ 2.5 Z 13 G (Table B .20 ) , good for 131 lb/ft, with a 2.06-in. 8 ⫻ 3.5 Z 12 G (Table B.21), good for 128 lb/ft, with a deflection of 2.36 inches. (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

141

FIGURE 5.50 Lateral brace for lateral wind column. (Nucor Building Systems.)

For beams supporting metal cladding with no inner cladding or attached equipment, horizontal deflections are not critical and a maximum ratio of L/120 can be used as discussed in Chap. 11. The 2.36-inch. (the larger of the two) corresponds to a ratio of 2.36 ᎏᎏ ⫽ L/127 ⬍ L/120 25 ⫻ 12

(OK)

SECONDARY FRAME: CIRCLES AND SPA 142

CHAPTER FIVE

FIGURE 5.51 Detail of side rail for side wall wind tube with flush insert and open web purlins. (Nucor Building Systems.)

Both sections are structurally acceptable, but the 10 ⫻ 2.5 Z 13 G section weighs less than the 8 ⫻ 3.5 Z 12 G section (4,606 versus 5,690 lb/ft) while providing greater stiffness (see Table B.6 in Appendix B). The belt carrying capacities listed are based on full lateral reinforcement of the belt flanges. Therefore, contract documents must specify the installation of interior ceiling tiles or discrete studs at a maximum spacing of 6.25 feet (in quarter span points). Note that girders in the outer spans of the building may need to be designed for a greater wind load than girders in the intermediate spans, so a larger girder size or smaller spacing may be required at these locations.

5.8 HOT ROLLED STEEL BEAMS Hot rolled steel beams are specified for the same reasons as hot rolled purlins – increased load capacity and designer familiarity (sometimes bordering on distrust of cold formed construction in general). Manufactured from U-profiles or wide flange sections, these straps can be particularly useful for bridging long distances and for framing large windows and bespoke sectional doors. Because continuity is difficult to achieve with hot rolled belts anyway, these sections are often performed with flush or semi-fluid plies. While the weight of the cold-formed C- and Z-belts is light, the hot-rolled structure is quite heavy, prone to sagging, and requires periodic support by the aptly-named drop bars. a canal

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

143

FIGURE 5.52 Attaching the slack strap. (A&S building system.)

The chord is usually analyzed as a single-span girder for wind loads and as a continuously supported girder for gravity loads, consisting of the chord's own weight and any supported wall material. The sag bars are ultimately supported by the girded eaves, which is also a hot rolled element. The issue of side bracing is just as important with hot-rolled belts as it is with cold-formed belts. With solid metal casing, slings can generally be considered to be reinforced on their outer flanges. Room surfaces such as ceiling tiles or drywall on steel beams or ceilings can support internal flanges that are otherwise considered unbraced. There are two ways to construct a sag rod supported strut with an unlocked inner flange. The first approach simply assumes that the inner flange is not supported from column to column and neglects any strengthening contribution from the sag bars. Steel sections designed under this assumption are so heavy that customers and contractors alike tend to question their design.

SECONDARY FRAME: CIRCLES AND SPA 144

CHAPTER FIVE

The second approach recognizes a limiting effect of the sag bars. It is assumed that the straps are laterally supported on each bar and the slack bars are placed as close together as is necessary for full effectiveness of the strap section. This seemingly non-conservative approach has been used for decades and has proven its worth. For those who are not convinced, here is a rationale the author previously put forward.27 In order for the inner flange to flex laterally—the most likely failure mode—it must rotate and move vertically. This movement is prevented both at the outer flange by the side fasteners and at the locations of the lowering rods. The inner and outer chords are of course connected by the web, which acts as a cantilever beam by restraining the slack chord (Fig. 5.53). It is generally accepted that the compression flange of a flexure can be considered stiffened if the stiffener can absorb approximately 2% of the compressive force in the flange. The support of the web therefore occurs when the web is strong enough to withstand this transverse bending force. (For this model to work, the flex bars must be attached to the foundation.) The effective web width for this action is a matter of engineering judgment. For circumferences greater than 8 inches, the web may be too thin for a cantilevered effect. In this case, a few continuous rows of internal flange bracing may be required, attached to the eaves and footings to supplement the slack bars. Since struts are hot rolled, your struts are usually also hot rolled, e.g. B. 3 ⫻ 1Ⲑ 4 inches. metal paneling.

5.9 NEIGHBORHOOD STRUCTURES The third type of secondary structural element, after purlins and beams, is the eaves brace. This unique structural element is located at the intersection of the roof and outer wall, thus acting as the first, third, and last (tallest) beam. The eaves height of the building is measured to the top of this component. The term brace refers to another important function of this element: it normally serves as a compression element in the wall bracing assembly and as a connection between such bracing assemblies located along the same wall. Consequently, eaves bracing is often designed for the combined effects of deflection and axial compression. Due to their importance in metal construction, eaves braces have been mentioned several times in this book. The traditional form of eaves rails is the trough section shown in Figure 5.1, which allows the cap to be attached to the top leg and the side cap to its web. Some manufacturers produce

FIGURE 5.53 Transverse bending of the network.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

145

unique shapes, like those in Fig. 5.54. Others manage to have no structure in the eaves - the first purlin and the highest chord are separate bars (Figure 5.55)! Depending on the size of the stress that weakens the web, the channel-shaped eaves clip can simply be screwed onto the main beam (Figure 5.56a) or connected to it with purlin clips (Figure 5.54). A gusset plate (Fig. 5.56b) can be attached to improve the transmission of the axial forces from strut to strut.

FIGURE 5.54 Proprietary eaves purlin kit. (VP building.)

FIGURE 5.55 A system with no eaves braces. (Steelox Systems, Inc.)

SECONDARY FRAME: CIRCLES AND SPA 146

CHAPTER FIVE

FIGURE 5.56 Connection of eaves brace to primary structure: (a) direct bolting; (b) Direct bolting to reinforcement plate. (Butler Manufacturing Co.)

Note another difference between the two figures: in Fig. 5.56 the eaves brace rests on the framing beam, a common construction in buildings with recessed joists, while in Fig. 5.54 the eaves brace is supported on a frame extension bracket for use with bypass straps. Where the eaves reinforcement is screwed directly to the beam, its upper and lower flanges usually have the same pitch as the roof - the so-called double pitch construction (Figure 5.1). However, an extension stanchion can be made with the upper level so that the eaves support can be angled with its upper flange and the lower level made - the sloping version (see Fig. 5.44). Some systems have the eaves element resting on the horizontal top of the post and also use individual drop sections as shown in Fig. 5.45. The eaves is a versatile structural element, but it is sometimes expected to do more than it can. The channel-like section works well with metal roofing and siding, but as discussed in Ch. 7, may lack the rigidity to laterally support walls made of non-metallic materials - masonry and concrete. In addition, their torsional capacity may not be sufficient to laterally support up-and-over doors, as discussed in Chap. 4 and 10. In both situations, a wide rim or tubular structural steel member would be more appropriate than a cold-formed eaves rail, although manufacturers prefer the use of cold-formed sections. Even when dealing with metal siding, it is sometimes assumed that eaves support plays a role that cannot realistically be provided. For example, as in Sect. 5.4.5, some manufacturers seem to think that simply tying the purlin gusset to an eaves brace ensures lateral support of the purlins, even though the lateral stiffness of the eaves brace section is comparable to that of the purlin (and the cantilevered side by a distance of several meters also does not provide a sufficiently stable support). The tables in Appendix B show the dimensions, cross-sectional properties, single span deflection capabilities, nominal axial values and combined axial and deflection values of typical eaves profiles manufactured by LGSI. Example 5.3: Pre-selection of an eaves brace. Select a preliminary size of dual slope eaves brace to support a 15 kip compressive load. Assume that purlins and adjacent joists are in place to withstand the wind load, so the eaves brace will carry only the axial load. Use LGSI Z-profiles The primary rigid structures are spaced 25 feet apart. Solution. The distance between the KL x frame supports is 25 feet. Suppose the eaves support is braced laterally at 6.25 feet (the quarter span point) by cross purlins or channel bracing. Using Table B.13 in Appendix B Section 8, select ⫻ 4 ⫻ 4 ⫻ 1 DSE 12G capable of withstanding a pressure of 18.3 kip (found by interpolating between the 6 and 7 foot lateral support spacings). Note that the axial load capacity of the eaves brace can be increased if the spacing of the side braces is reduced.

SECONDARY FRAME: CIRCLES AND MATERNAL FRAME SECONDARY FRAME: CIRCLE AND MATERNAL FRAME

147

REFERÊNCIAS 1. Specification for the Design of Cold-Formed Steel Structural Members, American Iron and Steel Institute, Washington, DC, 1996. 2. Cold-Formed Steel Design Manual, American Iron and Steel Institute, Washington, DC, 1996. 3. Norma de projeto para carga e especificação de projeto de projeto de fator de resistência para membros estruturais de aço formados a frio, American Iron and Steel Institute, Washington, DC, 1991. 4. Especificação norte-americana para o projeto de membros estruturais de aço formados a frio, American Iron and Steel Institute, Washington, DC, 2002. 5. Michael R. Bambach et al., „Distorional Buckling Formulas for Thin Walled Channel and Z-Sections with Return Lips“, Proceedings, 14th International Specialty Conference on Cold- Formed Steel Structures, St. Louis, MO, 15.–16. Oktober 1998, S. 21–37. 6. B. W. Schafer und T. Pekoz, „Laterally Braced Cold-Formed Steel Flexural Members with Edge Stiffened Flanges“, Proceedings, 14th International Specialty Conference on Cold-Formed Steel Structures, St. Louis, MO, 15.–16. Oktober 1998 , S. 1–20. 7. ASTM A 653-94, Especificação padrão para chapa de aço revestida com zinco (galvanizado) ou liga de ferro com zinco (galvannealed) pelo processo de imersão a quente, American Society for Testing and Materials (ASTM), West Conshohocken, PA , 1994. 8. Light Gage Structural Steel Framing System Design Handbook, Light Gage Structural Institute, Plano, TX, 1998. 9. Frank Zamecnik, „Erros na utilização de elementos de aço formados a frio“, ASCE Journal of Structural Division, vol . 106, Nr. ST12, Dezember 1980. 10. Ahmad A. Ghosn und Ralph R. Sinno, „Capacidade de Carga de Vigas de Seção Z de Aço Formadas a Frio“, ASCE Journal of Structural Engineering, vol. 122, Nr. 8, August 1996. 11. Howard I. Epstein, Erling Murtha-Smith und Jason D. Mitchell, „Analysis and Assumptions for Continuous Cold-Formed Purlins“, Practice Periodical on Structural Design and Construction, Mai 1998, S. 67 12. Eine Diskussion über das Stahlquiz, Modern Steel Construction, Januar 1997, p. 16. 13. Joseph A. Yura, „Fundamentos do Viga Bracing“, AISC Engineering Journal, 1º Trimestre, 2001, S. 11–26. 14. Wayne W. Walker, „Tabelas para ângulos únicos iguais em compressão“, AISC Engineering Journal, 2. Trimester, 1991, S. 65–68. 15. Roger A. LaBoube, „Capacidade de Elevação de Z-Purlins“, ASCE Journal of Structural Engineering, vol. 117, Nr. 4, April 1991. 16. Juan Tondelli, „Purlin and Girt Design“, Metal Architecture, Februar 1992, p. 26. 17. Roger A. LaBoube, „Schätzung der Auftriebskapazität eines leichten Stahldachsystems“, ASCE Journal of Structural Engineering, vol. 118, Nr. 3, April 1992. 18. James M. Fisher und Joe N. Nunnery, „Stability of Standing Seam Roof-Purlin Systems“, Proceedings, 13th International Specialty Conference on Cold-Formed Steel Structures, St. Louis, MO, 17 de outubro– 18, 1996, S. 455–463. 19. Um guia para projetar com painéis de telhado com costura vertical, Guia de design CF97-1, American Iron and Steel Institute, Washington, DC, 1997. 20. Joseph Minor et al., „Failures of Structures due to Extreme Winds, ”ASCE Journal da Divisão Estrutural, vol. 98, ST11, November 1972. 21. „Considerações de projeto de manutenção para edificios baixos“, Steel Design Guide Series no. 3, AISC, Chicago, IL, 1990. 22. Especificações padrão para vigas de aço de rede aberta, série K, Steel Joist Institute, Myrtle Beach, SC, 1985. 23. vigas de aço e vigas de vigas, catálogo da Vulcraft, uma divisão da Nucor Corp., Charlotte, NC, 2001. 24. Vigas de aço e vigas de viga, Katalog Nr. 307-1 von The New Columbia Joist Company, New Columbia, PA, 1998.

SECONDARY FRAME: CIRCLES AND SPA 148

CHAPTER FIVE

25. Standard Specifications for Open Web Steel Vigas, K-Series, Steel Joist Institute, Myrtle Beach, SC, 2002. 26. Design Manual, Steelox Systems Inc., Mason, OH, Mai 1992. 27. Alexander Newman, „The Answer em ‚Steel Interchange'“, Modern Steel Construction, AISC, Chicago, IL, September 1994, p. 10.

CONTROL QUESTIONS 1 Why is the third reinforcement necessary? 2 Select a preliminary size of continuous thirds spanning 25 feet and carrying the 20 psf roof load and 5 psf collateral load. Assume that the continuous load from purlins and copings is 2 psf. The maximum vertical deflection must not exceed L/240. Use profiles LGSI Z. 3 List two methods of tensioning the inner belt of cold-formed belts. 4 What are the three roles of the Eaves Member? 5 Explain the concept of effective construction width in cold formed profiles. 6 List three methods to increase the bending capacity of purlins in end spans. Why might it be necessary? 7 What does the baseline test measure? 8 What is the function of the anti-roll clips? What is a purlin roll? 9 Can the free-span roof with hidden clamps be considered as lateral bracing for open joists? If yes why? If not what could it be? 10 Explain the difference between a wind column and a wind mast.

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 6

METALLIC COVER

6.1 INTRODUCTION The function of a roof goes far beyond protecting the interior from the elements. Architecturally, a roof complements and accentuates the color and texture of the building and plays an important role in defining its character and appearance. Structurally, the roof covering withstands wind and traffic loads and can serve as reinforcement for roof purlins. Metal roofing is one of the most attractive features that prefabricated buildings have to offer and has greatly contributed to the growing popularity of metal building systems. This chapter examines available metal roofing materials and discusses the specification of metal roofing for new construction. Some particular challenges in roof renovation applications are discussed in Chap. 14. Metal roofing has been used in Europe and later in this country for centuries. Traditionally, it was hand formed into pots to be crimped or hand sewn later. The glorious golden dome that covers the Massachusetts State House was made and installed by none other than Paul Revere.1 Designed by Charles Bulfinch and completed in 1797, this building is still recognized as one of Boston's top five buildings. Popular roofing materials of the turn of the century included terne roofs made of steel coated with an alloy of 4 parts lead and 1 part tin. Unfortunately, those first Terne roofs eventually rusted and had to be painted. Contemporary metal roofs are a far cry from their predecessors. Today's products offer a long, maintenance-free lifespan, reflected in 20-year warranties; The best can last half a century with some regular maintenance and the occasional repair. In recent years, metal roofing has been installed at an annual rate of approximately 2 billion square feet (Ref. 2). Occasionally, prefabricated buildings are covered with non-metallic roofs - constructed or membrane. This fact can be due to several reasons, from the architect's desire to blend into the environment where metal roofing can be unusual, to the homeowner's hard-headed preference to avoid the metal-enhanced rain noise. In buildings that are no longer based on the one-craft concept, such as with beam constructions or hot-rolled purlins and steel roofs, cassette or foil roofs can be integrated more easily. Design guidelines for non-metallic roofs are widely used and will not be repeated here.

6.2 MAIN TYPES OF METAL ROOFS 6.2.1 'Watertight' versus 'watertight' roofs Basically, metal roofs can be classified according to how they resist the ingress of water. Water roofs, or "hydrokinetic" roofs, are functionally similar to roof tiles - they rely on a steep pitch to quickly drain rainwater. As with shingles, the minimum pitch for this roofing material is 4:12, although a pitch of 3:12 is often considered acceptable. Water drain covers are usually put on

149 Conveyed from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

METAL COVER 150

CHAPTER SIX

substrate, such as B. 30 lb. roofing felt, and is sometimes separated from it by an anti-friction paper slip sheet. In contrast, water barrier or waterproof roofs are designed to function under occasional standing water. Because of this fact, these roofs are sometimes referred to as "hydrostatic". Another name used for this roof is "low pitch" as opposed to the "high pitch" watershed roof. However, the system is not designed to be completely leak free during prolonged water immersion and still requires a minimum roof pitch for best performance. Tobiasson and Buska3 recommend a minimum pitch of 1:12 (1 in/ft) for "watertight" roofs and state that pitches greater than the minimum work better, especially in cold regions. Others find slopes as low as 1Ⲑ4:12 appropriate. In any case, as Ref. 4 points out, metal roofs with a water barrier “are generally not watertight in their valleys, eaves, ridges, rakes and penetrations”.

6.2.2 Architectural Versus Structural Roof The terms architectural and structural are somewhat misleading as each type of metal roof serves an architectural purpose and is available in a variety of finishes and profiles. The main difference between the two types is as follows: Architectural (or "non-structural") roofs rely on structural support provided by decks or spaced sub-purlins such as support rails, while structural roofs can bridge the distance between roof purlins themselves. In practice, the architectural covering resembles a waterproofing coating, or more precisely, a hydrographic covering. True to its name, architectural roofing can be used to create dramatic visual effects not possible with other types of roofing. It can be installed on very steep slopes, even vertically, although good sealing and strong structural supports are essential for installations on steep slopes (Fig. 6.1). So-called special forms of architectural roofing are based on clay tiles, roofs and roof tiles, but are supplied as panels. Individual metal tiles are also available to be used in place of the traditional variant.

FIGURE 6.1 The architectural roof provides a bold visual effect.

METALLIC COVER METALLIC COVER

151

Roof constructions are often referred to as "waterproof", although these two terms refer to different concepts. Structural coverage can be used on flat slopes, perhaps as low as 1Ⲑ4:12, although a steeper slope is preferred, as just mentioned. Structural roofing can be considered a form of roofing and as such must meet certain wind lift criteria and support the worker's weight (250 lb). The architectural cover does not have to meet these requirements. Both types of metal roofing typically weigh only 1 to 2 lb/ft2. As in chap. 3, the wind load is not uniform from one part of the roof to another. For example, the load is much greater along the perimeter of the roof, and sometimes along the ridge. Instead of using thicker structural roof panels in areas with high local loads, it is better to place the nearest purlins closer together. A common design involves supporting structural roofs 5 feet in the center of the roof span, but only half that in areas of greatest wind loading, such as attic. B. 10 feet or more from the roof edges. The contract documents should state the maximum deflection criteria for structural roofing. For steel roofs the L/180 limit is appropriate, although there are circumstances where a more stringent or softer limit may be warranted. A limit of L/60 is often specified for aluminum roofs, which have a significantly lower modulus of elasticity than steel.5 The subject of vertical deflection limits for roofs is dealt with in Chap. 11. The structural design of metal roofs follows that in Chap. 5. Some engineers specify the minimum properties of the roof section—the moment of inertia and the modulus of the section—right on the contract drawings.

6.2.3 Classification according to the type of attachment and direction of execution Metal roofs can also be classified according to the type of attachment to the supports. Fixed roofs are attached directly to purlins, usually by screws or rivets. On the other hand, the vertical seam cover is connected indirectly by hidden staples formed in the seams. It is more accurate to refer to this product as a "metallic cover with concealed fastening and vertical seam" to distinguish it from all other vertical seam covers ("standing") described in the following section, but the heavy term has not gained widespread popularity. The United States Government's Uniform Installation Guidelines refer to it as the "Permanent Seam Metal Roof Structural System (SSSMR)". 6 In this book we call the concealed fixing roof simply the vertical seam roof and the other type of vertical seam roof. It should be noted that the first vertical seam metal roof, introduced by Armco Buildings in 1934, featured exposed fasteners. It wasn't until 1969 that the hidden clip design was introduced by the Butler Manufacturing Company. A separate type of concealed roofing is insulated structural panels, also known as foam core sandwich panels. These roof panels consist of two layers of formed sheet metal with insulation in between as explained in Section 1.6.6. The metal roof consists of ribbed panels with joints usually located along the slope. An exception to this rule is the horizontal Bermuda cover. The panels of this unique canopy attach to brackets with hidden clips and resemble louvers with openings ranging from 9.5" to 11.5". Accumulation.

6.3 VARIOUS STITCHING CONFIGURATIONS Contemporary cover panels are available in the following stitching configurations. 1. The felled seam, typically found on cinched roofs, offers the simplest and most economical design (Fig. 6.2a). The edges of the corrugated sheets are simply overlapped, provided with a bead of sealant and fixed to the purlins. Despite being economical, the closures of the polished panels are exposed to the weather – and the view. This system lacks a certain degree of sophistication and is reserved for relatively simple functional structures.

METAL COVER 152

CHAPTER SIX

FIGURE 6.2 Different seam configurations: (a) overlapping panels; (b) flat seams; (c) batten seams; (d) vertical vertical stitching. (Adapted from Means Square Foot Cost Data 1995. Copyright R.S. Means Co., Inc., Kingston, MA, 617-585-7880, all rights reserved.)

2. The flat seam is formed by folding the sides of two adjacent tiles 180° and pinning them together (Fig. 6.2b). It's relatively rare. 3. The lath seam originated in the days of hand molding when the sides of two adjacent panels were folded up, separated by a wooden lath and covered with a snap cap (Fig. 6.2c). Most modern batten systems dispense with wood but retain the metal batten design (Fig. 6.3). 4. Roofs with vertical seams (vertical seam) have the seams raised 2 to 3 inches above the water bearing portion of the flat sheet (Fig 6.2d). The picture shows the so-called Pittsburgh Double Lock, a 360° rolled seam that resembles the seam on a tin can. Other types of vertical seams include those that simply snap together, usually with a sealant in between. Most modern metal roofing systems use vertically seamed metal roofing with concealed attachment, which is discussed in Section 1. 6.2.3. A 2001 poll2 of Metal Architecture readers found that about 80% of them specified roofs with vertical seams; 29% coverage with exposed grip ribs; 22% corrugated iron; 18% batten seams; sandwich panels with 12% foam core; 8% panels with tile, vibrating and shingle profiles; and more than 7% custom formed metal tiles. Metal panels can be made to resemble any traditional roof type - even thatch. Most panels are manufactured from pre-coated light alloy coils or 'pre-formed' at the factory where great care is taken to handle the coils to preserve the finish during forming. Attempts to apply coatings after molding typically result in quality issues related to color, thickness uniformity, and durability.7 Recently, portable roll-forming machines have become available, and many types of panels can now be formed on-site. However, factory forming will likely still provide better surface quality.

METALLIC COVER METALLIC COVER

153

6.4 PREVIOUS SALES COVERAGE 6.4.1 Pros and cons of the system Lap seam cover is the oldest design approach to metal roofing. The system remains a popular choice for small to medium sized industrial buildings and warehouses due to its low cost, simplicity and ease of construction. This structural covering has some membrane capacity and in many cases can provide lateral flange support for purlins. Fastening to purlins is usually done with self-tapping or self-drilling screws; Some manufacturers also use locking rivets or proprietary fasteners. Transverse attachment roofs suffer from two major disadvantages. First, the cover is punctured by fasteners, and any penetration is a potential leak in the cargo area. The only protection is a rubber or neoprene washer under the bolt head. The second problem is that the cover is hampered by thermal expansion and contraction fasteners. As in chap. 3, a long piece of metal may warp if temperature stresses become too high. Even if this is not the case, the metal around the connecting bolts will rupture from repeated expansion and contraction, causing leaks.8 Or the constant lateral movement can loosen the bolts enough that the roof can be blown off in a hurricane. To limit thermal stress build-up, buildings with solid roofs should not exceed approximately 60 feet in width. On such smaller buildings, fixed roofs may have slat joints. best chance of survival. (Carlisle Engineered Metals.) Another potential problem with bolted roofs is metal fatigue. Xu9 and others showed that this type of roof can fail locally through cracks around the fasteners when subjected to severe fluctuating wind loads. Lynn and Stathopoulos10 concluded that wind-induced fatigue was the sole possible cause of several permanent roof failures in Australia during Cyclone Tracy in 1974. One study attempted to identify the areas most susceptible to fatigue damage on hipped and gabled roofs. On gabled roofs, these areas appear to be near the gable ends and ridge; on hipped roofs they are close to the side walls and ridges.11

6.4.2 Roofing Products Continuously attached roofs are typically 1 to 2 inches in size. deep and made of 26 to 24 gauge steel. 24 gauge material has better dimensional stability and impact resistance. Manufacturers provide load charts for different panel configurations such as: 6.4 and 6.5 to facilitate roof selection taking into account traffic or snow loads, purlin spacing and wind rate. Another selection criterion concerns insurance requirements that can really control the project. For example, Factory Mutual often requires panels of greater depth or thickness than required for strength alone.12 Some manufacturers attempt to overcome the vulnerability of thermally bonded roofs by providing slotted holes in the panels. For example, Butler Manufacturing Company's Butlerib II roofing system uses pre-drilled slotted holes in the bottom plate of the end connection and regular holes in the top plate (Fig. 6.6a). A typical panel is shown in Fig. 6.6b. The manufacturer points out that they have taken extraordinary measures to perfect this system by providing a long return shank that increases dimensional stability during roof traffic, by using constantly tightening locking rivets instead of self-tapping screws and by incorporating a special gasket in the seam (Fig. 6.6c). These steps resulted in a premium mounting system with an exclusive 10-year waterproof guarantee. Some other manufacturers try to justify the lack of slotted holes in their panels, betting on the purlin roll - slight rotation under thermal stress. The purlin roll, while quite real, exists primarily on roofs with cold-formed C or Z profiles without top flange bracing. As in chap. 5, the practice of relying on the third throw raises some serious questions.

METAL COVER 154

CHAPTER SIX

FIGURE 6.4 MBCI R panel. (MBK.)

METALLIC COVER METALLIC COVER

155

FIGURE 6.5 Style Rib Coating and Covering. (Central.)

6.4.3 Fasteners As previously mentioned, direct attach covers are typically attached to purlins with self-tapping or self-drilling screws, with some manufacturers using locking rivets or proprietary fasteners (Fig. 6.7). Self-tapping screws, living up to their name, can be inserted directly through panels and purlins, while self-tapping screws require pilot holes to be drilled first. For #14 self-tapping screws in Fig. 6.7, the manufacturer recommends drilling 1Ⲑ4-in. in the top plate and smaller holes (1Ⲑ8 inch for panel-to-panel mounting) in the bottom plate. The threads of the fastener engage only the bottom member and tightening the screw joins the two panels together.13 Self-tapping hex screws have largely replaced self-tapping screws in metal-to-metal fasteners due to their faster installation, but self-tapping screws are in metal -Wood joints most popular.14 The size and spacing of fasteners, usually specified by the manufacturer, depend on the forces they are designed to withstand. In areas of the roof with strong wind loads (as shown in Chapter 3), the fastening distances can be smaller than in the roof area. If the building owner intends to purchase property insurance from a member of Factory Mutual Systems, smaller fixture spacing (and sometimes a stronger roof structure) may be required.

METAL COATING 156

CHAPTER SIX

FIGURE 6.6 Exposed mounting plates: Butlerib II from Butler Manufacturing. (a) Pre-drilled panels and purlins ensure correct alignment. (b) Seam details with locking rivet. (c) Cross-section of the plate. (Butler Manufacturing Co.)

There are three main failure modes for self-drilling screws: skipping, pull-out and shearing (Fig. 6.8). In a tensile fracture, the bolt loses its grip, while in a stall fracture, the material around the bolt breaks. For thicker materials (from 22 ga. to 1Ⲑ4 in. thick), draw capacity usually controls; For thinner gauges, a pullover could be the norm. Fastener pull-out capacity is primarily dependent on drill bit size, shank diameter, and threads per inch. The capacity of the sweater mainly depends on the head style and to a lesser extent on the size of the drill bit.14

6.4.4 Leakage and Corrosion Protection To minimize the susceptibility of fastener covers to leakage at the fastener points, rubber or neoprene washers are provided under the fastener heads. Unfortunately, this protective measure is only as good as the installer's job. In order for the sealing washers to function properly, the screws must be screwed in to the correct depth. This is done with electric screwdrivers or similar tools with precise depth control.

METALLIC COVER METALLIC COVER

157

FIGURE 6.7 Fasteners used by a fabricator to fasten roofs fastened by purlins and to each other. (The scrubolt is only used for purlin attachment.) (Butler Manufacturing Co.)

FIGURE 6.8 Different failure mechanisms for bolted fasteners. (MBMA.)

With a correctly installed screw, the neoprene washer will be slightly visible under the edge of the metal washer (Fig 6.9a). If the neoprene is not visible under the metal (Fig. 6.9b), the screw is probably not tight enough, but if the neoprene looks 'crumpled' (Fig. 6.9c), the screw may be over-tightened.15 An overdrive fastener can create pits in the metal plate and invite water to pool in the pit, making the situation worse. Screws must be placed perpendicular to the panel, otherwise the neoprene will be squeezed on one side and under-compressed on the opposite side. Exposed fasteners without a corrosion-resistant coating cause trouble. The best are made of stainless steel or aluminum; Galvanized or cadmium plated screws should be left indoors. To reduce complaints about fastener visibility, exposed fasteners can be fitted with color-matched head fairings (or fitted with colored plastic caps, a once popular but now largely obsolete solution).

METAL COVER 158

CHAPTER SIX

FIGURE 6.9 Proper and improper installation of neoprene washers: (a) With a properly installed washer, the neoprene is slightly visible under the metal washer; (b) the washer is not compressed enough; (c) Washer is too tight. (Adapted from Ref. 15.)

A popular style of hex head bolt has a stainless steel cap or cast zinc-aluminum alloy cast onto the carbon steel shank. According to the manufacturer, these round-head fasteners combine excellent corrosion protection from their exposed heads with the hardness of their heat-treated shanks, allowing them to be used in self-drilling screws. On the other hand, fasteners made entirely of 18-8 or 300 series stainless steel cannot be heat treated and therefore cannot match the hardness of heat treated carbon steel. Completely stainless fasteners are typically self-tapping screws. 410 stainless steel can be heat treated, but doing so can make it susceptible to red rust. The cast zinc-aluminum alloy is similar in composition to Galvalume and is chemically compatible with Galvalume-coated panels. The alloy can be color matched to match virtually any metal panel.16 The integrity of pinned roofs can be improved through the proper application of sealing tape applied to the panel seams within the side plies and at the overlapping end plies. Sealants must be fully embedded in the metal as they can be damaged by UV radiation. Sealants recommended by panel manufacturers are typically butyl, aliphatic polyurethane, high solids acrylic and neutral cure silicone.17 Many issues associated with exposed fasteners, such as visibility and corrosion, can be addressed with concealed fastening systems. Since these systems are the most commonly chosen as wall materials, their discussion is deferred to the next chapter.

6.4.5 Future The shortcomings of transversely attached metal roofs are serious but not fatal as long as the roof width is kept small and the installer's work is good. Despite this, the vertical seam roof is a much better product and has already replaced the bolt-on variant in most major projects. However, another trend in building design that could finally make roof anchoring obsolete: the need for better insulated buildings. As in chap. 8, modern building codes require high levels of energy efficiency and the old “hourglass” method of insulating metal buildings may no longer be satisfactory. With this method, the fiberglass insulation is simply placed over the purlins and this secures the roof to the purlins. The insulation is pushed into the supports (hence the name "hourglass") and its overall thermal performance is greatly reduced. The most popular method of improvement is to place so-called thermoblocks - rigid insulation strips - between the roof and the purlins. With vertical seam roofs, the thermoblocks are placed under the roof, but the hidden support brackets are attached directly to the steel purlins. On permanent roofs, rigid insulation, if used, must run the full length of the purlins. The metal plates must then be fastened through the rigid insulation. It goes without saying that the two faces connected by bolts must be in contact, but in this case this is not possible. As the metal plates expand and contract with temperature changes, loosen the screws by rocking them back and forth (Fig. 6.10). The next strong storm is likely to have devastating consequences. Pop rivets or small screws would probably work better, but they are more expensive to install than self-tapping screws, and their use would detract from the latch cover's main advantage: its low cost. Two other advantages of this roofing system - membrane capability and purlin reinforcement capability - would also be questioned. Instead of thermoblocks, one described in Chap. 8 could presumably be used, but this would also negate the cost advantages of fixed roofs. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

METALLIC COVER METALLIC COVER

159

FIGURE 6.10 Rigid insulation placed under mounting covers will cause bolts to loosen with constant movement of the cover.

6.5 STRUCTURAL ROOF WITH SALES 6.5.1 System Components The vertical seam structure metal roof is a vast improvement over the fixed variant. Rather than simply being folded and screwed together, the raised seams of adjacent vertical seam panels are formed on site by a portable sewing machine or, less reliably, by hand tools. Commonly, a factory applied sealant is placed in the corrugation of the female seam. To accommodate expansion and contraction, the panels are attached to purlins by concealed clips, which allow the cover to move (Fig. 6.11). Some common clip designs are shown in Fig. 6.12. Clips usually come in at least two versions - high and low. Tall clamps are used in combination with thermal blocks placed above the base of the clamp and under the roof, and the height of the clamp depends on the thickness of the blocks. High clips also allow air to circulate between the purlins and the decking. Despite visual differences, all clips consist of two parts - the rigid base, which is attached to the purlin, and the movable insert, which is rolled into the rabbet. Clips are usually self-centering, i. H. preset to allow equal tilt movement up and down. The amount of movement the clip allows depends on the length of the slot and the size of the insert in it. One of the best designs is the so-called hinge clip, which is designed to compensate for misaligned purlins.18 The clip, introduced by Elco Industries and now offered by MBCI, is shown in Figure 6.12 (lower right corner). Some other features found on higher-end systems are stainless steel rather than zinc-plated, floating clips and inserts, and pre-drilled holes in panels and purlins that reduce panel misalignment. This pre-drilling, if any, is reason enough to purchase a complete metalwork system from one manufacturer, rather than mixing and matching components from multiple vendors. The most common sewing settings are shown in Fig. 6.13. As it states, there are two distinct groups - vertical and trapezoidal. Both types have their supporters from different manufacturers. The trapezoidal seam is more popular, in part because it allows the clip to be easily concealed, and Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

FIGURE 6.11 Free-span roof structure with trapezoidal profile, hidden clamps and purlins. (Star building systems.)

METALLIC COVER

160 Conveyed from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

METALLIC COVER METALLIC COVER

161

FIGURE 6.12 Common vertical seam roof clip constructions. (From Nimtz,18 courtesy of Metal Architecture.)

partly because it is better able to accommodate the thermal expansion and contraction of the roof perpendicular to the slope (Fig. 6.14). Dimple spacing and plate width vary between manufacturers. Some panels with widely spaced seams have approximately 6-inch cross grooves down the center to improve rigidity and passage, as well as reduce vibration and wind noise. An example of the properties and configuration of vertical seam panels is shown in Fig. 6.15. After the corrugated sheets are sewn and clipped together, the individual sheets become parts of the metal roof skin, which moves as a unit with temperature changes. Clip mobility and expansion details limit the maximum uninterrupted roof width to approximately 200 feet, beyond which staggered expansion joints are required. A typical detail of the stepped roof expansion joint is shown in Fig. 6.16.

METAL COVER 162

CHAPTER SIX

FIGURE 6.13 Typical panel joints for roofs with vertical joints. (From Nimtz,18 courtesy of Metal Architecture.)

FIGURE 6.14 Trapezoidal seams allow for easy concealment of the clips and accommodate transverse panel movement. (Butler Manufacturing Co.)

METALLIC COVER METALLIC COVER

FIGURE 6.15 Example properties of a roof panel with a vertical seam. (MBK.)

163

METAL COVER 164

CHAPTER SIX

FIGURE 6.16 Details of cantilever roof with trapezoidal profile on roof step. (MBMA.)

METALLIC COVER METALLIC COVER

165

Standard plate lengths vary between manufacturers. The length is best kept under 40 feet, a practical limit imposed by shipping and handling restrictions, although panels up to 60 feet can be shipped by rail. In the case of wider roofs, the panels must be connected to one another. End connections, which are normally staggered from panel to panel, can be supported by special clamping plates and pre-drilled holes (Fig. 6.17). This detail avoids direct panel-to-yoke connections that limit movement. Edge details deserve special attention as edges, along with roof penetrations, represent the majority of metal roofing problems. With vertical seam roofs, through-roof fastenings are not completely eliminated - after all, the roof has to be fastened to supports somewhere with a form fit - but their number is reduced by approx. 80%. Fastening (fixing) of the panels is usually carried out on the eaves brace, which allows the panels to expand to the ridge covered with a flexible cover.

6.5.2 Overcoming System Limitations The greatest disadvantage of the structural vertical seam roof system can be attributed to its greatest advantage - the ability to move. In the absence of a direct attachment to the supports, the cover provides little or no lateral reinforcement for the purlins and offers little membrane action. Wherever this type of roof is used, separate purlin reinforcement and separate horizontal membrane construction is required. For architectural roofs that require a bearing surface, both functions can be fulfilled by a metal decking pad. The metal deck can also be used under structural roofs, but most manufacturers prefer purlin gussets and batten membranes. Alternatively, some manufacturers try to solve this problem by offering separate ceiling tiles that offer limited reinforcement and membrane action. However, to be really useful, cladding panels need to be fairly stiff, maybe as stiff as

FIGURE 6.17 Plate end splice. (Butler Manufacturing Co.)

METAL COVER 166

CHAPTER SIX

B. as a steel cover, which is not usually the case. In addition, cladding panels can only laterally stiffen the lower purlin flanges. They do little to reinforce their top flanges or provide torsional stability. (Because siding panels are more commonly used on metal walls than roofs, they are discussed in Chapter 7.) Other vendors simply claim that their vertical seam roof is different and can serve as purlin reinforcement and bulkhead—without providing any proof. . Another major disadvantage of roofs with vertical seams is that without many design complications, they are best suited for rectangular buildings. For example, many roof openings require many fixed connections around the edges, which can interfere with the roof's "floating". Likewise, complex plans tend to create situations for which standard design details geared towards a simple fixed end assumption of final extent are inadequate. Non-rectangular roof panels not only require expensive on-site cutting and adjustments, but can also compromise available fastener, sealant and finishing details. Even a rectangular floor plan can require a great deal of design ingenuity to allow for roof movement at corners and other critical spots. In some systems, a simple hipped roof can present enough complications to negate the economics of vertical joint design.19 The difficulty of ensuring free movement of the roof should be appreciated when examining Fig. 6.18. It shows typical locations of roof attachment points on hipped roofs as used by a manufacturer. Vertical seam roofs can also have some appearance issues. As Stephenson19 points out, the often utilitarian appearance of its closing and edge details may not lend itself to aesthetically demanding applications where traditional batten roofing (or other architecture) is better suited. On vertical seam roofs it can be difficult to find good looking solutions for roof pitch changes or fascia and soffit transitions. In such situations, vertical seam structures or architectural roofs are more appropriate. Sliding panels can produce a certain "metallic" sound that some people find uncomfortable. Noise can be masked to a certain extent by roof insulation. In short, all of these limitations are outweighed by the benefits of the system. Permanent seam roofs remain a prime choice for metal building systems. The superior performance of vertical seam roofs is reflected in longer warranties than double seam roofs. In fact, the popularity of these roofs often draws potential users to metal building systems. Although vertical seam roofs are initially more expensive, they often prove to be very economical in a long-term comparison when the life cycle costs are taken into account.

6.5.3 Construction Details for Trapezoidal Structural Roofing The metal vertical seam structural roof shall be free to move with respect to purlins in order to function without leakage, buckling and structural failure. Conceptually the sewn field system

FIGURE 6.18 Fixing points for a hipped roof. Complex shaped roofs affect the ability to move the roof. (Central.)

METALLIC COVER METALLIC COVER

167

Sheet metal, hidden cleats, and spliced ends should work well as long as the roof size is compatible with cleat mobility and the purlins are well aligned. Misaligned purlin tops tend to cause roof binding and restrict movement. Challenges arise at edges, eaves, ridges, wall transitions, penetrations, and other areas where metal roofing can be constrained without thoughtful detail. The design details used by different manufacturers differ in their effectiveness. The Metal Building Manufacturers Association has compiled many general details that represent best practice in its Design Guide for Metal Roofing Systems.14 Some representative designs for various conditions from this and other sources are reviewed here. Another good publication is the NRCA20 Roofing and Waterproofing Handbook, particularly the metal roofing section. As previously mentioned, at some point the cover will need to be positively secured to the supports to keep it from slipping under its own weight. The point of attachment is usually the eaves and the end of the extension of the ridge. This approach allows the eaves support to be laterally supported by the roof (at least to some extent) and helps withstand high wind uplift forces, which are typically greatest in eaves. According to some estimates, nearly three quarters of all wind damage to roofs occurs at the eaves. Eliminating panel movement at the eaves is an added benefit for roofs in cold regions where the eaves are prone to ice dam formation and snow slide. Details of a fixed eaves and a floating ridge are shown in Figures 10 and 11. 6.19 and 6.20. Some manufacturers prefer the opposite approach - attaching the panels to the ridge and letting them float in the eaves. Very wide roofs can be fastened at neither the eaves nor the ridge but in the middle, which maximizes the flexibility of the roof clips. Regardless of which end of the metal decking is fastened, it is important that the panels can slide on the rake. Often the roof is attached to the bulkhead paneling (Fig. 6.21), creating a connection that limits the ability of the roof to float on the rake. The result is a failure of the fasteners or the roof. Finer details allow the cover to move relative to the end panel while the panel remains laterally supported, or allow the panel to move relative to the side of the end panel as shown in Figs. 4 and 5. 6.22 and 6.23. Occasional details on high walls and parapets require some thought to avoid introducing unintended roof attachment points at these locations. A heavy angle bracket with a simple L-shaped configuration that attaches directly to the roof can make it difficult to move. The author is aware of at least one leaking metal building where this happened. Also the W-shaped bezel shown in Fig 6.24 may not provide sufficient flexibility if it is attached directly to the panel rather than to the panel housing as shown. A combination of plates/nubs (Fig. 6.25a) or W-shaped plates with more pronounced curves (Fig. 6.25b) allows the cover to slide more easily. It is even more difficult to design a good wall-to-ceiling transition detail at the edges of the roof, such as B. the slope (Fig. 6.26) and the inside corner (Fig. 6.27). These closure details must effectively seal moisture from the roof and maintain its buoyancy as much as possible. Flashy details contribute more than their share of design and construction problems. Ideally the flashing should be of the same material as the panels and any caulk used underneath should be continuous. Leaks are usually due to poor sealant in the flash welds. How should changes be made? Hardy and Crosbie21 suggest welding the fin when there is no movement between the joined fin plates, but the finish after welding is not as durable as a baked shop floor. It is best to weld metals that do not require painting, such as B. stainless steel or galvanized steel. External gutters, visible or concealed, are common to all metal roofs, since a bi-directional slope to internal drains is obviously impractical in this type of construction. Details of the gutter attachment to the trapezoidal structural cover are shown in Figures 6 and 7. 6.19 and 6.28. Roof details on building expansion joints parallel to the roof pitch deserve special attention. (The step joint in the vertical direction is shown in Fig. 6.16 above.) There are two common design solutions for these expansion joints. In Fig. 6.29 the connection between the upstands is made by trapezoidal roof panels supported by continuous sloping angles. The latter are used instead of cover clips. A discreet flat extension bar can be slid over the toggle bar. Tilt brackets have grooved holes on the underside to avoid impeding plate movement. This relatively simple solution requires the panel layout to start and end at the junction. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

METAL COVER 168

CHAPTER SIX

FIGURE 6.19 Detail of a fixed eaves for a vertical butt roof. (MBK.)

A more complicated detail of Fig. 6.30 can be used if the connection is made at the cut edges of the panels. Expandable, curved flash helps accommodate cross-coverage movement. This design allows for greater flexibility in how the roof panels are arranged, but introduces a number of additional metal parts. Appropriate expansion joints in the eaves and gable as well as in the gutters are required for all details. And finally, Fig. 6.31 illustrates how the roof can be 'fixed' in a depression as shown in Fig. 6.18. Two cuneiform secondary structures ("cee"), supported by an angle below, close the Z-purlins on each side of the throat. The channels support the throat-bearing throat plate with a raised midrib. The center rib allows and encourages some transverse movement of the roof

METALLIC COVER METALLIC COVER

169

FIGURE 6.20 Detail of the floating ridge for a vertical butt roof. (MBK.)

the water drain. Obviously a lot of thought has gone into this detail, but in general hipped roofs work best with roofs that have vertical seams.

6.5.4 Construction Details for Vertical Seam Roofs Vertical vertical seam roofs have their place in commercial and institutional buildings where the utilitarian appearance of trapezoidal sheet metal roofs may be out of place. Some examples of roofing products with vertical seams are Butler Manufacturing's VSR* Roofing System and MBCI's Battenlock† (Fig. 6.32).

*VSR is a registered trademark of Butler Manufacturing Co. †Battenlock is a registered trademark of MBCI.

METAL COVER 170

CHAPTER SIX

FIGURE 6.21 Detail of vertical seam capping with fixed pitch trapezoidal profile, panel placed outside the module. (In this common detail, plate motion is prevented and not recommended by the author.) (MBMA.)

METALLIC COVER METALLIC COVER

171

FIGURE 6.22 Detail of roof rake with vertical seam. (MBK.)

FIGURE 6.23 Sliding connection of the trapezoidal roof structure to the gable end on the slope. (Star building systems.)

Vertical seam panels are often chosen for architecturally demanding applications such as roof-to-parapet transitions (Fig. 6.33) and roofs with ridges and valleys (Fig. 6.34). Indeed, vertical seam products were originally developed for architectural roofing but have since become popular in structural roofing applications. Unfortunately, some manufacturers continue to refer to these products as architectural roofing, causing confusion. In our definition, these panels represent load-bearing roofs with vertical seams, since they can cross the distance between the purlins unaided, as shown in the load table in Fig. 6.32. The best aesthetics cannot compensate for the fact that vertical joints cannot accommodate thermal expansion and contraction perpendicular to the slope as well as trapezoidal joints. Thin vertical seams leave no room for hidden clips with moving inserts, and movement must take place under the panels. One of the clips used on vertical seam roofs is in the top right corner of Figure 6.12. The seams can be rolled or snapped together with a portable machine.

METAL COVER 172

CHAPTER SIX

FIGURE 6.24 Section through parapet at high side wall. (A&S Building Systems.)

Vertical seam panels may look better than those with a trapezoidal profile, but are generally less strong due to the large flat areas between seams. As a result, the transition line from the roof to the real panel is not always as sharp as expected (Fig. 6.35). The manufacturers warn that panels of this type can lead to oil canning (slight wave formation). Aside from these fundamental differences, the perimeter details of the vertical and trapezoidal covers are similar. Both types require fixation (via fixation) at a specific point along the

METALLIC COVER METALLIC COVER

FIGURE 6.25 These roof and flashing details on a top wall allow for different movements. (Central.)

173

METAL COVER 174

CHAPTER SIX

FIGURE 6.26 Wall-to-ceiling transition at slope. (VP building.)

Tilt. A detail of the unventilated fixed ridge is in Fig. 6.36 and the ventilated floating ridge in Fig. 6.37. As can be seen, the difference between the two details lies not only in the degree of attachment of the roof, but also in the ventilation or lack thereof. Fixed and floating ridges can be supplied with or without ventilation. The floating edge detail in Fig. 6.37 shows a vented version where hot air collected at the apex is allowed to escape through a slotted soffit. As in chap. 8, roof ventilation is important to ensure the comfort of metal building occupants. A detail of the metal vertical seam cover attached to the eaves is shown in Fig. 6.38. To illustrate another point, a detail for the high eaves of a single pitched roof is drawn next to an adjacent high ("head") wall. Since the roof is assumed not to move on the eaves, a fixed L-shaped flash is used here. The ridge is not attached to the vertical seams, which are too far apart and too narrow for this, but to the fitted closure pieces clamped between the seams. As in chap. 5, the eaves can rarely be a full limit for roof assembly. Therefore we prefer that some roof movement, even on 'fixed' eaves, be accommodated by a W-shaped or curved flashing. A good example of a bent sheet is shown in Fig. 6.39, which illustrates how a high floating eave with a vertical seam roof can be constructed. With "hard" end walls (masonry or concrete), slightly different details are used. Depending on whether the walls are load-bearing or not, the roofing and roofing may need to accommodate not only horizontal but also vertical movement. Two representative details are shown in Fig. 6.40.

METALLIC COVER METALLIC COVER

175

FIGURE 6.27 Wall-to-ceiling transition at inner corner. (VP building.)

Figure 6.40a shows a condition where the vertical seam ends very close to the end wall, while in Figure 6.40a. 6.40b the cover is cut into modules.

6.6 INSULATED CONSTRUCTION PANELS As mentioned in our discussion of fixed metal roofs and in Ch. 8 Insulation issues have not traditionally been a priority for suppliers of metal building systems. However, the situation is changing quickly. Traditional methods of insulating metal buildings may no longer meet the energy saving regulations imposed by the latest code regulations. In what used to be the "hourglass" method, fiberglass insulation is laid over the purlins and squeezed through the roof on the purlin supports. This "short circuit" significantly reduces the overall thermal performance of the roof. Newer and better methods of insulating metal buildings include fiberglass thermal block insulation systems (Chapter 8), rigid insulation, and insulated structural panels (sandwich panels). The advantages of insulated panels are predictable insulation performance, finished floor area and, in some designs, the ability to provide purlin flange bracing. These panels have been used in cold storage for years and are now becoming popular in many other applications as well.

METAL COVER 176

CHAPTER SIX

FIGURE 6.28 Cover detail of vertical spans with trapezoidal profile in low eaves with fixed gutter. (MBMA.)

METALLIC COVER METALLIC COVER

FIGURE 6.29 Covering details of trapezoidal profile vertical joints in the expansion joint of the building. (MBMA.)

177

METAL COATING 178

CHAPTER SIX

FIGURE 6.30 Details of the expansion joint in a building with a trapezoidal roof. (Nucor Building Systems.)

Insulated structural panels typically extend between purlins and are attached to them with concealed clips. To reduce air and moisture leakage through side seams from panel to panel, the best products have intricate tongue and groove edges (Fig. 6.41). In addition to illustrating an element-to-element connection, Fig. 6.41 shows an inclined wall-to-roof connection with pre-insulated wall elements. With the exposed panel edges shown here, properly installed edging and sealant are critical to watertightness. As with other types of roofing, the transition to a high wall at the eaves must allow for movement of the panel. The aforementioned W-shaped or curved bezels can help; One of these details is shown in Fig. 6.42.

6.7 METALLIC ARCHITECTURAL COVERING In our definition, the architectural covering is that covering that requires a sub-grade for support. The substrate is typically plywood or metal decking, but other products such as oriented fiberboard, wood planks, and cementitious fiberboard are occasionally used. With properly designed accessories, the substrate can laterally support purlins and act as a fascia. The architectural cover is usually attached to the substructure with hidden clips instead of being it

(Video) Download PDF Design of Steel structure, Railway Engineering, PERT- CPM

METALLIC COVER METALLIC COVER

FIGURE 6.31 Details of the roof made of vertical beams with a trapezoidal profile in a fixed valley. (MBMA.)

179

METAL COVER 180

CHAPTER SIX

FIGURE 6.32 Example data for vertical seam panels (MBCI Battenlok). (MBK.)

METALLIC COVER METALLIC COVER

FIGURE 6.33 Detail of roof transition with batten locomotive. (MBK.)

181

METAL COVER 182

CHAPTER SIX

Fig. 6.34 Detail of the ridge to hip transition with Butler's VSR panels. (Butler Manufacturing Co.)

completely stuck. Self-drilling screws are generally used for fastening to metal terraces, and screws or nails for fastening to wood. Building panels generally fall into the water pollution category. As such, they cannot hold standing water and require an underlay (waterproofing sheet) underneath. They are most commonly specified for roofs with relatively steep pitches. Its seams are often fitted with or without battens, and the height of the seam is less than in structural roofs. Sealant is not normally used on the side overlaps. Building shingles are shorter than building roof shingles and are typically specified for roofs less than 60 feet wide. Therefore the cover does not require the complex clips of Fig. 6.12. Instead, the plates simply slide back and forth over loops or fixed loops as shown in Figures 1 and 2. 6.43 and 6.44. Clip spacing ranges from 1 to 5 feet depending on resistance to wind lifting forces. The most common spacing is 2 to 3 feet, and perhaps closer in roof areas exposed to severe negative wind pressure. The panel ends can be connected with fixed or floating loops. In any case, at least one plate is attached directly to the substrate. With the fixed overlap (Fig. 6.44), the panel lying on the slope overlaps the lower panel with a sealant in the middle and both panels are fixed with a continuous locking strip. Fasteners are typically self-tapping screws, chosen for their compatibility with the substrate. The lock is covered with a pad for additional weather protection. As with vertical seam roofs, the fixed overlap is generally provided at a single point on each roof panel. With a floating support, only the bottom plate is attached to the substrate. The top plate is attached to an offset clip attached to the substrate (Fig. 6.45). This connection allows the panels

METALLIC COVER METALLIC COVER

183

FIGURE 6.35 The transition line from roof to face occasionally becomes less sharp.

they move slightly relative to each other, although the sealant placed between them tends to limit their movement. (If the movement becomes excessive, the caulk may fail by excessive stretching or delamination.) In any case, as mentioned earlier, the primary waterproofing membrane on architectural roofs is not the metal but the base. As such, it is important to accurately detail the underbody seams, penetrations and finishing details, as with any other type of waterproofing. Diaphragm material must be chosen carefully as some products tend to melt under metal roofs in the summer heat. Products available range from traditional 30 pound biofelt to a self-sealing rubberized asphalt membrane. One of the most common problems with architectural roofs is the failure of installers to extend the membrane beyond the edge of the eaves support. Here a parallel to wall paneling is instructive. As curtain wall designers well know, effective glare extends beyond the face of the wall. If the bolt stops on the face of the wall - or worse, inside it - water can get under the bolt and into the wall. A similar situation can occur with improperly processed backing. The underlay is best laid in the gutter so that the water carried by the track can drain off. Or at least the edge of the base coat should point down over the siding and eaves as in Fig 6.46.

6.8 PANEL SURFACES Nothing spoils the appearance of a metal building system more than the sight of a rusted corrugated iron roof. Fortunately, modern metal finishes not only look good, they also protect

METAL COVER 184

CHAPTER SIX

FIGURE 6.36 Coverage details of vertical ribs in non-ventilated fixed ridges. (MBMA.)

Roofs from moisture, your biggest enemy, and from pollution. Durable is a surface that does not peel, crack or discolour over a reasonable period of time. Good lightfastness is particularly important for roofs in sunny locations where ultraviolet radiation often destroys darker colors such as reds and blues.

6.8.1 Anti-corrosion coatings The most common anti-corrosion coatings for steel roofs are based on metallurgically bonded zinc, aluminum or a combination of both. The ASTM A 924 specification covers both zinc and aluminum applied by the hot dip process. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

METALLIC COVER METALLIC COVER

185

FIGURE 6.37 Details of the roof of the vertical ribs in the ventilated floating ridge. (MBMA.)

The zinc coating found on known galvanized steel is primarily due to a sacrificial chemical action of zinc, which slowly melts while protecting the underlying metal. Of course, the thicker the zinc layer, the greater the protection; a coating according to the new ASTM A 65322, designated G60 or G90, is suitable for most applications. The G90 coating contains 0.9 oz/ft2 of zinc - applied in total to both sides of the sheet - and measures approximately 0.001 inch. on each side. In addition to its victim protection, galvanization provides a barrier against the elements, although this effect is secondary. The barrier effect is supported by a white film formed by zinc oxidation products. Industrially galvanized steel has a familiar glossy finish, while a hot-dip galvanized finish has a rough and dull appearance. According to some estimates, hot-dip galvanized panels can lose approximately 1Ⲑ2 mils in coating thickness every 5 years. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

METAL COVER 186

CHAPTER SIX

FIGURE 6.38 Detail of vertical roof ribs on an unvented solid high eaves adjacent to a main wall. (MBMA.)

METALLIC COVER METALLIC COVER

FIGURE 6.39 Detail of vertical roof ribs on an unvented solid high eaves adjacent to a main wall. (MBMA.)

187

METAL COATING 188

CHAPTER SIX

FIGURE 6.40 Roof detail and flashing on masonry sidewalls must allow for movement. (Central.)

METALLIC COVER METALLIC COVER

189

FIGURE 6.41 Detail of the insulated panels on the rake. (Central.)

The aluminum coating, on the other hand, acts primarily as a physical barrier formed by a clear, chemically stable residue of aluminum oxide, a product of aluminum oxidation. ASTM A 46323 Type 2 specifies the minimum weight of aluminum headliner on both sides as 0.65 oz/ft2 (the T2-65 designation for panels). Aluminized steel has a uniform matte finish. Aluminium-zinc coatings combine the sacrificial effect of zinc and the barrier protection of aluminium. Two compositions are common - Galvalume and, to a lesser extent, Galfan. Galvalume* coating introduced by Bethlehem Steel Corp. In 1972, it consists of 55 percent aluminum, 43.5 percent zinc and 1.5 percent silicon and is described in ASTM A 792.24. This ratio is based on weight; In terms of volume, aluminum makes up about 80 percent of the total volume and therefore represents a barrier measure

*Galvalume is a registered trademark of BIEC International Inc.

METAL COVER 190

CHAPTER SIX

FIGURE 6.42 Transition from insulated roof panels to high wall. (Central.)

Protection, in addition to a sacrificial effect of zinc. Galvalume board is available in commercial, blocking and textural grades and each grade can have one of three coat weights - AZ50, AZ55 and AZ60. A popular AZ55 coating weighs 0.55 oz/ft2 and has a nominal thickness of 0.9-1.0 mil (0.0009-0.001 inch) 5% aluminum, specified in ASTM A 875. Galvalume-coated steel looks like a mix of galvanized and aluminized steels, while Galfan

METALLIC COVER METALLIC COVER

191

FIGURE 6.43 Fixed clips used for short architectural panels. (Butler Manufacturing Co.)

Finish is difficult to distinguish from pure electroplating. Zinc and aluminum are usually combined with steel in a continuous hot dip coating process. Aluminium-zinc coatings are said to provide excellent metal protection for decades. A recent survey of 82 Galvalume coated metal roofs in the eastern United States by Galvalume Sheet Producers of North America found the roofs to be in excellent condition.26 The organization estimates that these roofs will last 30 years without problems in most regions can last before major maintenance is required; estimates that Galvalume coatings should last two to four times longer than G90 coatings in marine, industrial and rural environments. Part of the explanation for the superior performance of Galvalume and other zinc-aluminum coatings is that these coatings are less reactive and therefore retain their protective barrier longer than galvanizing alone. become the industry standard. Galvalume coated covers are usually guaranteed for 20 years by the manufacturer. Galfan coated steel is particularly well suited for field bending and plate forming applications as it is virtually unaffected by the cracking or spalling common to hot dip galvanized bends. Plates of Galvalume were traditionally coated with lubricating oil before rolling to avoid damaging the coating.14 It was believed that runaway oil would largely evaporate when the plates were delivered to the job site. In fact, a lot of oil was left behind and installers had to lift and walk over slippery, difficult-to-handle tiles. This disadvantage was largely overcome with the introduction of clear coated Galvalume plates. Leading zinc aluminum sheet manufacturers have developed many proprietary formulations of Galvalume coated with acrylic or other clear resins. According to Fittro27, some of the brands marketed in North America and their trademark owners are Galvalume Plus (BIEC International Inc. and Dofasco Inc.), Acrylume (USX Corp.), Galvaplus (Galvak, S.A. de C.V.), Zincalume (Steelscape) and Zintro-Alum Plus (Industrias Monterrey S.A.). The clear coating not only prevents oil leakage, but also minimizes smearing and scratching during plate storage and installation. Clear coated Galvalume is quickly gaining ground over the original oil lubricated variant. For unpainted roofing applications, the clear resin coating is designed to dissolve naturally within 12 to 18 months without spattering or flaking. cut edges. Despite the fact that zinc can extend its healing properties to raw edges to some extent, it's still best to have all covers cut and finished at the factory. Using the factory supplied repair compound will improve the corrosion resistance of the roof in areas of nicks and handling damage.

METAL COVER 192

CHAPTER SIX

FIGURE 6.44 Details of the architectural cover with vertical ribs on solid ground in the final support of the solid slab. (MBMA.)

It should also be noted that roofs with aluminum and zinc coatings must not come in direct contact with exposed steel to avoid galvanic effects. Contact of this coating with chemically treated wooden decks can also be harmful; The two materials must be separated by a properly installed primer layer.

6.8.2 Painted Coatings Clear Galvalume panels are perfectly acceptable for industrial buildings, warehouses and similar utility buildings with structural roofing. Some architects specify these silver roofs

METALLIC COVER METALLIC COVER

193

FIGURE 6.45 Details of architectural cover with vertical ribs on solid ground in final overlay of floating slab. (MBMA.)

many other applications - also for demanding work in the living area. But for architectural roofing and even structural panels where an additional layer of protection is desired, a durable finish is typically used. Finishes are sprayed onto the metal and baked at the factory. Metal roofing paints are predominantly based on organic (carbon) compounds such as polyester, acrylic and fluorocarbon. Acrylic and polyester-based paints that are common in residential homes are specified in the American Architectural Manufacturers Association (AAMA) Standard 603. These synthetic polymers have a durable, abrasion-resistant finish. A slightly different and better product is silicone polyester.

METAL COVER 194

CHAPTER SIX

FIGURE 6.46 Architectural canopy details of vertical ribs on solid ground in a fixed low eaves with hanging box gutter. (MBMA.)

METALLIC COVER METALLIC COVER

195

which offers good resistance to chalking (gradual film removal) and gloss retention. Acrylic and polyester based inks typically offer 3 to 5 year warranties, while siliconized polyesters can have 10 year or even 20 year prorated warranties. Fluorocarbon based paints are made from polyvinylidene difluoride (PVDF) resin which was introduced around 1970. PVDF resin is an exceptionally stable compound that offers excellent durability, color stability, and resistance to UV, heat, and peeling. The finished surface is dense, smooth and stain-resistant. According to the manufacturer, PVDF does not absorb ultraviolet rays, which are responsible for the frequent deterioration of the color of metal roofs. In one study, it was the only type of paint that was virtually unaffected by the Florida sun and salty air after 12 years of exposure. PVDF-based coatings can easily exceed their 20-year performance guarantee. PVDF resin is supplied by Elf Atochem North America, Inc. under the Kynar 500 trademark and by Ausimont USA, Inc. under the Hylar 5000 trademark. These two companies license their products to certain US paint companies; The complexity of the production process narrows the list down to a few. To reflect the dual origin of this finish, it is often referred to as Kynar 500/Hylar 5000 fluoropolymer resin. The actual coating is typically 70% resin and 30% pigments and solvents by weight, which is referred to as full strength or a 50 to 50 split , where some of the PVDF resin is replaced with the acrylic. The 70 percent formulation has slightly better resistance to paint and chalk fading, but the 50 percent resin offers better scratch and abrasion resistance. Therefore, the former has long been used for curtain walls in high-rise buildings, while the latter may be better suited for low-rise metal building systems that are subject to physical stress. AAMA Standard 605.2 specifies criteria for PVDF-based inks such as: B. Acceptable limits for gloss retention, color shift and coating erosion. In addition, it requires testing of some other properties such as salt spray resistance, heat resistance, moisture resistance, adhesion and chemical resistance. The PVDF surface can have different thicknesses. The standard liner is 1 mil (0.025 mm) thick, often stated as 0.9 mil. This coating may consist of a 0.7 mil top coat and a 0.2 mil primer. Perfect for mildly corrosive environments, it may fall short for moderately corrosive situations where a high quality 2 mil (0.05mm) finish may be required. (This premium finish may require special coil coater setup and will likely require longer hold time.) For exceptionally harsh or abrasive environments, a 0.1 mm (4 mil) thick specialty finish may be considered. However, the cost and difficulty of achieving this thickness, as well as the fact that field-cut plate edges, holes and scratches reduce the effectiveness of the "superfinish", make stainless steel or aluminum better suited for such applications. A relatively new and increasingly popular protective coating for metal roofing is two-layer PVC plastisol. The system includes a corrosion-resistant primer and a topcoat of polyvinyl chloride (PVC) resin dispersed in a plasticizer. When applied at a thickness of 4 mils (0.004 inch) or greater, plastisols offer excellent resistance to corrosion (including common acids, alkalis and inorganic compounds) and abrasion, even exceeding that of PVDF. Its color performance and gloss retention are generally inferior to PVDF inks. Should the back of the panel be painted? Obviously, additional finishing is not required for aesthetic reasons, but it can help resist abrasion during shipping and installation and improve resistance to corrosion from internal condensation. Some manufacturers suggest applying a full coat, while others are happy with a "backer" coat. Traditional shop-applied paint has recently been complemented by an on-site PVDF architectural coating introduced by Ausimont USA Inc. According to the manufacturer, Hylar 5000 ACS (Ambient Cured System) can be applied by spray or brush. Panel surface warranties are generally pro-rated. Upon completion of the work, the manufacturer lays down a chip of the material that can be used for comparison if a fade-out claim is made. 6.8.3 Stainless Steel, Copper and Aluminum For special applications requiring a cut over coated steel, or for purists who believe that every color will eventually fail, stainless steel, copper or aluminum may be the materials of choice.

METAL COVER 196

CHAPTER SIX

Stainless steel has excellent corrosion resistance, although it is not corrosion resistant as many believe. The popular 302 and 304 stainless steels contain approximately 18% chromium and 8% nickel, the two main components responsible for corrosion resistance. Type 316 stainless steel offers even higher corrosion protection at a premium price because it contains the additional ingredient molybdenum. To further increase corrosion resistance, stainless steel can be coated with a delicate alloy, already mentioned in Section 6.1 as a popular turn-of-the-century material. Although the first carbon steel and terne panels didn't last long, modern terne coated stainless steel is the most durable roofing material available. Terne coated stainless steel is clearly not intended for routine use due to its cost and is found in some of America's best known corporate headquarters including IBM, Procter & Gamble and Coca-Cola. Working with stainless steel roofing requires some knowledge on the part of designers and installers. To avoid galvanic action, stainless steel must be physically separated from all structural steel items such as purlins, fasteners and angles. This can be accomplished by installing a moisture barrier such as 30 pound roofing felt, using stainless steel fasteners, and following good industry practices when brazing or brazing the material. See ref 29 for some useful information on finishing stainless steel. Copper covers, an ancient and respected material, are still used to reproduce the rich and beautiful look of years ago. In addition to the refurbishment of historical buildings, copper has its place in new construction of all kinds. Panels are formed on the spot, as in the days of Paul Revere. The disadvantages of copper, in addition to cost, include problems with water drainage, which pollutes the underlying materials. Copper in contact with aluminum, stainless steel, or galvanized or bare steel can produce a galvanic effect. Aluminum is the most common non-ferrous metal roofing material. It offers excellent corrosion resistance, so it can be suitable for buildings near the sea and in saline environments. Because aluminum is relatively soft, it's easy to bend and extrude, but also easy to damage and dent. For this reason, aluminum roofs are not recommended for hail areas. Another disadvantage of aluminum is its high coefficient of expansion; Aluminum roofs expand about twice as much as steel roofs. Again, the issue of joining dissimilar metals needs to be approached carefully. Aluminum covers must be separated not only from steel purlins, but also from other non-aluminum structures and piping. Copper pipes and water derived from them must not come into contact with the roof. Fastening elements must be made of stainless steel. Aluminum panels are generally anodized by immersion in an electrolyte tank. The plate length is limited by the available size of the electrolyte tank. Electric current flowing through the tank deposits a layer of aluminum oxide, forming a chemically resistant, hard and durable top layer. Panels can be left a natural color or pigment can be added during anodizing to create a choice of chemically bonded colors such as bronze and black. The anodized finish retains colors well but is difficult to repair once scratched; it is susceptible to damage from pollutants. The design of aluminum structures is covered by the Aluminum Association (AA) standards30 and specifications.31 For stress analysis, the properties of the structural cross-sections are calculated using the actual dimensions of the cross-section. The concept of “effective width” is used for deflection testing. Aluminum alloys used for panels normally conform to ASTM B 209.32. Panels must be at least 0.032 inch thick. (0.8 mm) thick and 0.04 in. on larger spans. (1 mm).

6.9 FIELD FORMED METAL PLATES Despite the already mentioned and obvious quality advantages of factory made metal plates, there are situations where roll forming is done on site. Panels formed on site are not constrained by shipping limitations and can extend from ridge to eave, eliminating problem-prone overlaps. In addition, transport costs are saved, although expensive labor costs are incurred.

METALLIC COVER METALLIC COVER

197

rather than. In short, panels formed on site are typically cheaper than those supplied by large fabrication fabricators. Workplace roll forming was introduced in the 1970's and has steadily expanded since then as the quality of portable roll forming equipment has improved. One of the leading manufacturers in the development and use of such devices is Knudson Manufacturing. The company has state-of-the-art roll formers with rubberized drive rollers that are said to handle coils of steel, aluminum and copper to form vertical seam roofs without damage. Knudson can fabricate continuous panels up to 150 feet long and can form some C, Z and hat channel sections on site. Prefabricated curved panels can be laminated on site by Berridge Manufacturing Co. from Houston and some others. Despite increased product quality, roll forming in the workplace must be approached with caution as many panels produced on site still do not perform as well as panels made in big box shops and their installers are not necessarily as experienced. In addition, there is still a possibility that the coil surface will be damaged during forming; For this reason, galvanized steel, aluminized steel, and anodized aluminum are not recommended for field shaping. Incidentally, the metal strip formed in cold weather needs to be preheated before forming, which is often overlooked on the job site.

6.10 WINDLIFT RATINGS OF METAL ROOFING Photographs of blown and damaged roofs often accompany media reports of hurricanes, tornadoes and tropical storms. Damage to metal roofing caused by strong winds that create high suction forces can manifest itself as panel buckling, fastener breakage or tearing, seam warping or splitting, and seam staple failure. To ensure their product specifications, fabricators seek a windlift designation from one of the major testing bodies: Underwriters Laboratories (UL), Factory Mutual (FM), or U.S. Corps of Engineers (Corps). ASTM (formerly American Society for Testing and Materials) is also active in the search for the perfect test method. Unfortunately, no tests at this time can accurately predict how roofs will behave during a “real world” disaster. A brief explanation of the procedures available will help put the claims of roofing salespeople into perspective. 6.10.1 UL 580 standard for wind lift tests The classic UL58033 test has been used since 1973. It includes a 10 foot by 10 foot roof sample built on a dyno to typical manufacturer specifications. The sample edges are sealed and secured to the perimeter with spaced fasteners (6 centered on panel edges). In addition to the perimeter supports, two interior purlin supports are provided 5 feet apart. Plate clips attach to each row of brackets. The sample is then alternately exposed to wind and suction pressure. After the sample has safely withstood a 100 mph wind for 1 hour and 20 minutes, it is given the Class 30 designation. To advance to the next classification level, Class 60, the same sample must withstand a pressure equivalent to 140 mi/ corresponds to. h Wind for an additional 80 min. The highest designation, Class 90, can be obtained by testing the same sample a third time for an additional 1 h and 20 min under pressure generated by a wind speed of 170 miles per hour. Panels rated UL 580 Class 90 generally perform well - until exposed to a real hurricane. Partial roof chipping and seam detachments have been reported to occur at wind speeds producing only about one-fifth the roof lift capacity that would be expected from a UL 580 Class 90 rating.34 How can this happen? Experts point out that the test was developed to evaluate the bond strength of engineered roofs and not mechanically fastened metal roofs. In addition, this "static" pass/fail test, performed at constant pressure, does not take into account real world wind gusts and changing pressure patterns, does not accurately represent the actual behavior of metal roofs, particularly spliced type. As already mentioned, hurricane failures on these roofs mostly occur at the edges. For all of these reasons, the

METAL COVER 198

CHAPTER SIX

UL 580 test results cannot be directly translated into allowable wind uplift pressures, although some roofing suppliers claim otherwise. A UL 580 90 Wind Uplift rating does not mean the roof can safely withstand a 90 lb/ft2 uplift, although the sample must withstand a combined upward load of 105 lb/ft2 for 5 minutes to qualify. In essence, test results can only be used for "indexing" - to compare the tested panel with other similar products tested in the same way and under rigorous conditions.

6.10.2 ASTM Testing Methods Once the serious limitations of UL 580 testing were understood, many architects and building designers began to look for alternative testing methods. One such method, called Modified ASTM E 330, has become quite popular. The original ASTM E 33036 test was developed for curtain walls, not roofs. Walls and roofs behave differently under wind loads and the correct procedure for a very rigid wall attachment is not readily transferrable to flexible metal roofs. Despite its widespread use, ASTM's "modified" test method is not approved for testing metal roofs. ASTM recognized the need to develop an adequate standard for metal roof testing and formed its subcommittee E06.21.04 to address the complexity of the problem. The subcommittee's idea, ASTM E 1592, Structural Performance of Sheet Metal Roofing and Siding Systems by Uniform Static Air Pressure Difference, covers sheet metal panels and their anchorages. Essentially retains the basic approach of the E 330 test method and modifies it slightly to allow for flexible coverage. The sample size is 5 panels wide (10 feet total) by 25 feet - much larger than in UL 580. Intermediate purlin supports can be placed at variable spacings and the coverage is continuous over multiple spans. Panel clips are installed on each row of brackets, including the panel ends. No other fasteners are provided at the ends and edges, allowing the panels to move freely under load. The new test specifies the load to be applied so that slowly developing failures, such as confirm the load-bearing capacity of roofs under uniform static loading conditions. The test runs to failure and therefore allows the determination and tabulation of the final load bearing capacity of the roof. The procedure corresponds to the testing methodology of the AISI manual. It took the subcommittee more than 5 years to break the initial deadlock38 and reach consensus. Once published, however, the document was quickly endorsed by a leading industry group, the Metal Roofing Systems Association (MRSA), which was formed in January 1994 and produced a technical bulletin explaining the standards for designers. MRSA recognizes the fact that transversely fastened roofs behave differently under wind loads than vertical seam roofs and, unlike vertical seam roofs, can in fact be rationally analyzed for height. Consequently, the Association does not believe that ASTM E 1592 testing is required for fully attached roofs40, a reasonable argument supported by the MBMA and MCA.

6.10.3 FM Global Standard 4471 FM Standard 4471 contains another well-known test for wind lift. As a general rule, all metal buildings insured by a member company of FM Global (formerly Factory Mutual Systems) must have their roofs conforming to FM Standard 4471 “Class 1 Panel Roof Approval Standard”. While FM 4471 specifies design and construction requirements for Class 1 metal and plastic roofing panels, a related standard, FM 4470, is used for flexible roofing coverings such as single-ply membranes and engineered roofs. In addition to resistance to wind uplift, FM 4471 assesses non-static criteria such as resistance to fire, walkability and leakage. The roof assembly size used for wind lift testing is 12 ⫻ 24 feet. The kit must include the connecting fasteners and clips used in service. After the roof edges are sealed and secured at the perimeter, the panels are subjected to increasing wind pressure at ground level until the assembly either fails or can withstand a specified pressure for 1 minute.

METALLIC COVER METALLIC COVER

199

This maximum stroke pressure forms the basis for panel ratings such as 1-60, 1-90, 1-120, 1-150 or 1-180. For example, a panel that can withstand a wind uplift pressure of 90 psf for 1 minute will receive an FM rating of 1-90 (number 1 means Class 1 ceiling panels). Using a safety factor of 2 gives a maximum allowable design load of 45 psf.41 Like UL 580, FM 4471 does not account for wind gusts and is a 'static' test.

6.10.4 US Corps of Engineers Test Method Concerned that the available test methods listed above do not accurately measure the wind uplift resistance of roofs with vertical seams, the Corps developed (and later largely abandoned) their own test method. The older editions of their Military Construction Specification Guide, Section 07416, prescribed a test procedure that more closely reflected the effects of actual field conditions. For this test, Standard Test Method for Structural Performance of SSMRS by Uniform Static Air Pressure Difference, edge details must match actual panel construction; The perimeter grip is turned off. The test must be conducted under the supervision of an independent professional engineer using Corps-approved procedures. Some consider the process developed by the Corps to be a hybrid of the "modified" ASTM E 330 and ASTM E 1592 methods and the new regulatory requirements. It draws the same criticisms because it is "static"—does not account for uneven roof pressure distribution—requires only a small number of load cycles, and is not designed to determine an allowable roof load capacity.35 Although it is one of the more realistic tests for Metal roofs, the Corps procedure has been emphasized in favor of ASTM E 1592. Roofing systems previously tested and approved using the Corps test procedure may still be acceptable.

6.10.5 Which test to specify and why Why test metal roofs rather than full scale testing of for example prefabricated buildings? The answer is that while the static behavior of primary building structures exposed to wind uplift can be reasonably predicted computationally, that of metal roofs with vertical seams cannot. Steel plates bend and deform under lifting loads to such an extent that analysis based on a plane section is as relevant to their actual behavior as beam theory is to arches. In addition, many panel failures occur because the lateral overlaps of the panels at the clip locations become loose and excessive bending stresses are introduced into the “hook” portion of the clips.35 Therefore, if a rational design is not available and testing procedures are not perfect, what should a specifier do? For critical applications, it may make sense to require the vertical seam roof to be tested according to ASTM E1592. It is also wise to carefully consider any alternative testing methods suggested by the vendor, as some manufacturers are already performing dynamic testing of their products, which is arguably superior to the static ASTM method. The manufacturer's background should also provide some reassurance. Meanwhile, the search for the perfect test continues. Among other studies, MBMA and AISI sponsor research being conducted at Mississippi State University, Starkville. The new 32-foot by 14-foot hyperbaric chamber in this facility can simulate “real world” wind gusts by electromagnetically delivering an uneven pressure distribution that can change almost instantaneously. We hope that a better understanding of how real wind acts on roofs will lead to more reliable testing methods for roofs with vertical seams.

6.11 SOME TIPS ON ROOF SELECTION AND CONSTRUCTION Metal roofs can provide a long, maintenance-free life if properly designed and installed. The designer's contribution is to select the appropriate roofing product and details for the building.

METAL COVER 200

CHAPTER SIX

We hope the information in this chapter will help you determine if architectural or structural roofs with vertical or continuous seams are appropriate.

6.11.1 Avoid Penetrations Don't compromise the benefits of metal roofs by cluttering them with a forest of pipes, ducts, vents, and roof-mounted devices. Any penetration or opening in the roof will result in panels cut on site, which limit temperature expansion, exposing the metal to corrosion and leading to future leaks. It is better to make breakthroughs in the walls; these are much easier to protect against leaks. It is also best to combine multiple potential roof openings into one by connecting ventilation ducts together.21 Where this is unavoidable, penetrations and openings should be carefully detailed to allow for panel movement and to resist water infiltration . One of the product selection criteria should be the clarity of the manufacturer's specifications for difficult conditions. For felt-primed boards, penetration into the felt must be detailed at least as carefully as into the metal. In addition, the underlay must be able to drain into the eaves, which is impossible if the eaves edge is blocking it. To drain properly, the felt should run over the cladding and into the gutter, not behind it.42 Where roof detail makes this impractical, at least the edge of the bottom sheet should be laid down over the wall cladding and protected with interrupted eaves, as in Fig. 6.46.

6.11.2 Choosing the right products When a vertical seam roof is required, carefully examine the seam details offered by different manufacturers. As Stephenson43 notes, "Some so-called weatherseal designs are more weatherproof than others, and some may not prove waterproof at all." The Pittsburgh-style seam (Fig. 6.2d) is clearly superior to the snap-on types. Roof constructions recommended for low pitches (1Ⲑ4:12) are likely to be more water resistant than most architectural products designed for a pitch of 3:12 or steeper. From a hurricane resistance standpoint, all sewing projects are a bit vulnerable, but snap-button types seem to fare the worst. For roofing we recommend the use of Galvalume sheet clear coated for industrial, warehousing and similar applications or PVDF finished for all others except for smaller and temporary structures where acrylic or polyester will suffice. Galvalume roofs must be accompanied by sheets of the same material (and color line if the roof is colored) or aluminum. Galvanized, copper or lead coatings common in other types of construction should not be used on metal construction systems. Galvanized burr does not offer the same high corrosion resistance as Galvalume, while copper and lead burr can cause galvanic corrosion when in contact with zinc-aluminum coated steel. We recommend the provision of gutters and downspouts in most large metal buildings. Manufacturers bill these must-have rainwater removal agents as an additional cost item, and some homeowners are tempted to save money by omitting them. Without gutters, as in Fig. 6.46, water can attach to the underside of the roof and back up into the building, particularly in the absence of properly installed gaskets and seals. Finally, thin metal tends to be damaged by hail, as many car owners have found. In areas where hail is common, stronger panels should be used - or maybe even non-metallic roofs.

6.11.3 Special Considerations for Roofs in Cold Regions Accumulation of snow and ice can severely test the strength of a metal roof. Slipping and sliding snow can overwhelm some areas of the roof and create local depressions in others where standing water can collect. This water can enter the building through the water joints.

METALLIC COVER METALLIC COVER

201

scale plates or even through supposedly waterproof covers with a few small gaps in its defenses. A classic source of leaks in metal and other roofs are ice dams. Ice dams begin to form as the building's increased heat melts snow from the roof in some areas. Water seeps under the snow until it hits a cold eaves where it freezes. Newly melted water collects behind this ice dam and eventually runs back into the unprotected seams or sealing gaps of the roof. The key to ice and ice dam formation is a combination of warm roofs and cold eaves, so it's important to avoid both. As snow cottage builders long ago recognized, overhanging eaves and cathedral ceilings are not meant to be combined. Roof insulation keeps the roof cooler and reduces snowmelt. In some areas of the United States it is common to heat eaves with special heating cables. The risk of leakage can be further reduced by using a 'watertight' rather than a 'watershed' panel design and by increasing the roof pitch (up to 2 to 12 according to Tobiasson and Buska3). If water-bearing roofs are used in snowy regions, they should at least be equipped with additional waterproofing layers in the eaves and valleys. As Hardy and Crosbie21 suggest, valleys should be of flat metal and flared at the underside of the roof to encourage snow movement away from these difficult to seal areas. Vertical roof seams that terminate in valleys tend to impede snow sliding out of the valley and can become bent or torn as a result. Likewise, ventilation chimneys should be located away from the eaves so they do not become unwanted snow shields that can be damaged by snow slides. But what about snow guards in general? Architectural coverage, even when used on steep slopes, does not work very well with snow guards. Snow and ice trapped by the guards can melt and exit through the gaps, causing ice jams. When powder containment is required, it's best to specify quality Pittsburgh-style stitching. In any case, it is unclear how effective snow guards are on metal roofs. Sometimes they can pick up too much snow and cause roof overload, or they can be blown away by a large snow buildup, causing roof damage and leaks. Building codes often require that roofs with snow caps be rated for the snow load of flat roofs rather than a potentially lower roof load. It's important to know that snow catchers are designed to hold roof snow in place until it melts, but they cannot prevent snow from sliding off the roof. Snow accumulated on a steep roof tends to pull the roof down. Although the snow load is vertical in nature, on pitched roofs it can be broken down into components acting both parallel and perpendicular to the roof, similar to the one shown in Fig. 5.14a in Chap. 5. The steeper the roof, the greater the component parallel to the roof, the so-called drag force. Roof fasteners must be able to withstand this tensile force and the panel manufacturer must demonstrate that the fasteners are suitable for the task.

6.11.4 The Importance of Proper Design In this chapter we have stressed the importance of proper roof design and selection, but even the best designed system will fail if installed incorrectly. For this reason, the guarantees and reputation of manufacturers and installers are important. A wealth of information on various metal roofing manufacturers can be found in the NRCA Commercial Low Slope Roofing Materials Guide. Shop drawings should clearly show the installation details for all trim, fasteners, and sealant. Even more important, however, is the qualification of the installer. Despite the similarity of design flaws, the majority of wind-related roof losses can be attributed to improper attachment of the roof to the structure.20 The common problems of avoiding bad apples when selecting contractors are discussed in Chap. 9, but it's also important to gain confidence in the contractor's craftsmanship during roof assembly. How to know if the roof will be installed according to best building practices? Here are some signs of poor workmanship (and system design) in roof installation: 1. All panel end seams are flush, which means that four corners of the panel overlap and need to be sealed in one place, which is difficult to accomplish. Quality builders often require installers to stagger roof seams. Preferably the changes should be made on the purlins and not between them.

METAL COVER 202

CHAPTER SIX

2. Vertical roof seams form a wavy line when viewed perpendicularly rather than a straight line. This can impede the movement of the roof and lead to panel binding, clip damage and eventual leakage. 3. Bonding can also occur when the tops of the purlins are not flush, causing the roof rabbet to appear wavy when viewed from the side. Proper alignment of the purlins is the responsibility of the steel erector, not the roofer, but the results affect the roof and can be partially corrected with shims. Misalignment also occurs when the installer places some brackets too low or too high in relation to the roof seams. 4. Workers carelessly walk around the roof, denting the flat areas and pinching the raised seams instead of following the manufacturer's instructions, which usually suggest stepping next to but not on the raised seams, and using special ironing boards. 5. Instead of sheet metal closures (e.g. Fig. 6.26), metal roof transitions and penetrations are sealed by caulking. Additionally, the penetrations lack flexible sleeves or flexible cover plates for roof movement around them. For large penetrations, roof curbs should be used as described in Chap. 10. 6. Roof fasteners are not placed square to the roof or are improperly tightened (see Figure 6.9), incorrect fasteners used, etc. Aside from the obvious structural reasons for their occurrence, e.g. B. the use of very thin metal without frequent stiffening ribs, more can be added. Some are directly related to the list above. For example, oil retention can be caused by residual stresses in metal coils, stresses introduced during lamination, unevenness of the substrate, misalignment of seams relative to brackets, damage during construction, and restricted panel movement.45

REFERENCES 1. Paul D. Nimtz, "Metal Roofing: Past, Present and Future", Metal Architecture, November 1992. 2. "Market Trends/2002", Metal Construction News, January 2002. 3. Wayne Tobiasson and James Buska, " Vertical Seam Metal Roofing Systems in Cold Regions,” CRREL Misc. Document 3233, Hanover, NH, 1993. 4. Wayne Tobiasson, "General Roofing Considerations", CRREL Misc. Document 3443, Hanover, NH, 1994. 5. Section 07410N, Metal Roof and Wall Panels, Unified Installations Guide Specifications, Huntsville, AL, September 1999. 6. Section 07416A, Vertical Splice Metal Roof System (SSSMR), Unified Installation Specifications Guide, Huntsville, AL, September 1999. 7. "Masterspec-Evaluations", Sec. 07411, American Institute of Architects, Washington, DC, 1990. 8. R. White and C. Salmon (eds.), Building Structural Design Handbook, Wiley, New York, 1987, pp. 1048-1050. 9. Y.L. Xu, “Fatigue Performance of Bolted Lightweight Steel Roofing Tiles,” ASCE Journal of Structural Engineering, vol. 121, No. 3, March 1995. 10. Brian A. Lynn and Theodore Stathopoulos, "Wind-Induced Fatigue in Low-Metal Buildings", ASCE Journal of Structural Engineering, vol. 111, No. 4, April 1985, pp. 826-839. 11. Y.L. Xu, “Wind-induced fatigue loading and damage in hip and gabled roof coatings,” ASCE Journal of Structural Engineering, vol. 122, No. 12, December 1996. 12. Metal Building Systems, 2nd ed., Building Systems Institute, Inc., Cleveland, OH, 1990. 13. Widespan Buildings Erection Information, Butler Manufacturing Co., Kansas City, MO. 1978. 14. Metal Design Manual for Roofing Systems, Metal Building Manufacturers Association, Cleveland, OH, 2000. 15. Product and Procedure Manual, HCI Steel Building Systems, Arlington, WA, 2001.

METALLIC COVER METALLIC COVER

203

16. Tom Hulsey, “Self-drilling cap head fasteners are an excellent choice for corrosion resistance”, Metal Architecture, July 1999, p. 61. 17. “Fasteners”, an article on the World of Steel Roofing website dedicated to Galvalume roofs, www.steelroofing.com. 18. Paul D. Nimtz, "Metal Roof Systems", Metal Architecture, January 1993. 19. Fred Stephenson, "Specifying Standing Seam Roofing Systems for Conventional Buildings", Roof Design (now Exteriors Magazine), March 1984, as reprinted in The Construction Specificationr, November 1986. 20. The NRCA Roofing and Waterproofing Manual, 5th ed., National Roofing Contractors Association, Rosemont, IL, 2001. 21. Steve Hardy and Michael J. Crosbie, "Metal Roofing Fundamentals", Architectural Record, January 1997, pp. 161–165. 22. ASTM A 653-00, Steel sheet, galvanized (galvanized) or zinc-iron alloy coated (galvannealed) by hot dip method, ASTM, West Conshohocken, PA, 2000. 23. ASTM A 463-00, Steel sheet, hot-dip galvanized aluminum, ASTM , West Conshohocken, PA, 2000. 24. ASTM A 792-99, Steel sheet, 55% aluminum-zinc alloy, coated by hot-dip galvanizing, hot-dip galvanizing, ASTM, West Conshohocken, PA, 1999. 25. Specification data for Section 07600, standing seam Metal roofs, Galvalume Sheet Producers of North America, Bethlehem, PA, 1995. 26. Angelo Borzillo, "Galvalume Sheet Steel for Versatile Long Lasting Building Panels", Metal Architecture, October 1996, p. 10. 27. Bob Fittro, “Industrial Coatings and Substrates Offer Designers Many Choices,” Metal Architecture, June 2001, p. 28. 28. Joan Urbaniak, “Earning and Understanding Zincalume Coated Steel”, Metal Construction News, Nov. 1996, p. 58. 29. Standard Practices for Stainless Steel Roofing, Flashing, Capping, North American Stainless Steel Industry, Washington, DC. 30. Aluminum Standards and Data, Aluminum Association, New York, 1990. 31. Aluminum Design Manual, Sec. 1, Specifications and Guidelines for Aluminum Structures, Aluminum Association, New York, 2000. 32. ASTM B 209-00, Aluminum and Aluminum Alloy Sheet and Plate, ASTM, West Conshohocken, PA, 2000. 33. UL 580, Standard for Tests for Uplift Resistance of Roof Assemblies, Underwriters Laboratories, Inc., Northbrook, IL, 1989. 34. Charles H. Gutberlet, Jr., “Designing against Wind,” The Military Engineer, September-October 1992. 35 Harold Simpson, “Wind Uplift Tests”, Metal Architecture, January 1993. 36. ASTM E 330, Test Method for Structural Performance of Exterior Windows, Curtain Walls, and Doors by Uniform Static Air Pressure Difference, American Society for Testing and Materials, Philadelphia, PA, 1990 37 “ ASTM's Metal Roofing Standard Intended to Improve Performance", Metal Architecture, March 1995. 38. John Gregerson, "Bracing for Wind Uplift", Building Design & Construction, September 1990. 39. "MRSA Recommends Wind Uplift Tests for Standing Seam Roofing," Metal Ar chitecture, March 1995. 40. “Technical Bulletin on Wind Uplift Testing…” Metal Architecture, May 1995. 41. Richard Coursey, “Keeping Wind from Raising the Roof”, Architecture, August 1988. 42. Richard Schroter, “Leak Prevention for Metal Roofs,” Architectural Specificationr, November/December 1995. 43. Fred Stephenson, “Design Considerations for Architectural Metal Roofs,” Commercial Roofing Systems Handbook, Edgell Communications, Inc., Cleveland, OH, 1989. 44. Commercial Low Pitch Roofing Materials Guide, NRCA, Rosemont, IL, 1993. 45. Jeffry J. Ceruti, "Metal Roofing Systems Do's and Don'ts", The Construction Specificationr, September 1997, p. 64

METAL COVER 204

CHAPTER SIX

REVIEW QUESTIONS 1 List three types of buildings for which the structural roof might be used and three for which the architectural roof would normally be required. 2 What metals is Galvalume made of? What Makes Galvalume Superior to Coating? 3 How do I know if the self-drilling screws for full attachment roofs are installed correctly? 4 What is the minimum recommended pitch for "impermeable" roofs? Does it matter if the roof is in snow land? 5 Name at least three types of roof connections. 6 What are the advantages and disadvantages of vertical seam coverage versus full closure coverage? 7 Why are expansion joints necessary as roofs with vertical seams can slide in relation to purlins? 8 Which roof shape is best suited for a hipped roof? 9 What wind lift tests are typically required for vertical seam roofs? Are they necessary for fixed roofs?

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 7

WALL MATERIALS

7.1 INTRODUCTION Gone are the days when metallic structures were uniformly coated with galvanized corrugated sheeting. If Rip van Winkle was taking a nap in the 1940s and waking up today, he might have trouble spotting these buildings. Now, exterior wall material options for metal building systems are as numerous as for conventional structures; It's hard to tell which structure hides behind an elegant contemporary facade. The discussion in this chapter focuses on common wall materials for metal building systems: metal siding, masonry walls, concrete, and some modern lightweight finishes. Some possible combinations of these materials are studied to meet different functional or aesthetic needs. Our study does not exhaust all available options; We can only allow a brief discussion, omitting such well-known materials as glass, wood and stone.

7.2 METAL PANELS The first mass-produced prefabricated buildings were clad with unpainted galvanized steel panels. Color was introduced in the late 1950s; Paint was applied by spraying and baking, as on refrigerators and car fenders. In contrast, modern metal panels are formed from factory coated coils and are available in many durable finishes. Wall panels of metal buildings are usually supported on cold-formed C or Z beams. Most panels are made of 24, 26 or 28 gauge galvanized steel with additional coatings detailed in Chap. 6. Metal roofing and wall cladding are similar in many respects and some products can be used for both applications. Wall panels tend to be shorter than their roofing brethren, so they don't expand as much with temperature changes. Therefore, the vertical seam panel construction popular on roofs is not required for walls. Metal Wall Panels can be assembled in store or on site and with visible or concealed fasteners.

7.2.1 Field Mounted Panels Field mounted panels consist of an outer wall cladding, a fiberglass insulation mat and in some cases cladding panels. Ceilings offer finished interiors and can easily accommodate (or be substituted for) acoustic finishes. They also offer side bracing for straps. A panel can be mounted using one of two methods. In the first method, insulation mats are attached to the edges of the eaves and left hanging and held in place with retaining strips; then the outer sheets are attached to the straps through the insulation (Fig. 7.1). Finally, coatings are applied if necessary.

205 Conveyed from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

WALL MATERIALS 206

CHAPTER SEVEN

FIGURE 7.1 Field assembled metal panels with fasteners exposed. (Butler Manufacturing Co.)

WALL MATERIALS WALL MATERIALS

207

Fig. 7.2 shows a cross section through the wall assembled in this way (without insulation). To finish off the top of the wall cavity, some manufacturers offer a continuous capping strip that covers the top edge of the cladding panels. The cross-section of a typical cladding panel is shown in Fig. 7.3. The second method uses cladding panels with special edge ribs. The ceiling panels are first attached to the wall studs; Then the hat-shaped bottom chords are attached to the frames of the panel, the insulation is applied and finally the outer sheets are attached to the bottom beams (Fig. 7.4). The main advantages of site assembled panels are quick installation, low cost and easy replacement of damaged parts.1 The panels are lightweight and do not typically require cranes for assembly. Window openings are easily fabricated on site and easily trimmed using wall studs or frame rails (Fig 7.5). A fire resistant wall mount can be made by installing fire resistant drywall.

FIGURE 7.2 Wall cladding and cladding panels provide a finished appearance - and lateral reinforcement for beams. (Star building systems.)

WALL MATERIALS 208

CHAPTER SEVEN

FIGURE 7.3 Typical dimensions of cladding panels. (Star building systems.)

FIGURE 7.4 Field Mounted Isolated Panel. (Carlisle Engineered Metals.)

between the lining and the front panels. The arrangement may include several layers of plasterboard attached to hat channels running between the struts. Disadvantages include loss of thermal performance at the panel attachment points where the insulation is compressed. Some manufacturers solve this problem by using thermal blocks, similar to those found on roofs, or by inserting special tape between the metal parts that are in contact. It is of course possible to fit a layer of rigid insulation between the belt and the liner and omit the glass fibre, but this solution is unusual. Where a high level of wall insulation is required it is best to use factory insulated panels detailed in the following section.

7.2.2 Factory Fitted Panels With factory fitted panels the same three system components - outer cladding, insulation and inner cladding - are supplied as a unit. In addition to better customization, workshop assembly saves on-site labor. Rigid insulation, primarily urethane or polystyrene foam, offers better R-values than fiberglass. it depends on

WALL MATERIALS WALL MATERIALS

209

In terms of insulation, panel thickness can range from 21Ⲑ2 to 6 inches; The width can vary between 24 and 42 inches. Custom panels up to 12 inches. Insulation can be manufactured for specific applications such as cold storage and refrigeration equipment.2 The R-value of a popular 2-in. Thickness is usually between 15 and 16.3. Factory insulated panels are filled with foam insulation and have interlocking gaskets to reduce thermal bridging. A typical panel is fastened at the top and bottom to brackets and intermediate braces with concealed fasteners or with expandable fasteners installed from the inside (Fig 7.6). There are two ways to get rigid insulation in the panels. First, panels can be expanded by injecting liquified polyisocyanurate or urethane insulation between two continuously fed sheets of metal. The insulation is fused to the metal. In the second FIGURE 7.5 Detail on the reveal. (Carlisle Engineered's process called lamination, the two sheets of metal are bonded together under high metal pressure.) Pressure to form an expanded solid core of polyurethane, polyisocyanurate, or polystyrene.2,3 Composite panels consist of outer sheets of steel or aluminum laminated over rigid insulation or corrugated cardboard. Bonded by high performance epoxy adhesives, the assembly is lightweight, strong, durable and energy efficient. Composite panels are often manufactured with PVDF coatings, which are discussed in Chap. 6. For high-end applications, they can also be given a special finish similar to granite or marble.2 However, as Hartsock and Fleeman4 have shown, the recent increase in the popularity of these slabs has not been matched by increased technical knowledge. Foam insulation plays a vital role in the composite construction of foam-filled sandwich panels, allowing them to span greater distances than would be possible with a cover sheet alone. However, the composite design cannot be dictated by the limits of deflection and bending stress under load, but rather by thermal deformation and skin buckling. In fact, thermal deformation due to unequal temperature expansion or contraction of the two surfaces is one of the main causes of composite panel failure. Thermal deformation or bending plays a lesser role in single-support composite panels than in multi-span panels. Likewise, choosing lighter colors can reduce surface temperature and consequently warping. Another potentially critical issue is panel support, as the ability to design fasteners can control panel anchorage. Insulated and composite wall panels are usually attached to secondary structures with hidden roof-type clips (see Fig. 7.6 and Fig. 6.41 in Chap. 6). Like most wall coverings, store mounted panels usually extend vertically, but horizontal products are also available. The required load-bearing capacity of the bracket can be determined by multiplying the feeder area of the bracket by the projected wind load specified by the applicable building code for wall members. A clip supporting a 42 inch wide panel spanning between wall studs spaced 7 feet apart on center must withstand the wind load acting on a 24.5 ft2 secondary area. For example, with a wind suction load of 50 psf, this equates to 1225 lbs per clip, which can exceed the extraction capacity of traditional self-drilling screws attached to thin wall frames.3 To increase the capacity of the fastener, the metal the thickness of the secondary structure may need be increased beyond what is required for strength alone. Therefore, the metal building manufacturer must supply wall joists, eaves supports, hat channels and other secondary structures made of a material thick enough to develop the panel attachments, and not just to withstand a consistent wind load. In theory, the designer could determine the desired clamp connection capacity by the feeder area method shown above, but the design of bonded panels and their attachment is best left to experienced contractors. Ideally, the panel supplier coordinates the connection requirements directly with the metal fabricator. With workshop assembled panels, field modifications are difficult to make and the locations of all wall openings must be determined before the panels are fabricated. Some plumbers try to "fix" misplaced openings by cutting the metal with scissors and clippers and cutting through the insulation with saws, but the results are rarely perfect. Likewise, a sloppy erection will result in a poor fit, negating the benefits of this system. Installation of factory fitted panels should only be performed by experienced installers.

WALL MATERIALS 210

CHAPTER SEVEN

FIGURE 7.6 Typical installation of factory insulated panels. (Star building systems.)

WALL MATERIALS WALL MATERIALS

211

7.2.3 Exposed built-in panels These panels descend from the simple corrugated panels that came with early prefabricated buildings. A typical exposed mounting plate is attached to brackets with self-tapping screws or similar fasteners (Fig. 7.7). Side mounts can be spaced 24 inches or closer together. Today's possibilities go far beyond the basics and are very diverse (Fig. 6.4, 6.5 and 7.8). Wall panels with visible fasteners behave similarly to fixed covers; those in Figs. 6.4 and 6.5 can be used both as a top coat and as a coating. Zinc plated steel siding, or Galvalune siding, is manufactured in gauges from 29 to 18 gauge, with medium gauges being the most popular. Aluminum profiles are also available. Panel width can vary from 2 to 4 feet; Panel length is limited to approximately 40 feet due to shipping and handling restrictions. Exposed fastener cladding is still the most cost effective exterior wall material on the market; It is commonly used for buildings ranging from factories to schools. For added versatility, the panels can run horizontally when the architectural intent calls for it (Fig. 7.9). The horizontal alignment implies changes in the secondary framing, as discussed in Chap. 5. Deep ribbed slabs (3 to 4 in depth) may be particularly suitable for this purpose if they are able to bridge the gap between the portico columns (Fig. 7.10). Although the visible fastening system is easier to install and more forgiving of errors in the field than bonded panels, the experience of the installer is still important. A mistake as simple as over-tightening the fasteners can cause ripples or damage to the panels and allow water ingress. Panels fastened with fasteners are prone to fastener corrosion (Fig. 7.11). This problem can be solved by using corrosion resistant fasteners as described in chap. 6 - or by hiding them in the panel as described below.

7.2.4 Hidden Fastening Plates In this system, the fasteners connecting the plates to the brackets are hidden by interlocking edge joints (Fig. 7.12). In addition to being aesthetically pleasing, concealed mounting panels generally offer better protection against water ingress than exposed mounting panels. Physical protection is also enhanced as these panels are difficult to remove. Panels are typically 1 to 11⁄2 deep and made of 18 to 24 gauge steel; Lengths of 30 feet and more can be purchased. Concealed mounting structures are often found in field assembled standalone panels discussed in Section 7.2.1. The top and top plies may be connected via hat-shaped bottom chords (Fig. 7.4) or directly to one another if both the top and top plies have identically spaced hanging legs. Factory insulated panels also often use concealed fasteners (Fig. 7.6). The one in Fig. 7.6 has a flat surface, but in many cases, hidden mounting plates (and their exposed counterparts) have some sort of opening, grooves, or a rough texture to prevent oil buildup. As in chap. 6, canned oil is a small surface waviness, tending to affect smooth and light metal plates.

7.2.5 The rain protection principle Wall panels made of metal drain water mainly through the joints. The only watertight protection that joints normally provide is a bead of caulk applied between the edges of adjacent panels. When—not when—the sealant fails, the joint begins to leak. The situation is exacerbated by a common installation technique in which the sealant is applied between two panels of panels and the fasteners are driven across the interface, compressing the sealant in the process. The flattened seal does not last long and failure is desirable. A radically different method of preventing water ingress is based on the cavity wall principle, which dispenses with the idea of a single water barrier. The rainproof principle was formulated in 1963 by Kirby Garden, a Canadian researcher who recognized the impossibility of sealing every small opening that might appear in an exterior wall. Instead, the seals should be placed on an interior wall where protection from the elements and sunlight is easier to achieve.

WALL MATERIALS 212

CHAPTER SEVEN

FIGURE 7.7 Exposed mounting plates. (Star building systems.)

WALL MATERIALS WALL MATERIALS

FIGURE 7.8 Exposed Mounting Plate. (Panel “A” from MBCI.)

213

WALL MATERIALS 214

CHAPTER SEVEN

FIGURE 7.9 Metal plates extending horizontally. (Photo: Maguire Group Inc.)

A rainscreen consists of three parts: an outer layer, which most resists water penetration, an inner seal, and the air space between them. The inner waterproofing membrane is fully waterproof, with waterproof joints and flashing. It is partially protected from the elements by the outer film ("veneer"), which allows most water to bead off. The air gap physically separates the two layers and facilitates the drainage of accumulated condensate. Exterior rainscreen wall panels have no exposed sealants. Instead, some specific design steps are taken to block the various pathways of water intrusion into the cavity, which are described below for a horizontal plate connection:5 1. Near-horizontal wind-driven kinetic energy of rain, the most common form of water intrusion. The best defense is to build a ledge on the bottom plate and make an internal baffle. 2. Surface Tension: Water adheres and flows along the underside of the top plate. Fixed by a drip edge. 3. Gravity: The water simply follows the outer surface of the plate down. It can be overcome by tilting the joint surfaces upwards. 4. Capillary action: Water penetrates a thin joint like a wick. Disappears when the joint is at least 1Ⲑ2 inches long. Broad. 5. Air pressure drop and drafts: Water is drawn into the cavity by a pressure difference between the cavity and the outside. In a cavity wall, water ingress can only be resisted by the additional watertight air barrier. Walls constructed according to the ventilated cavity wall principle require two completely separate layers, as illustrated by the facing of brick over steel studs. The pins are covered with an internal seal, complemented by ridges and drainage holes. Normal single-layer wall coverings are obviously out of the question, at most only joints can be formed according to this principle. The design steps listed above greatly reduce, but do not completely eliminate, water intrusion into the cavity, particularly due to an air pressure differential. The Inner Waterproof Layer Connected and Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

WALL MATERIALS WALL MATERIALS

215

FIGURE 7.10 Exposed Deep Rib Fastening Side. (Super Rib by Centria.)

Adequate cavity ventilation is therefore critical to success and must be carefully planned and engineered. Another type of rain protection is the pressure equalization wall. The principle of pressure equalization assumes that when an air pressure difference between the cavity and the outside is eliminated, a water leakage problem caused by it, the Achilles tendon of hollow-wall umbrellas, can be solved. In order to achieve pressure equalization, enough openings are left in the outer skin to essentially make the cavity part of the outside world. The cavity itself is divided into relatively small chambers that restrict air movement within it and allow for rapid changes in air pressure. Some panel manufacturers now take the principle of pressure equalization into account in the joint design of their composite panels. An example of a factory insulated wall panel with joints suitable for cavity wall and rain screen design is the Formawall* shown in Fig. 7.13. The “void” in Fig. 7.13 is confined to the joint area and ends with a hidden seal between the panels. Note that the seal contains a pressure equalization hole.

*Formawall is a registered trademark of Centria.

WALL MATERIALS 216

CHAPTER SEVEN

FIGURE 7.11 Rusty screws in wall penetrations.

The best way to determine the effectiveness of these and other joint construction methods in containing water and air leaks is through extensive testing, in which water is sprayed at a specified rate and an air pressure differential is maintained between the interior and exterior surfaces outside the wall . The promised benefits of pressure equalizing rainscreens have yet to be fully proven. It is not even entirely clear whether perfect pressure equalization is even possible. Meanwhile, when properly designed and constructed, a simple cavity wall rain shelter offers a reliable, simple, and inexpensive method of protecting against the elements and water leakage.

7.2.6 Pedestal details Metal plates can be supported on the pedestal in various ways, as shown in Fig. 7.14. A base tube (a) embedded in the foundation concrete is preferred by many panel installers, mainly because it requires no fixing to the concrete on your part. However, the pipes must be delivered in time for concreting, which increases site coordination requirements. In contrast, the two most popular designs - the base angle (e) and the base rail (c) - are anchored to the foundation after it has been placed and cured.

WALL MATERIALS WALL MATERIALS

217

FIGURE 7.12 Concealed Fixing Wall Panel: (a) Exterior Wall Panel (WP by Carlisle); (b) inner liner (#SL100 from Carlisle); (c) Detail of interlocking connection and concealed fastener. (Carlisle Engineered Metals.)

The base channel has two main advantages over the base angle. First, the fasteners can be placed farther from the edge of the foundation than is possible with the base angle. The further away from the edge, the less likely the concrete will spall, which sometimes happens when fasteners are placed too close to the edge. Second, the base channel simplifies the anchoring of the cladding panels (Fig. 7.15) compared to the detail in Fig. 7.2, which requires two base linkers. The base beam construction (Fig. 7.14d) does not require concrete anchoring, but at the expense of an additional beam and its connections to the supports. Obviously the belted base construction does not provide a tight seal at the bottom of the wall. The belt is displaced by the hurricane's wind pressure or suction, allowing wind and rain to enter the building and damage its contents. Therefore, we do not recommend this detail. The basic details discussed above can be used when the wall paneling protrudes beyond the concrete edge. They all require a base cap or trim at the bottom (see Fig. 7.7 for a trim design used with a base tube). Closing or finishing the base is necessary for several reasons. It separates aluminum or zinc clad panels from uncoated steel and concrete, closing the siding and preventing moisture, insects and other creatures from entering the building. It also helps reduce air intrusion into the wall and maintain insulation at the bottom.

WALL MATERIALS 218

CHAPTER SEVEN

FIGURE 7.13 Element joints based on the curtain wall principle. (Formawall from Centria.)

Bottom closure strips can be made of metal, foam, or rubber as needed to perform their intended functions. End strips can be attached to outer panels before assembly or attached later. In a common design, the contour of the metal base trim matches the side profile (Fig. 7.16). Fastenings to the angle or channel are made only in the corrugations so as not to interfere with the flat part of the panel. This metal fastener works well as an insulating retainer and barrier against bugs, but it doesn't fit well enough to serve as an effective air barrier. A snug foam or rubber seal that fits snugly into the indentations will control air movement better, but will not hold the bottom edge of the insulation as effectively.

WALL MATERIALS WALL MATERIALS

219

FIGURE 7.14 Base frame details: (a) base tube; (b) base angle and blinking; (c) base channel; (d) girded base; (e) base angle. Note: A notch can be provided in the concrete to align the outer surfaces of the metal plate and concrete. (Star building systems.)

The base/flange angle (Fig. 7.14b) is useful when the outer edges of the casing and foundation are flush. This project requires a notch in the concrete, a minor complication for concrete workers. The base/bezel angle is preferred by some installers who simply stake the ends of the panels against the recessed angle, thus avoiding a separate base finish. A better looking, better performing design includes a separate color-matched trim at the bottom of the bezel, placed on the recessed base angle (Fig. 7.6).

WALL MATERIALS 220

CHAPTER SEVEN

FIGURE 7.15 The base channel simplifies the attachment of outer panels and ceiling. (Nucor Building Systems.)

FIGURE 7.16 The metal base matches the profile of the wall panel. (Butler Manufacturing Co.)

WALL MATERIALS WALL MATERIALS

221

7.3 HARD WALLS: GENERAL QUESTIONS 7.3.1 Why use hard walls? As the name suggests, hard walls are those made of masonry or concrete, not metal plates. Solid walls offer beauty, stability, security, sound, fire and side load resistance, in addition to a host of other benefits. Masonry and concrete exterior walls, long an integral part of conventional construction, are increasingly found in prefabricated buildings. Its increasing use can be attributed not only to property owners realizing the benefits, but also to greater involvement by architects in the specification of metal building systems. Additionally, some cities frown on issuing building permits for metal-clad prefab homes, insisting on traditional brick facades. There are two main types of hardwalls: single Wythe (one sheet) and cavity. The weather resistance of individual walls, also known as barrier walls, depends largely on the strength and tightness of a wall layer. These walls are typically made of concrete masonry units (CMU). The idea behind cavity walls is that a single line of defense against moisture is almost by definition unreliable and a second line of defense is needed. Contemporary cavity walls are typically made of masonry veneer combined with a load-bearing wall, either masonry or steel studs. Masonry walls are dealt with in the following sections and concrete walls are dealt with separately in Section 7.7.

7.3.2 The Challenges of Using Hard Walls in Metal Building Systems Some difficulties in incorporating masonry and concrete walls into metal building systems are quite obvious; others require careful study. The obvious challenges relate to the inherent mismatch of heavy, brittle hard wall materials and light, flexible metal structures. In addition, masonry and cast-in-place concrete are produced in a relatively slow field process - the exact opposite of the concept behind metal construction systems, which relies on rapid construction. Prefabricated masonry panels are available but not yet widely used. The less obvious difficulties lie in the structural interaction between the hard walls and the metal structure. Perhaps the most important is the structural role played by solid walls, such as load-bearing or enclosing walls. Among other things, hard walls require lateral support at two or more of their edges, but the means to provide this support may not be present in standard metal building systems. Likewise, standard details designed for all-metal buildings may not work with masonry or concrete exterior walls. For example, flange bracing of primary structure supports provided by beams in all-metal systems requires custom details in hard-walled buildings. These and some other challenges are discussed in the following sections.

7.3.3 Use of Fixed Walls as Retaining Walls External concrete and masonry may or may not be used as retaining walls, shear walls or mere enclosure. Design details for the different wall types depend on the wall materials in question. In general, some building manufacturers seem to prefer to use rigid walls as the containment (non-structural) walls, perhaps to preserve the design assumptions of the all-metal system in their design software. Some even keep the metal wall mount on their solid wall buildings. As in chap. 3, the lateral stiffness of solid walls far exceeds that of rod or cable supports, making support beyond the building construction phase unnecessary. Solid wall constructions are usually excluded from the metal construction manufacturer's scope of services. Therefore, whenever hard walls play a major role in the metallic building system, their design must be closely coordinated with that of the metallic structure. This can cause some difficulties when the solid walls are designed before the builder's decision. (A similar problem is faced by those who design foundations for metal buildings, as discussed in Chapter 12.)

WALL MATERIALS 222

CHAPTER SEVEN

Exemplary support details for load-bearing solid walls are shown in Fig. 7.17. Open I-joists can be supported on a mild steel wall angle (Fig. 7.17a) or placed in a wall pocket on a cemented backing plate. Cold-formed purlins can be supported by supports bolted to their webs (Fig. 7.17b) or supported on a steel wall angle, in which case the lower support purlins require anti-roll clamps or similar devices for lateral stability in the supports .

7.3.4 Use of hard walls as shear walls If a hard wall is used as shear walls, there must be sufficient force transmission from the roof disc to the wall. One way to achieve this is to connect the roof fascia rods to wall mounted steel angles. Brackets can consist of single snap angles with holes welded to bolted or flush mount panels (see illustrations in section 7.3.6). The bars may be welded to the supports or anchored to them with washers or similar devices. When a shearwall is non-structural, care must be taken that its attachment to the structure is not overly rigid in the vertical plane, but still allows shear transfer. For example, consider a shear wall next to a sidewall structure. The connection must be strong enough to carry the lateral load between the wall and the structure, but the wall must not be continuously connected to the structure by heavy angles or similar rigid elements that could limit the deflection of the structure under gravity loading.

FIGURE 7.17 Beam (a) or purlin (b) connection to structural precast wall. (Star building systems.)

WALL MATERIALS WALL MATERIALS

223

A common manufacturer detail for this condition is in Fig. 7.18. Using this detail, the connecting intermittent thrust plate should be as thin as possible - but still suitable for thrust transmission - so as not to constrain the frame too much in the vertical direction. However, for significant frame deflections, this detail may not be very effective due to the geometry involved. As in Fig. 7.19, when the frame is deflected downwards, the plate becomes a hypotenuse of the resulting triangle and not its side. Consequently, either the board must stretch (an unlikely scenario) or the wall must move inward to keep the length of the board constant. If the wall is inflexible, the panel or its attachments may break when trying to restrict movement of the frame. A more effective detail is angles with vertical slots that allow the frame to move without pulling on the wall (Fig. 7.20). If the solid wall is neither load-bearing nor a shear wall, its attachments to the structure need only be designed to transmit out-of-plane wind and seismic forces, as described in the following sections.

FIGURE 7.18 Attachment of non-structural shear panel to side panel frame. (Butler Manufacturing Co.)

WALL MATERIALS 224

CHAPTER SEVEN

FIGURE 7.19 As the structure sags downward under gravity, the length of the connecting plate must increase or the wall must move inward. If the wall cannot move, the panel or its connections may break.

FIGURE 7.20 Connection between bulkhead frame and shear wall. The use of an angle or bent sheet metal with vertical slots allows the frame to deflect without affecting the shear transmission to the wall.

WALL MATERIALS WALL MATERIALS

225

7.3.5 Clamp horizontally or vertically? In conventional construction, concrete and CMU walls are generally constructed to extend vertically from top to bottom. The lower support is provided by the foundation, the upper lateral support by the surrounding ceiling or floor beams. This approach does not lend itself to a typical prefab building where there are slight eaves rather than wraparound roof joists. As in chap. 5, the gutter shape of an eaves allows easy connection to both the metal cover and the siding. As originally intended, the eaves serve as a transition point between these two elements. While it works well with flexible metal roofs and sidewalls, standard beams generally lack the strength and rigidity to laterally support hard, brittle walls. It is easy to demonstrate that horizontal wall reactions to wind acting perpendicular to the wall create forces that require a much larger cross-section than that of a typical lightweight eaves. The size of the forces can be found in the design examples in Section 7.4. How stiff should the top brace be? Many engineers specify the maximum allowable horizontal deflection of steel members used to support masonry as the member length L divided by 600 (the L/600 criterion). Some use even more restrictive criteria. The same L/600 criterion can also be used for concrete. Alternatively, concrete and masonry walls can be constructed to extend horizontally between the building's metal columns, with wall ties or mortise angles transferring lateral reactions from the walls to the columns. Although relatively simple for two-way reinforced concrete walls, horizontally running CMUs require the primary structural reinforcement to be placed on closely spaced horizontal tie beams. The horizontal span arrangement works with regular eaves and seems to be preferred by many manufacturers. However, it has obvious issues with the doors and all wall control joints not matching the pillar positions. In addition, the bearing and shear wall capabilities of horizontal span CMU walls are not well known. Finally, the system becomes uneconomical with large column spacings and even with common eaves heights and spans (see example 7.3 in the next section).

7.3.6 Construction Details for Vertical Span Rigid Walls To ensure a high degree of rigidity and to meet the L/600 criteria discussed above, girders running horizontally between frame supports should be constructed of mild steel rather than cold formed metal. There are two basic options: using mild steel pipe at the top of the wall or a wide ring behind the wall. Both solutions have their advantages and disadvantages. In the first solution, the pipe is hidden in the thickness of the wall (Fig. 7.21). Since the wall is connected to the lower flange of the pipe, the pipe is subjected not only to bending but also torsion. Fortunately, tubular members have superior torsional strength. The main disadvantage of this solution is that the depth of the tube is limited by the wall and is economical only in areas with light and moderate winds. For example, a solid wall on the building with a 25 foot frame spacing and a 24 foot eaves height with a 20 psf wind load would likely require an HSS section 8 ⫻ 8 ⫻ 5Ⲑ8 to meet the L/600 criterion - quite a heavy steel link. The pipe can be attached to the top of the wall in a number of ways. Perhaps the simplest is to weld shear pins to the bottom flange of the pipe and embed them in the still-plastic concrete or CMU grout applied to the top tie beam (Fig. 7.22a). This method requires the solid wall to be erected and grouted first, with the exception of the concrete or CMU top layer which is grouted immediately prior to pipe installation. Obviously, the construction process requires close coordination between the trades. The opposite way - placing the tube in front of the wall - tends to make construction more difficult, especially with CMU walls. In order to place the vertical rebars and place the grout, workers would have to remove some side shells from the blocks and pump the grout through these openings. To make their job easier, they might be tempted to simply stop the vertical bars before the first blocks, at the expense of the strength and ductility of the wall.

WALL MATERIALS 226

CHAPTER SEVEN

FIGURE 7.21 A tubular member at the top provides lateral support for the outer wall of the full-length, full-height single-sheet CMU. (From a drawing by Butler Manufacturing Co.)

WALL MATERIALS WALL MATERIALS

227

FIGURE 7.22 Two details of attaching the pipe strap to the CMU wall: (a) by welded pins; (b) by post-installed screws.

Another connection detail requires less site coordination but a little more work. First, the fixed wall is erected, then the pipe with welded plates is placed and finally the plates are screwed to the cast connection beam on the side (Figure 7.22b). For both pipe attachment details, the size and spacing of fasteners and connectors are determined by analysis. If the wall ignores the primary framing, as in Fig. 7.21, the tubes may be attached to the outer flange of the column in such a way that they effectively resist the applied horizontal forces and torsion. This usually requires elbow or plate joints at the top and bottom flanges of the annulus. The second design solution is to place a wide flanged beam, extending from column to column, behind the wall, as in Fig. 7.23 for masonry and concrete walls. The main difference between using CMU and concrete in this detail is in the wall connections: it is easier to use slabs cast in concrete than in CMU. In this construction, the wall extends to the bottom of the metal roof. Note the welded bracket for the rod membrane connection. The main advantage of this solution is the material saving, since the wide flange belt can be manufactured in the required depth instead of being limited by the wall thickness. Difficulties include connecting the harness to the frame: it must be made on a thin core rather than a relatively thick flange to which the top wall tube can be attached. Another complication is that, unlike a section of pipe, the wide-flanged belt requires flange reinforcement (Fig. 7.23) to be effective, increasing labor costs on site.

7.3.7 Column flange stiffener in hard walls The cold-formed steel beams to which the column flange stiffener is usually attached (see Chapter 4) must not be present in hard walls. What happens to the flange bracing when using masonry or concrete walls? It is possible to dimension columns without flange bracing, but a more economical solution is to add flange bracing to the walls. The details of flange bracing for solid walls are slightly more complex than for metal clad walls, mainly because both column flanges now have to be braced. The outer column flange can be attached to the wall with retrofitted anchors. Anchors can be drilled directly through the flange (Fig. 7.24a) or through short gusset plates attached to the column (Fig. 7.24b). The cutting angle version is used when the column depth or

WALL MATERIALS 228

CHAPTER SEVEN

FIGURE 7.23 Wide-flanged beam behind a vertically extending concrete or masonry wall. Note the flange clamp and diaphragm mount.

The flange width is not sufficient for a direct screw connection or if a reduction of the flange area due to screw holes is to be avoided. Section angles should also be used when the solid wall has a column centerline connection; otherwise the screws would be placed too close to the joint edges. The inner flange of the column is reinforced in the usual way - by a pair of flange brackets bolted to the solid wall. Electrical tape, paint, or putty is needed to isolate the steel from direct contact with masonry or concrete. Alternatively, the outer flange of the column should be at least 1 inch. away from the wall (Fig. 7.24b). As the manufacturer similar to Fig. 7.24 normal state, fixed wall attachment is "foreign". Steel frames are usually erected well before masonry or concrete, and by the time the wall is up the steel erectors are likely to be long gone. However, it is extremely important to check that all required frame-to-wall connections have actually been made - by experienced personnel, preferably the steel fabricators contracted to do the work. If the flange reinforcement is forgotten or installed incorrectly, frame columns can be dangerously stressed under load.

WALL MATERIALS WALL MATERIALS

FIGURE 7.24 Column flange bracing details for rigid walls. (a) On concrete walls without joints in pillars; (b) on CMU walls with gaskets.

229

WALL MATERIALS 230

CHAPTER SEVEN

7.4 SINGLE STRUCTURE MASONRY 7.4.1 A Popular Choice CMU single frame walls are popular in prefabricated buildings for their fire and sound resistance, toughness and other masonry advantages. Visible CMU walls don't have to be gray and nondescript: a wide range of split, sanded and scribed block elements are available in attractive colours. Units are typically 8 to 12 inches thick. Like any solid wall, external masonry can be designed as a load-bearing, penetrating or just enclosing wall. Concrete wall blocks used in structural walls and in exterior applications where freezing is possible must conform to ASTM C90. Instructions. The steel reinforcement improves the strength and ductility of the wall. A reinforced wall gradually flexes under load and tends to form a multitude of very narrow cracks rather than the few wide cracks that can occur in an unreinforced masonry wall. Vertical reinforcement is placed in potted cells and spliced as needed to increase the overall height of the wall. Many masonry designers avoid using more than one layer of vertical reinforcement because bar placement techniques are much less precise in masonry than in concrete construction. However, on the west coast, due to the traditionally high level of inspection in these seismic areas, two layers of staff are often used. The horizontal reinforcement may comprise deformed bars or fabricated wire-bonding reinforcement. Bars are placed in special units, either U-shaped connecting beams or blocks with notched webs. Joint reinforcement is placed in horizontal joints between blocks. The height of the reinforcement is determined by calculation according to the applicable building code. Most building codes recognize ACI 5307 as a trusted source. There are many computer programs and masonry design tables, a useful resource is the National Concrete Masonry Association (NCMA) Concrete Masonry Design Tables. Factors such as the applicable building code, the building's seismic performance category, and whether or not the wall is a shear wall. For example, for seismic performance category D, which represents areas of high seismicity, ACI 530 requires CMU walls to be strengthened in two directions, with the sum of the vertical and horizontal cross-sectional areas of reinforcement being at least 0.002 times the gross area of the wall . The minimum area of the bars in each direction must not be less than 0.0007 times the gross wall area. The reinforcement must be evenly distributed; Maximum spacing in any direction is 48 inches, excluding masonry laid in pile bond. For CMUs placed in pile tie, the maximum rebar spacing is reduced to 24 inches and the units must be fully cemented. Additional provisions concern the shear wall reinforcement. For other lower seismic performance categories, more lenient requirements apply. There are a few points to consider when specifying CMU walls in metal buildings. Because masonry cannot be easily removed and reused, it is ill-suited to buildings where future expansion is likely. In addition, masonry is heavy and requires continuous foundations for support. The insulating value of CMU walls is low; To increase it, interior walls will be needed, lined with insulation and finished with drywall or metal siding panels. Blocks will absorb moisture unless treated with a water-repellent compound at the time of production or given a waterproof coating that needs to be renewed periodically. To reduce moisture penetration through mortar joints, the joints must be properly finished, preferably in a concave shape. Adding a water-repellent mixture to the mortar can also be beneficial. Like any exposed masonry, CMU requires closely spaced control joints, e.g. B. 20 to 25 feet. Fortunately, this joint spacing corresponds to common building well sizes. Despite all the precautions taken, the single-layer CMU still does not enjoy the benefits of cavity construction, and cracks in the wall or lack of mortar allow moisture to enter the interior (Fig. 7.25).

WALL MATERIALS WALL MATERIALS

231

FIGURE 7.25 The watertightness of this single-layer CMU wall is compromised by the lack of mortar. Daylight can be seen through the head gaskets.

7.4.2 Vertically extending CMU walls The main discussion of vertically extending solid walls occurs in Section 7.3. Here we examine some design issues specific to CMU walls, still the most common form of hard walls, and provide some design examples. Whenever the wide-brimmed perimeter of Fig. 7.23 behind the wall is used, its height must be carefully determined. Often the strap cannot be placed at the very top of the wall where the plastered connecting beam is usually located, as some manufacturers provide diagonal reinforcements on the knees of the frame. This makes any corset connection in the knee area impractical and requires placement of the corsets below the knee. The CMU wall now needs to be fully cemented at the level of the waling where the anchor bolts are located. To avoid placing the anchor bolts close to the mortar edges, at least two layers of blocks closest to the bolts must be cemented. This requirement must be communicated to the masons and the grout monitored accordingly to ensure that the anchors are not placed in empty block cells. A simple and foolproof but more expensive solution is to cover all exterior walls with solid grout. A solidly plastered wall is naturally stronger than a partially plastered wall.

7.4.3 Reinforcement of the walls of the CMU by intermediate beams? With a combination of wall span and loading, the strength of a standard 8-inch CMU wall extending vertically becomes insufficient, or the wall may require an excessive amount of steel reinforcement. For

WALL MATERIALS 232

CHAPTER SEVEN

To increase the wall's ability to bend, it is generally more economical to make the wall thicker than to support it with horizontal beams or wind columns. (Note that the design thickness of ribbed blocks does not include the ribs and ribbed blocks must be thicker than normal blocks for the same span and load). of stiffness mismatch, a consideration often absent in some secondary calculations that treat masonry the same as metal cladding. Masonry tends to crack at relatively small lateral deflections, while metal siding and struts can deflect severely without breaking. Rather than locking the masonry as it deforms under horizontal loading, flexible straps simply move without being fully taut. By the time the straps are sufficiently tensioned to exert the intended restraining forces on the masonry, the brittle wall may already have cracked. To be effective as a sidewall brace, braces must deflect less laterally under load than the CMU walls they are intended to reinforce. This usually requires lower and heavier waist sections than is required for strength alone. Conventional 8" deep cold-formed Z-beams generally do not provide adequate lateral support for 8" CMU walls, and deep steel structure sections are typically required. As previously mentioned, the maximum allowable horizontal deflection of steel members used for masonry bracing is typically assumed to be L/600 and perhaps even less. The design process is similar to Example 7.2 below. Wherever CMU walls are present, significant lateral rigidity is required not only from the beams but also from the primary building structures. Otherwise, heavy-duty belts become encased in a structure that moves excessively ("drifts") under lateral loading, rendering them ineffective. The question of lateral drift criteria for steel structures with masonry walls is dealt with in Chap. 11. Example 7.1 Design a single leaf CMU exterior wall for a rigid frame prefabricated building with a span, eaves height of 24 feet, and frame width of 80 feet. Wind load is 25 psf and roof payload is 20 psf. The wall extends vertically with the top brace placed behind the masonry and below the knee as in Fig. 7.26a. The wall bears no vertical load and its own weight can be neglected. Use ACI 5307 Seismic Performance Category D to determine the maximum vertical and horizontal wall reinforcement spacing and minimum reinforcement percentage. For this example, simply consider the wind load perpendicular to the wall; neglect seismic loading and any shear wall behavior. Assume specified masonry fm' with a compressive strength of 2000 psi and "partially plastered" masonry. Solution First determine the approximate position of the top beam to find the construction wall span (distance L in Fig. 7.26a). Consult the frame tables in chap. 4, find the distance from the base of the spine to the bottom of the knee as 21 feet. Since this number is approximate, conservatively position the waist 20 feet above the base. The wall can then be analyzed as a cantilever beam with L ⫽ 20 ft and ⫽ 4 ft under a uniform load w ⫽ 25 lb/ft (Fig. 7.26b). The horizontal reactions R1 and R2 and the maximum design bending moments M1 and M2 can be found by standard beam formulas: 25 R1 ⫽ ᎏᎏ (202 ⫺ 42) ⫽ 240 lb/ft (2) 20 25 R2 ⫽ ᎏᎏ (202 ⫹ 42) ⫽ 360 lb/ft (2) 20 25 M1 ⫽ ᎏᎏ2 (202 ⫹ 42) (202 ⫺ 42) ⫽ 1152 ft-lb/ft ⫽ 13,824 in-lb/ft (8) 20 (25)42 M2 ⫽ ᎏᎏ ⫽ 200 ft- lb/ft ⫽ 2400 in-lb/ft 2 The required instantaneous strength of the wall can be determined using any accepted masonry design method. Instead of manual calculations or computer programs, you can use concrete masonry

WALL MATERIALS WALL MATERIALS

233

FIGURE 7.26 CMU wall for example 7.1.

NCMA8 design diagrams for 2000 psi fm′ 8″ CMU with bars placed at mid-wall depth. Because the building falls within seismic performance category D, the maximum vertical and horizontal bar spacing is 48 inches. Since the wall is not load-bearing, the load combination only includes wind and permanent loads (the latter is neglected in this example). Most recent codes do not allow a one-third increase in allowable stresses (or a 25 percent reduction in total load) for this case, as has been allowed in the past. (If the applicable building code still allows such an increase in stresses or a decrease in loading, multiply the allowable moments by 1.33.) Locate the most economical vertical reinforcement in NCMA Table 3.2.13: #7 bars, 40 spacing ok (14.833 in-lb/ft tensile moment) Check the minimum rebar percentages. Vertical bars #7 spaced 40 in o.c. Enter a gain ratio (using the actual block size) of: 0.60 ᎏᎏ ⫽ 0.00196 ⬎ 0.0007 (40)(7.625)

Also 0.00196 ⬎ (2Ⲑ3)(0.002) ⫽ 0.0013 Since the vertical bars account for more than two-thirds of the total required, the horizontal area of reinforcement need only be: Ah ⫽ (0.0007)(48)(7.625) ⫽ 0.256 in2 This can be satisfied by two bars #4 (Ah ⫽ 0.39 in2) or one bar #5 (Ah ⫽ 0.31 in2). It is common to supply two bars in glued beam units so choose two No.4.

WALL MATERIALS 234

CHAPTER SEVEN

So select #7 vertical bars spaced 40 in o.c. and two No. 4 bars mounted on horizontal tie beams spaced 48 o.c. are arranged. Example 7.2 Design the horizontal steel beam used as side support for the top wall in Example 7.1. Use steel with a yield strength of 50 ksi. Use L/600 wind load deflection limit. Solution As the reaction R2 per foot of beam was found in Example 7.1. B. 360 lb/ft, the maximum design bending moment for a horizontally extending beam is (360)252 Mmax ⫽ ᎏᎏ ⫽ 28.125 lb-ft ⫽ 28.125 kip-ft 8 When the beam is fully stiffened, Fbx ⫽ 0.66 (50) ⫽ 33 ksi The required section modulus is Sx, rq ⫽ (28.125)12/33 ⫽ 10.23 in3 Using a deflection limit of L/600, the required moment of inertia is approximately (10.23)( 25) Ix, rq ⫽ ᎏᎏ ⫽ 219 in4 1.17 Use the AISC manual9 and try W14 ⫻ 26 (Sx ⫽ 35.3 in4, Ix ⫽ 245 in4). Determine the lateral bracing requirements for the beam. Since the size of the beam is determined by stiffness and not strength, the beam is lightly loaded and its lateral braces can be spaced apart by more than the distance Lc. Using the allowable moments tables in the manual for W14 ⫻ 26 with Mall ⫽ 28.125 kip feet, determine that the free length is about 12 feet, about half the span. Use the equations from Chapter F of the AISC10 specification to verify that midspan-only bracing is acceptable. For l ⫽ 12.5 feet, rT ⫽ 1.28, d/Af ⫽ 6.59, Cb ⫽ 1.0, and Fy ⫽ 50 ksi, calculate:

Energy

510.000Cb ᎏᎏ ⫽ 101 Fy (12,5)12 l ᎏrᎏ ⫽ ᎏ 1,2ᎏ 8 ⫽ 117,2 ⬎ 101 T

Then use equations F1-7 and F1-8. 170,000(1.0) Fb ⫽ ᎏ (117ᎏ 0.2)2 ⫽ 12.38 ksi ⬍ 0.6 Fy

← Use

(F1-7)

12.000(1,0) Fb ⫽ ᎏ (12,5)(12ᎏ )(6,59) ⫽ 12,14 ksi ⬍ 0,6 Fy

(F1-8)

Use Fb ⫽ 12.38 ksi. 28.125(12) fb ⫽ ᎏᎏ ⫽ 9.56 ⬍ 12.38 ksi 35.3

Check the lateral deflection of the beam at design wind load:

WALL MATERIALS WALL MATERIALS

5(0.360)244(1728) ⌬hor ⫽ ᎏᎏ ⫽ 0.445 in (384)(29,000)(245)

0,445 ᎏᎏ ⫽ L/674 ⬍ L/600 (25)(12)

235

Check the vertical deflection of the beam under its own weight for l ⫽ 12.5 ft and Iy ⫽ 8.91 in4: 0.0054(0.026)12.54(1728) ⌬vert ⫽ ᎏᎏᎏ ⫽ 0.023 in 29.000(8.91 )

(insignificant)

Therefore select the circumference W14 ⫻ 26 with flange struts at the span center distance.

7.4.4 CMU Walls that Extend Horizontally As discussed in Section 7.3.5, solid walls can be constructed to extend horizontally. It is worth considering some design nuances related to the CMU. The first is the attachment of the CMU wall to the portal columns at the locations of the connecting beams. A note on typical connection details (see inset in Fig. 7.27) states that masonry ties or anchors are excluded from the scope of the manufacturer. This means that part of the responsibility for your design lies with the architect/engineer. We say "partial" because these attachments must be matched to the design of the column flange stiffener (Fig. 7.24), which is normally obtained from the manufacturer. The second nuance concerns the need for a plinth recess in one-star CMU walls, as shown in Fig. 7.27. According to one school of thought, the hollow CMU acts like a cavity wall, and any moisture that seeps into the outer shell must be removed - hence the flash. However, bidirectionally reinforced CMU walls, as required by contemporary building codes, must not have a series of have full-height vertical cavities through which water can flow. Unless walls contain a lot of heavy horizontal connecting rebar, they can have fully cemented connecting beams at close intervals (perhaps 4 feet on center, measured vertically). For glued beams, the cavity wall analogy applies only to each wall segment in between. Logically, a collar should be provided over each tie bar - an unusual construction - or not at all. There may also be another independent reason for the base recess: breaking the connection between the wall and the foundation to allow rotation of the wall under horizontal loads, a problem discussed in Chap. 11. Example 7.3 Design an 8-inch building using the design load, masonry strength, and seismic performance category from Example 7.1. Steel bands or wind columns are not provided. Solution The maximum bending moment for the wall extending horizontally a distance of 25 feet and subjected to a wind load of 25 psf is (25)252 Mmax ⫽ ᎏᎏ ⫽ 1953.125 lb-ft ⫽ 23,438 lb-in 8

(per foot of wall height)

This moment is much larger than the moment calculated in Example 7.1 for a vertically extended wall, but the situation is aided by the fact that a double curtain of connecting beam elements provides greater effective beam depth. Using Table NCMA 3.2.15 and neglecting the one-third increase in allowable stresses, find the most economical horizontal composite beam reinforcement: Two #8 bars spaced 40 o.c. (23,544 in-lb/ft moment resistance) Note that bars larger than #8 may have difficulty fitting into an 8 inch cord and still provide the required grout wrap around them. (Some may suggest that even bar #8 might be too big for the task.) This gives a horizontal gain percentage of

WALL MATERIALS 236

CHAPTER SEVEN

FIGURE 7.27 Horizontal single-layer CMU wall. (Butler Manufacturing Co.)

WALL MATERIALS WALL MATERIALS

1,57 ᎏᎏ ⫽ 0,0051 ⬎ 0,0007 (40)(7,625)

237

Also 0.0051 ⬎ 0.00133 Since the horizontal bars are more than two thirds of the total, the vertical reinforcement must not be closer than 48 in o.c. be spaced. and your area just needs to be Av ⫽ 0.0007(48)(7.625) ⫽ 0.256 in2 Use #5 vertical bars in 48 in o.c. (Average ⫽ 0.31 in2). Determine the reaction at the anchors connecting the CMU to the steel column (anchors are spaced 40 inches apart at each tie beam): Rmax ⫽ 25(40/12)(25)/2 ⫽ 1042 lb This is a significant force , and the connections must be carefully selected. So select #5 vertical bars spaced 48 in o.c. and two #8 bars mounted on horizontal tie beams spaced 40 in o.c. are arranged. (Compare the total reinforcement in this example with that in example 7.1 for a vertical wall extension.)

7.5 BRICK WALLS 7.5.1 Clad brick over CMU This system combines the advantages of CMU walls, such as durability and fire resistance, with elements of a curtain cavity wall. Figure 7.28 shows a version of this wall as designed by some of the major manufacturers, with the CMU stretching horizontally between the columns. As mentioned in the previous section, we prefer a wall that extends vertically, with a pipe element at the top or a wide ring placed behind the wall. A cold-formed belt is probably not stiff enough for this application; A hot rolled duct would tend to sag under its own weight if no flex bars are present. In this system, the composite brick and block are connected by adjustable tie rods, which also transfer the lateral design loads to the outer limits of their movement. The function of the brick is not structural and all lateral loads are resisted by the CMU; 4-inch split-surface blocks or bricks can be used in place of bricks without changing the CMU design. All of the structural considerations discussed above for a single sheet masonry wall also apply to the CMU in this assembly. To improve the insulating properties of masonry cavity walls, rigid insulation can be added to the cavity; It can also run on the inside surface of the CMU between the support channels and the drywall. The wall void should be at least 2 inches wide, but the width can be reduced to 1 inch if rigid insulation or a drainage board is placed there. The base connection should face up behind the brick and be firmly embedded in a block joint. Masonry cavity walls are rarely viewed by designers or legislators as 'systems' in the sense that other curtain wall systems such as ETICS (see Section 7.7.2) are designed and tested. Instead, masonry is still viewed as a block-and-mortar assembly; both components are specified and tested separately. As Kudder11 points out, many masonry tests are available, but most relate to cavity walls. For example, there is no standard leak test for the entire cavity wall structure with ridges and control joints. And yet, as architects know, problems stemming from improper detailing and installation do more damage than masonry materials that do not meet specifications. Other common occurrences that can affect watertightness are void fill mortar and poorly filled mortar joints. Brick and CMU are not identical in nature and performance, although both are masonry materials.

WALL MATERIALS 238

CHAPTER SEVEN

FIGURE 7.28 Full height brick wall and CMU cavity wall with horizontal extension. (Butler Manufacturing Co.)

WALL MATERIALS WALL MATERIALS

239

Once manufactured, clay and concrete based products tend to move differently with changes in humidity. The brick, fired in the kiln, at first has practically no moisture, later it increases, constantly expanding. In contrast, CMU is born wet and hardens in a water-saturated state, gradually losing moisture and shrinking.12 One can only hope that by the time the two materials are installed, most of the initial volume changes have already taken place. Likewise, bricks and blocks react differently to temperature changes; Regular weight CMU can expand or contract up to 15% more than brick. More importantly, the outer brick is directly affected by solar radiation and experiences greater temperature swings than the brick separated by an air space. (A simple but often overlooked method of reducing thermal movement in brick is to use light-colored brick.) Differential thermal movement can be controlled by closely spaced expansion and control joints, e.g. B. 18 to 25 feet, to be mitigated. confused, but they are quite different in nature. Expansion joints in the brick are filled with compressible materials, while control joints in the CMU allow for shrinkage and are usually filled with rigid inserts to transfer forces perpendicular to the wall between adjacent blocks. The most common standard for exposed brick is ASTM C 216.13, which covers about 93 percent of all bricks sold in this country.14 ASTM C 216 includes three types of bricks: FBS for general use, FBX for applications requiring tight controller sizes, and FBA for deliberately non-uniform size and texture of units, sometimes favored by architects for a rustic look. Mortar affects the performance of the brick and must be chosen carefully. The lower the strength of the mortar, the more deformation the wall can withstand. A good general choice for brick facings, particularly those that will be exposed to severe freeze-thaw cycling, is mortar that conforms to ASTM C 27015 Type N. Type S mortar has greater flexural strength than Type N; Type M mortar is reserved for load-bearing brick applications. Unlike brick, CMU retaining walls often contain steel reinforcement and require S or M type mortar. Because brick and CMU require two different types of mortar, proper supervision of the installation is essential. There are many other intricacies of masonry design that experienced architects and engineers learn throughout their careers. It is critical to the good performance of masonry that designers provide all relevant details in the contract documentation and insist on good construction supervision, even when the rest of the building system is 'pre-engineered'.

7.5.2 Brick Cladding over Steel Beams This type of wall is common in conventional construction and valued for its visual appeal, light weight, ease of insulation and economy. Developed in the 1960s, the system consists of steel studs spaced 16 or 24 inches on center, interior drywall, construction paper or other caulking on the exterior siding, and brick veneer separated from the siding by an air space and fastened with steel pins adjustable metal loops (Fig. 7.29). The space between the wall studs is usually filled with fiberglass insulation. Some argue that this complex wall system quickly became very popular before its weaknesses were fully understood. In fact, there are very few sources of information on its long-term durability and proper construction details. One of the best is the Brick Industry Association (formerly the Brick Institute of America) Technical Note 28B.16 The bulletin provides detailed advice on design criteria, avoidance of water intrusion, minimum air space size, maximum anchor spacing, and other important aspects of brick facade design. It is not our intention to enter into a comprehensive discussion of this complicated system. Instead, we describe some crucial points for their performance in prefabricated buildings. A brick veneer wall over steel joists is a simplified example of a cavity wall rainbreak. It is expected that some water will eventually enter the cavity. Proper functioning of the waterproof construction paper, burr strength and drain holes is therefore critical to leak-free performance. To be effective, the bolt must extend through the wall and extend at least 1Ⲑ4 beyond its extent.

WALL MATERIALS 240

CHAPTER SEVEN

FIGURE 7.29 Brick cladding over steel wall. (Used with permission from the Brick Industry Association, Reston, VA, www.brickinfo.org.)

front face; the raised tip should form a trickle when bent 45°.17 Some architects hate the look of the exposed glossy edge, but the ridge terminating inside the brick invites water back into the cavity underneath, or worse to flow into the brick. freeze there. The port should be enlarged and extended 6 to 9 inches above the penetration point and embedded in the case or in a reglet. It is important to lift and seal the edges of the flashing on the sides as well to prevent water escaping through the sides; Prudent architects provide an isometric detail of this state. The most durable deburring materials are stainless steel and lead-coated copper, but they are not easy to work with. Drain holes are located just above the flashing at 16 to 24 inches. in the centers. The cavity width, as mentioned above, should be at least 2 inches; Everything else is difficult to keep free of mortar fouling, which can carry water through the cavity. Why is it important? Doesn't water have to get into the cavity anyway? The answer is yes, but the less water the better as the water or even moisture left in the cavity will attack the metal joints and their joints. Adjustable anchors anchor bricks to steel studs and transfer lateral loads to them, acting as mini columns. Loops are best made from thick wire (3Ⲑ16 inches); The fine corrugated metal loops that house builders are familiar with are taboo in civil engineering. An adjustable loop is attached to the anchor, connected to a steel pin; the anchor must allow a height adjustment of at least half the brick height. Some common types of adjustable loops are shown in Fig. 7.30. In seismic zones, building codes may dictate that longitudinal wire reinforcement be embedded in the faced masonry and attached to the anchors (Fig. 7.31). The threads improve the ductility of the veneer and hold it in place when severe cracking occurs during strong earthquakes. Most often, brick trusses are attached to studs with self-tapping screws driven through the siding. The bolt-to-bolt joint is a crucial bond that holds the veneer upwind, which explains why moisture in the cavity is such a big problem: after all, corrosion at this vital joint can result in a tiled floor. Cable ties and pins can be made of galvanized steel, but screws are usually protected only by a corrosion-resistant coating that is easily damaged during installation. Stainless steel fasteners offer excellent corrosion resistance, but may cause corrosion when in contact with bare or galvanized steel. At least one prominent engineer points out that all masonry contains some salt and laments this

WALL MATERIALS WALL MATERIALS

241

FIGURE 7.30 Building block assemblies. (Used with permission from the Brick Industry Association, Reston, VA, www.brickinfo.org.)

the brick "is literally hanging by a thread from the building, and the thin, fine threads of the exposed wire of a steel screw may be bathed in a salt solution from time to time."18 While others see the situation as less apocalyptic, the threat looms large, clearly there. The author's practice is to specify zinc plated studs (G90 coating designation) of at least 18 gauge, regardless of strength requirements, simply to provide greater metal thickness in the joint. This practice is now endorsed by the BIA.16 Many engineers, including the author, have looked for ways to completely eliminate bolts from anchors. Unfortunately, possible alternatives using pop rivets or small screws instead of screws would be much more labor intensive than current practice. In addition, special waterproofing requirements must be considered to prevent water from penetrating through the resulting large holes. Currently, the use of self-drilling screws and dowels of the type shown in Fig. 7.30. A special baked copolymer coating such as Stalgard* has been found to be more effective in terms of corrosion protection and abrasion resistance than cadmium or zinc coating. To further restrict water access to the joint, Gumpertz and Bell19 suggest installing a piece of compressible gasket made of EPDM or similar material behind the base anchor. In addition, the screw must have a built-in neoprene washer or a separate rubber ring. As if the issues with curbs and brick sleepers weren't complicated enough, there's also a controversy surrounding lateral stiffness requirements. Remember that flexible steel studs are intended to provide "structural" support (side support) for "non-structural" brittle brick veneers. The problem is that the brick can break long before the flex pins take their bent shape. An obvious solution is to harden the pins with deeper sections of thicker metal. But harden by how much? BIA Note 28B recommends that the maximum deflection of the steel pin reinforcement when considered alone at full lateral working load is L/600. There are people who consider this limit not strict enough, pointing out that at deflections less than a third of it - L/2000 - the brick will crack from L/ 360.20 To arrive at a solution is a rigorous analysis considering the stiffness of the veneer, Steel *Stalgard is a registered trademark of Elco Industries Inc.

WALL MATERIALS 242

CHAPTER SEVEN

Brick nails and loops need to be made. Such finite element analysis and investigation by Gumpertz and Bell19 concluded that the BIA limit of L/600 is reasonable. The investigation also found that the distribution of wind power between multiple brick sleepers along the wall height is not uniform: two or three sleepers located at the top and bottom of a beam carry a significant portion of the total wind power. Therefore the total wind load inflow for one stud should be divided into four or six and the result should serve as the basis for the design wind load on all studs. BIA Note 28B recommends that one masonry veneer tie be installed for every 2.67 ft2 of wall area. Loops should be spaced no more than 18 apart vertically and 32 apart horizontally. They must be recessed at least 1.5 inches. in veneer, but still at least 5Ⲑ8-in. Mortar covering the outer surface of the brick. As mentioned above, this information only scratches the surface of the complex issues surrounding brick veneers. Those involved in specifying bricks in metal building systems should be aware of the latest relevant regulations and ever-improving standards for brick wall design and detailing.

7.6 COMBINED WALLS Metal building systems offer many opportunities to combine different materials in the same wall. For example, external masonry may only be provided for areas on the ground where the potential for physical abuse is greatest. of high seismicity. (Used with permission from Freemasonry. May also be chosen as wall base material for aesthetic purposes, Brick Industry Association, Reston, VA, to add depth and interest to a flat or ribbed facade. www.brickinfo.org.) Partially face high elevation masonry walls pose some design challenges in terms of their lateral support at the top. As previously mentioned, typical cold-formed wall beams are generally not suitable for lateral support of masonry due to differences in stiffness. The best method of stabilizing a half-height masonry wall ("panelling") is to drop vertical rebar into the foundation wall so that a cantilever effect is created. Making these walls taller results in a dramatic increase in rebar size, so it is best to keep part-height walls relatively short (Fig. 7.32). The most expensive and difficult project is taking the CMU almost to the eaves and completing it in a tape window. If such a situation arises, consider adding a few discrete windows at the top instead of one solid window. If the ribbon window design is unavoidable, steel construction beams can be provided above and below the windows. The waist design is similar to Example 7.2. A common wall combination is a partially tall CMU with metal panels above. Panels can be attached to the CMU using a parapet rail or parapet angle (Fig. 7.33a and c), whereby the masonry must be statically designed for the additional wind load of the panel. In another possible solution, the siding is laterally supported at the bottom by its own belt (Fig. 7.33b). Metal wall panels with different profiles can be combined to form accent strips and enliven the façade (Fig. 7.34). In this project, the bottom slab can be made not only of metal, but also of masonry (self-supporting, as mentioned above) or even precast concrete. Occasionally, a combination of multiple wall materials can be avoided when metal panels can play a role in other building components such as shutters. On a project in Puerto Rico, custom shutters were made from 18-gauge metal siding panels that were bent into a shutter-like configuration and extended horizontally. The resulting product not only blended well with the surrounding wall panels, but was also much less expensive than standard extruded metal blinds.21

WALL MATERIALS WALL MATERIALS

243

FIGURE 7.32 This partially tall CMU wall faces a hillside and provides some protection from rolling rocks and boulder slides.

7.7 CONCRETE MATERIALS 7.7.1 Precast concrete elements Precast concrete elements offer some of the advantages of masonry, such as: B. Impact, sound and fire resistance without the disadvantage of slow construction. Load bearing and shear wall applications of prefabricated panels are particularly popular in metal building systems. Non-structural prefabricated wall panels can add depth to the building (Fig. 7.35). To develop a distinctive wall structure, prefabricated panels can be designed with deep horizontal or vertical grooves, possibly with different groove spacing; Raised rib panels can also add visual interest. Even unlikely elements such as prefabricated double-T roof panels have been used successfully as low-cost, long-range curtain walls. One of the main disadvantages of precast concrete is its low thermal insulation value, but this problem can be overcome. Rigid insulation can be attached to the inside of the panel and covered with plasterboard, or it can be placed between a thick structural member and a thin outer layer. For example, a 12 inch insulation slab may be made from 8 inch solid or hollow structural member, 2.5 inch expanded polystyrene and 1.5 inch exposed aggregate concrete exterior. These panels commonly come in widths of 8 to 12 feet. Many engineers neglect any structural contribution of the outer layer and make it completely independent from the structural part. Bonding the two layers can allay concerns about panel delamination, but can also result in panel curling and some loss of R-value. Insulated wall panels appear to be particularly popular in food processing plants, where strict housekeeping regulations call for hard, smooth interior surfaces of crevices, holes and horizontal protrusions. For this application, plates are examined with binoculars to detect the smallest surface imperfections that can become havens for dust and bacteria.22

WALL MATERIALS 244

CHAPTER SEVEN

FIGURE 7.33 Details on top of half-height masonry wall: (a) with parapet duct; (b) with a girded base; (c) with the angle of the threshold. (Metal construction systems.)

FIGURE 7.34 Detail of accent stripe. (Star building systems.)

Designers are increasingly choosing to provide panels with an exposed aggregate finish. The American Concrete Institute's ACI 533R, Guide for Precast Concrete Wall Panels23 distinguishes between light, medium and low aggregate loading for a variety of visual effects. Architects' arsenal of surface treatments for precast concrete also includes formwork coatings, which can create surfaces ranging from wooden planks to double-sided masonry and the use of white and pigmented cement. The structural design of precast concrete is governed by ACI 31824 and is similar to cast-in-place concrete except as noted in ACI 533R. Typically, concrete is specified with a 28-day compressive strength of 5000 psi or greater for durability. The reinforcement can consist of deformed bars, welded wire mesh or tendons; Epoxy coating or electroplating is common to protect against corrosion. A typical structural slab spans the distance from the foundation to the roof. The lower connection is made by welding the plates embedded in the plate with those in the foundation, any gaps bitumened and sealed. The upper connection that supports the purlins of the roof is made by field weld angles that fit embedded plates (Fig. 7.17). Panel joints are treated similarly. The minimum thickness of conventionally reinforced precast slabs varies between one twentieth and one fortieth of the unsupported length. Non load-bearing prefabricated panels may be laterally supported by heavy wall joists at their tops. Although reinforced concrete has a higher degree of ductility than masonry, careful consideration of the required chord and frame stiffnesses is required, especially when the floor slab connection is close to solid. If the bot

WALL MATERIALS WALL MATERIALS

245

FIGURE 7.35 Prefabricated wall panels add depth to the facade. (Photo: Maguire Group Inc.)

When the tom joint is pinned, the forces acting on the strap can be determined by the direct method described in Example 7.1. The lattice construction can follow Example 7.2. The detail of a lace connection maker is in Fig. 7.36. In this detail, the strut is attached to the panels embedded in the panels via H-shaped welded clips. This very schematic detail does not show how the strut is welded to the chord and panel, how (and if) the roof membrane is attached to the Precast slab and how the walers are braced. Some of this information is shown in Fig. 7.23. Precast concrete design and detailing is a very specialized field; is best done by panel vendors. Nonetheless, the information needed to communicate design intent must be included in the contract documentation. The amount of this information varies greatly between design firms: in some, each panel is designed and detailed by the architect-engineer; in others only the panel layout is shown on the building elevation drawings. A common approach is to design and plan in detail a typical panel in-house, with the project including panel thickness, joint size and a general method of attachment. The rest can be left to the prefabricator who must provide full drawings and calculations including structural calculations for handling, transport and thermal loads. Architects are often concerned about the acceptability of panel finishes and casting or installation tolerances. Designers should read the relevant provisions of ACI 533R and the PCI Design Handbook25 and be familiar with realistic visual and dimensional variations from factory manufactured panels.

7.7.2 Swing Panels While precast concrete elements are formed in the workshop and transported to the site, Swing Panels are generally cast directly onto the building slab. After a one-week hardening phase, the panels are "tilted" - lifted by crane at suitable points - and supported. The 'righting' usually takes place prior to assembly of the prefabricated structure to avoid degrading the steel.

WALL MATERIALS 246

CHAPTER SEVEN

An obvious benefit of the tilt-up process is local manufacturing, which avoids transportation costs and shipping restrictions on panel sizes; Swing panels can be as large as the available crane will allow. A downside, of course, is the loss of facility sophistication and quality control. While swing panels are typically flat, a variety of reveals and surface treatments such as exposed aggregate and sandblasted finishes expand the designer's options. As with precast construction, wall cladding can simulate the appearance of brick or stone. To facilitate removal of the panel, the cast panel is sprayed with a link breaker; Board curing compound is generally suitable for this purpose. Most tilting panels have an uneven or blotchy appearance that will improve slightly over time; but the panels are often painted, often with acrylic-based coatings.26 Some architects take this opportunity to create a trompe-l'oeil effect, a painted pattern that looks like a three-dimensional structure from a distance.

FIGURE 7.36 Connection in non-structural precast slab. (Constructive appendix systems.)

WALL MATERIALS WALL MATERIALS

247

The sides of swing panels are usually formed from custom lumber, with 51Ⲑ2 or 71Ⲑ4 thick panels being common. A rule of thumb limits the panel thickness to one fifth of the unsupported height. (This thickness is added to the depth of all ribs.) In general, the structural considerations for swing panels are the same as for precast panels. Calculations related to lifting loads and the selection of flush anchors are often performed by specialist engineers. A common panel design has a single reinforcement mat located at mid-depth with the bars spaced 12 to 16 inches apart. Downtown. Like precast panels, concrete swing walls can be made up to 6 inches. built-in insulation. A typical sandwich panel may consist of a 2 to 3 inch thick outer layer of concrete, 2 inch thick extruded polystyrene insulation, and an inner structural layer. Recently, fiber composite connections between panel layers have become popular, freed from the disadvantages of thermal bridging endemic to panels with traditional metal ties. The inclined construction places special demands on the building ceiling used for the formwork: the ceiling has to bear the weight of the panels and the loaded crane (assembly usually takes place from inside the building). A minimum panel thickness of 5 to 6 inches is recommended.26 In addition, panel finishing requires special attention as any surface irregularities will be reflected in the finished product. For best results, panel joints are pre-planned to match panel joints. Plate joints are typically 1Ⲑ2 to 3Ⲑ4 inches. Broad; they are sealed on both sides after assembly. For additional technical guidance, the reader is referred to ACI 551R, Tilt-Up Concrete Structures27 and Brooks.28 Tilted construction can be very economical for medium-sized buildings with high sidewalls (20 ft⫹) and a repetitive appearance. Most oscillating plates are used as load-bearing elements and shear walls. Tilt-up is especially popular in good weather areas like Florida and Southern California. According to the Tilt-Up Concrete Association, this wall system accounts for more than 15 percent of new industrial building construction. to put in place. The cast-in-place exterior may also be required to accommodate lateral pressure from loose materials stored within, as in most material recycling and resource recovery facilities that require external concrete "compression walls". As cantilever retaining walls, these compression walls must be rigidly connected to a horizontal base or foundation walls; There are few alternatives to throwing them instead. The structural design and finishes of the full-height cast-in-place walls are similar to the tilting structure, although the cast-in-place walls are thicker to allow for placing the concrete in an upright position. For weather resistance of exterior walls, high strength concrete mixes with entrained air are used. Control joints in cast-in-place walls are often made with rough strips spaced 20 to 25 feet apart, perhaps to match cove spacing. To facilitate cracking in a joint, the amount of horizontal reinforcement going through the joint can be halved.

7.8 OTHER WALL MATERIALS 7.8.1 Glass Fiber Reinforced Concrete (GFRC) This relatively new material - used in this country since the 1970's - is made by mixing glass fibers into a sand-cement paste and spraying the mixture into moulds. Thin (approximately 5Ⲑ8 inches), lightweight and durable, GFRC panels offer an attractive alternative to precast concrete and stone. Since the shape of the molds is only limited by the designer's imagination, a variety of impressive shapes can be achieved. GFRC offers unsurpassed sharpness of detail due to the smaller aggregate size. Finishes available include grit blasting and special coating mixes for decorative aggregate in the 1Ⲑ8 to 1Ⲑ2 inch range. of thickness.

WALL MATERIALS 248

CHAPTER SEVEN

Randomly distributed glass fibers may only account for 5% of the panel weight, but the resulting increase in ductility is remarkable. The GFRC skin is supported by a lightweight steel frame with rigid ("gravity") and flexible ("wind") anchors embedded in thick plate pads. Total panel weight is 8 to 20 psf. Installation costs are high—ranging from $20 to $50 per square foot30—and this limits the use of GFRC for accent pieces, decorative parapets, column copings, elaborate cornices and fascias. While GFRC panels can add significant design interest, they must be carefully specified by following PCI Recommended Practice. Damage during manufacture or assembly may go unnoticed for a long time. Nicastro32 tells the story of some severely warped GFRC panels that were "fixed" on site by a contractor and developed multiple cracks a year after assembly. In the author's own experience, panels are sometimes cracked (mainly during shipping), discolored and shipped with detached anchor points. It is therefore advisable for architects to inspect the panels prior to installation and to insist that the manufacturer's operations comply with the PCI Quality Control Manual.33 All of these precautions do not guarantee success. The GFRC just hasn't been used long enough to reveal all of its limitations. For example, the GFRC cladding of a 21-story office building in Texas mysteriously developed widespread cracks and eventually had to be replaced with a different cladding system.34 The problem was attributed to several factors. On some panels, the "wind" anchors have been made too stiff, limiting the expansion and contraction of the panels. Other delaminated plates on the GFRC surface and base blend interface; The delamination was attributed to differential heat and moisture movement - and the build-up of stresses - between the two materials. Numerous manufacturing defects were also noted. One can only hope that once the weaknesses of GFRC are clearly understood, this promising material will overcome its problems and become commonplace in metallic building systems. 7.8.2 Insulation and Facade Finishing Systems (ETICS) Born in Europe, ETICS were introduced to this country by Dryvit System, Inc. in the late 1960's and bore that name for some time. Today this system is offered by many other companies represented by the EIFS Industry Members Association (EIMA). The association publishes policy specifications, technical notes, and other useful information about the product.35 The system most commonly used in metalwork involves steel bolts spaced from 12 to 32 inches. in the middle, outer cladding, rigid insulation glued to cladding with adhesive, base course, reinforcement fabric and top layer (Fig. 7.37). There are two generic classes of ETICS: polymer-based (PB) and polymer-modified (PM). PB systems are made from thin, flexible materials and are much more popular than the thicker PM or "hard coat" cement products. The main advantages of ETICS are design flexibility, high insulation value and low cost. ETICS allow a variety of shapes and surface structures (Fig. 7.38); The systems can be mounted to existing surfaces in the field or fabricated in lightweight steel framed panels. EIFS became incredibly popular, but the failure rate was also dramatic. The author was closely involved with the production of EIFS sheets in the early 1980's and subsequently witnessed failures in many of these applications. Recent reports show that the first ETICS introduced in the United States were not as good as their European brethren. Today's designs are much more advanced and will likely perform better. However, products that only meet the minimum EIMA criteria may not be safe; Some additional requirements greatly increase the chances of a long-lasting ETICS case, albeit at an additional cost. Below is an example of the expert recommendations made by Piper and Kenney36 et al.37,38 ●

PB coatings are flexible but cannot cover large gaps and defects in the substrate without cracking. A classic source of cracking is the sloppy joint between rigid pieces of insulation that, instead of being filled with insulation chips, are left unfilled or filled with glue.

WALL MATERIALS WALL MATERIALS

249

FIGURE 7.37 Detail of the base of the ETICS board.

●

Although gypsum siding is acceptable under EIMA, it can absorb moisture and lead to failure. Instead, Piper and Kenney recommend using fiber cement board or proprietary products that conform to ASTM C 1177, such as B. Georgia-Pacific's Dens Glass* or the newer Dens Glass Gold. Expanded polystyrene insulation materials must have a good weld; their joints should not be flush with the edges of openings, lining joints or rustication. The base course should not contain more than 33% cement, although many products on the market contain 50% cement. The extra cement not only reduces the flexibility of the coating, but also increases its alkalinity, which can degrade an alkali-resistant grid coating and eventually corrode the grid. A common base coat thickness is 1Ⲑ8 inches, which does not provide adequate moisture protection. For best results, the base coat should be at least 3Ⲑ32 inches thick and applied in two coats. Use low modulus sealers applied over the primed basecoat areas, not over the topcoat as is commonly done. Silicone sealant will probably do the best service.

*Dens Glass is a registered trademark of Georgia-Pacific.

WALL MATERIALS 250

CHAPTER SEVEN

FIGURE 7.38 ETICS enables various economic options.

● ●

Avoid using the ETICS on inclined surfaces with a gradient of less than 1:1. To protect against delamination, attach the ETICS to the substrate with mechanical fasteners and adhesives.

While some of these recommendations may be questionable, they do illustrate the evolving state of the art in the ETICS industry and the vast differences between the products available. Do not treat this wall system like a commodity.

7.8.3 Thin Veneer Panels on Steel Frame Occasionally architects wish to use a wall system that looks like stone, brick or precast concrete but weighs much less. Thin veneer panels can provide an answer. A typical panel consists of thin sheets of veneer supported by light steel posts or a mild steel frame. The thin veneer, which can be stone, precast concrete, and thin sheet of brick, is attached to the frame with steel anchors and brackets. Panels can be insulated. Thin veneer panels were primarily developed for medium to high rise building and building retrofit applications where light weight and rapid panel installation are highly valued. The system is no less suitable for metallic construction systems that value the same properties. The main disadvantage of thin veneer construction - as well as GFRC, EIFS and any other single ply system - is the lack of an air cavity. All of these systems do not conform to the rainproof principle, but rely on joint sealant to protect against moisture and are susceptible to corrosion of the fasteners, which requires regular inspection and maintenance. It is true that many EIFS manufacturers now offer so-called drainable assemblies, also known as water management systems or rain covers, in which the rigid insulation boards are supplied with grooves scored in the factory. The slots are provided

WALL MATERIALS WALL MATERIALS

251

act as drainage channels that allow moisture to escape. However, a true cavity wall must contain ridges and drainage holes; these are absent from some of the drainable systems. Both elements could spoil the appearance of the wall if they were included in every panel wherever it is on the facade. The long-term performance of these assemblies needs to be further investigated.

7.9 SELECTING A WALL SYSTEM The selection of exterior wall materials fits logically into an overall assessment of available building systems, which is discussed in Chap. 3. In this chapter, the advantages and disadvantages of different wall materials have been discussed. The environment often dictates the choice of finish, color and texture; Sometimes the customer makes a big contribution, for example by insisting on easy-care surfaces. Functional requirements often determine whether a 'hard' wall is required, at least at the base, and whether the wall needs to be insulated. A building with a lot of forklift traffic may need hard walls unless some type of wall protection is used. Likewise, a warehouse used to store valuable goods is not well protected by metal siding. As previously mentioned, a food processing facility requires a hard, smooth interior finish, which is best achieved with precast concrete. Fire resistance criteria can also limit the choice of masonry or concrete. Aside from these considerations, the trade-off is between aesthetics and cost. It is extremely important to remember that the exterior walls are part of the metal construction system; they have to fit in visually, structurally and functionally. Non-ferrous wall systems, and especially their joints, must be carefully evaluated for compatibility with metal structures. Unfortunately, low-rise curtain walls seem to have more problems than their high-rise counterparts, perhaps due to the limited design budgets commonly available. However, the designer's potential liability for poor coordination or reckless product selection knows no such bounds. A useful reservoir of ideas on how to combine different materials to achieve the desired effect can be found in Metal Architecture and other metal fabrication industry publications.

LITERATUR 1. Metal Building Systems, 2. Aufl., Building Systems Institute, Inc., Cleveland, OH, 1990. 2. Bob Fittro, „Wide Variety of Metal Wall Panels Available for Any Application“, Metal Architecture, November 1998, S. . 38–39. 3. W. J. Fleeman, „Insulated Metal Panels: Past, Present and Future“, Metal Architecture, Dezember 1997, S. 26–38. 4. John A. Hartsock und W. J. Fleeman, III, „How Foam-Filled Sandwich Panels Work“, Metal Architecture, Mai 1994. 5. Richard Keleher, „Rain Screen Principles in Practice“, Progressive Architecture, August 1993. 6 ASTM C 90, Standard Specification for Load Bearing Concrete Masonry Units, ASTM, West Conshohocken, PA, 2002. 7. Building Code Requirements for Masonry Structures (ACI 530/ASCE 5/TMS 402), Society American Institute of Civil Engineers, New York, 2002 8. „Concrete Masonry Design Tables“, TR 121, National Concrete Masonry Association, Herndon, VA, 2000. 9. Manual of Steel Construction, Allowable Stress Design, 9. Ausgabe, American Institute of Steel Construction, Chicago, IL, 1989 10 . Specification for Structural Steel Buildings, Allowable Stress Design and Plastic Design, American Institute of Steel Construction, Chicago, IL, 1989. 11. Robert J. Kudder, „Deliver Masonry as a Cladding System“, Masonry Construction, August 1994. 12. Stephen Szoke und Hugh C. MacDonald, „Combining Masonry and Brick“, Architecture, Januar 1989.

WALL MATERIALS 252

CHAPTER SEVEN

13. ASTM C 216, Standard Specification for Facing Brick (Solid Alvenry Units Made from Clay or Shale), ASTM, West Conshohocken, PA, 2002. 14. Brian E. Trimble, „Building Better with Brick“, Building Design and Construction, August 1995. 15. ASTM C 270, Standard Specification for Mortar for Unitary Masonry, ASTM, West Conshohocken, PA, 2002. 16. „Brick Clad Steel Stud Panel Walls“, Brick Construction Technical Notes No. 28B, rev. II, Brick Industry Association, Reston, VA, 1999. 17. Douglas E. Gordon, „Brick Basics“, Architectural Technology, Herbst 1986. 18. Clayford T. Grimm, „Brick Veneer: A Second Opinion“, The Construction Specificationr, April 1984. 19. Werner H. Gumpertz und Glenn R. Bell, „Engineering Evaluation of Brick Veneer/Steel Stud Walls“, Teil 1, „Flashing and waterproofing“, Teil 2, „Structural Design, Structural Behavior and Durability“, Proceedings , Third North American Conference of Freemasonry, Arlington, TX, Juni 1985. 20. Hier sind die Fakten über Stahlrahmen – Design of Brick Veneer Systems, Metal Lath/Steel Framing Association, Chicago, IL. 21. „Modified Panels Used for Wall Shutters in Fabrication Facilities“, Metal Architecture, Oktober 1995. 22. Craig A. Shutt, „Food Processors Savouring Insulated Precast Precast Wall Panels“, Ascent, Sommer 1995. 23 Guide for Precast Concrete Wall Panels , ACI 533R-93, American Concrete Institute, Detroit, MI, 1993. 24. Building Code Requirements for Reinforced Concrete, ACI 318, American Concrete Institute, Detroit, MI, 2002 25. PCI Design Handbook, Precast and Protensioned Concrete, 5. Aufl . ., Precast Pretensioned Concrete Institute, Chicago, IL, 1999. 26. Ed Santer, „Tilt-Up Basics“, Concrete Construction, Mai 1993. 27. Tilt-up Concrete Structures, ACI 551R-92, American Concrete Institute, Detroit, MI, 1992. 28. H. E. Brooks, The Tilt-Up Design and Construction Manual, 2. Aufl., H.B.A. Veröffentlichung, Newport Beach, CA, 1990. 29. Maureen Eaton, „Tilt-up Concrete Offers Economical Construction Option“, Building Design and Construction, Mai 1995. 30. Gordon Wright, „GFRC Provides Lightweight, Vielseitige Verkleidung“, Building Design and Construction, Dezember 1988. 31. Recommended Practice for Glass Fiber Reinforced Concrete Panels, MNL-128, Precast/Protensioned Concrete Institute, Chicago, IL, 2001. curtain wall", Exteriors, Summer 1988 33. Quality Control Manual for Plant and Production of Glasfaserverstärkte Betonprodukte, MNL-130, Precast/Prestressed Concrete Institute, Chicago, IL, 1991. 34 Gordon Wright, „Unangemessene Details verursachen Beschichtungsprobleme“, Konstruktionsplanung und Konstruktion, Januar 1996. 35. EIMA-Richtlinienspezifikation für Außendämmung und Finish Systems, Klasse PB und andere Veröffentlichungen, EIFS Industry Members Association, Clear Water, FL, 1994. 36. Richard Piper und Russell Kenney, „EIFS Performance Review“, Journal of Light Construction, Richmond, VT, 1992. 37. Michael Bordenaro, „Avoiding EIFS Application Pitfalls“, Building Design and Construction, Apr. 1993. 38. Margaret Doyle, „Trends in Specificing EIFS“, Building Design and Construction, August 1988.

REVIEW QUESTIONS 1 What changes need to be made to the metal structure of the building if hard walls are used? 2 How do façade panels improve the load-bearing behavior of external beams?

WALL MATERIALS WALL MATERIALS

253

3 Why should EIFS walls be specified in metallic building systems? What might be some of the issues to be aware of when doing this? 4 Explain cavity wall concepts and the rain protection principle. 5 What components, not normally found in metallic building systems, should be added when the exterior walls of the CMU are vertical? What are some design criteria for these elements? 6 What happens to column flange braces when using hard walls? 7 Explain what measures should be taken to improve the weather resistance of external steel beam and sheet piling walls. 8 List at least two methods of laterally supporting an 8 foot high CMU exterior wall in a metal frame system. 9 Suggest at least four common structural details at the base of the metal panel. Explain their relative advantages and disadvantages.

WALL MATERIALS

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 8

ISOLATION

8.1 INTRODUCTION * Since the oil crisis of the 1970s, the demand for energy efficient buildings has increased steadily. Model building codes have already issued many regulations on energy saving; Federal regulations such as the Energy Policy Act of 1992 also emphasize the construction of energy efficient buildings. To define uniform design requirements for energy conservation, the American Society of Heating, Refrigeration and Air-Conditioning Engineers (ASHRAE) issued its Standard 90.1.1, which is quickly being adopted by building codes. The 1999 edition of this standard considers metallic roofs and walls of buildings as distinct elements of the building envelope. In the past, insulation issues were not the focus of the metal fabrication industry. In fact, inadequate insulation is still one of the most common complaints about prefab homes. This chapter provides an overview of some available insulation products, systems and details to help designers make informed decisions about materials and installation methods. It will not delve too deeply into the realm of HVAC engineering and will cover topics such as thermal stresses, energy conservation and equipment selection. It will also avoid the problems of mass consideration, annual heating loads, and life cycle costs for metal building systems, which are well covered in Metal Building Systems from the Building Systems Institute.2

8.2 THE BASICS OF INSULATION DESIGN Heat loss or gain in a building can occur in three ways: radiation, conduction and convection. Of these, conduction through the building envelope accounts for most of the heat transfer. Heat losses through conduction can be reduced with additional insulation, but cannot be eliminated. Convection from air leaks can and should be avoided by 'squeezing' the building. Radiation primarily affects glass surfaces and can be minimized with reflective coatings. Because the outer walls contain openings and are thermally non-uniform, heat transfer through multiple parallel heat flow paths can be considered separately. Building Codes and ASHRAE 90.1 specify a certain required level of thermal conductivity (or thermal transmittance) U0 for roof and wall attachments at various locations, expressed in Btu/(h)(ft2)(°F). Thermal transmittance is a reciprocal function of an assembly's thermal resistance, R, which is a sum of the R-values of the components, including those of the inner and outer layers of air and any air voids. Heat transfer is usually accompanied by moisture movement because warmer air contains more water vapor than cold air. The movement of the vapor need not coincide with the actual movement of the air containing it. When warm air cools or hits a cold surface, it loses some of its moisture and condensation forms. The air temperature at which condensation begins is the dew point.

*The author would like to express his sincere gratitude to the Steel Construction Committee of the North America Insulation Manufacturers Association (NAIMA) for their contribution to this chapter and for permission to reproduce illustrations from their publications.

255 Conveyed from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

INSULATION 256

CHAPTER EIGHT

Condensation can lead to metal corrosion, mold growth, loss of insulating properties and ruined surfaces. It can be minimized by installing a vapor barrier on the hot side of the wall (more precisely, on the side with the highest vapor pressure). The vapor barrier slows the transport of moisture to a cooler surface. Unfortunately, it is not always easy to determine where to place the vapor barrier. A cliché of keeping it inside in cold weather and outside in hot weather only goes so far, for in many places the warm side is the inside of the wall during the winter months; during the hot summer months it is the outer surface. Most roofing and siding are virtually impervious to vapor and may well act as a vapor barrier in hot weather; is the cold weather anti-condensation normally needed for indoor surfaces. The term vapor barrier is more appropriate than a vapor barrier that is often used, since building materials cannot completely stop the movement of moisture, only slow it down. Retarders may not be necessary in moderately dry climates, but are important in humid locations and in buildings where moisture is released.

8.3 TYPES OF INSULATION The type and thickness of insulation have the greatest impact on a building's thermal efficiency. In fact, spending money on insulation can be one of the best investments a building owner will ever make. Properly selected roof insulation contributes even more to the thermal performance of a one-story metal building than a multi-story building. All insulating features trap stagnant air, which slows conductive heat transfer through the insulating medium. The different types mainly differ in how this is done. Four basic types of insulation are available for metal building systems: fiberglass, rigid, sprayed and foam core.

8.3.1 Fiberglass Mat Insulation Fiberglass works in a similar way to a fur coat: both trap air on the surface of many individual fibers. Fiberglass mats are the most common type of insulation used on roofs and walls of prefabricated buildings due to their low cost, fire and sound resistance, and ease of installation. The R-value of fiberglass insulation ranges from 3 to 3.33 per inch of thickness. Mats are usually supplied with a vapor barrier 'laminated' to the fiberglass (Fig 8.1a). In addition to fulfilling their primary purpose, vapor barriers often serve as the only ceiling finish for metal buildings. therefore the coating is generally white for better light reflection. Fiberglass insulation for metal buildings is not exactly the type used by homeowners to insulate their attics. First, it's wider, closer to typical metal joist and purlin spacing (5 feet or more) than wood joist and joist (16 or 24 inches). Second, a different type of vapor control layer is used, as discussed in Section 8.4. Insulation certified to the North American Insulation Manufacturers Association (NAIMA) Standard 202 must meet strict criteria, such as: It has been tested for fire safety, corrosion, mildew, odor and moisture resistance.3 The NAIMA 202 certification is printed on the uncoated side of fiberglass rolls. This insulation has been tested with all UL-approved coatings, according to the association. In addition to NAIMA certification, some designers also require compliance with ASTM E 5534. Another product name designers should be aware of is Certified Faced Insulation by the National Insulation Association (NIA). This standard was formerly known as NIA 404, a Standard Product Specification for Metal Building Insulation with Flexible Fiberglass Covering. The use of NAIMA 202 fiberglass insulation allows the fairing laminators to meet the NIA standard.

INSULATION INSULATION

257

FIGURE 8.1 Four basic types of insulation used in metal construction: (a) fiberglass with a laminated surface; (b) foam boards; (c) pre-insulated panels; (d) powdered cellulose. (Courtesy of the NAIMA Steel Construction Committee.)

Some information about NAIMA, which represents manufacturers of fiberglass insulation, can be found in Chap. 2. In contrast, the NIA primarily represents contractors, traders and mills. Yet another type of fiberglass insulation for prefabricated buildings is metallic building insulation (MBI), which lacks a vapor retardant and is intended to serve as a second layer over the exposed insulation. MBI is available in batts and rolls. Metal Building Insulation - Plus from Owens-Corning is an example of this type. Fiberglass insulation is manufactured by a few large companies (mainly NAIMA members) and is usually purchased from the manufacturer of the laminators that apply vapor control layers. Some of the leading metal building manufacturers, such as VP Buildings and Butler, offer a full range of insulation products for all purchases.

8.3.2 Rigid Insulation Rigid insulation, also known as foamboard, works by trapping air in multiple individual foam cells. Foam is most commonly made from polystyrene, polyisocyanurate, and phenolic materials (Fig. 8.1b). Rigid insulation offers excellent thermal efficiency (R-value), vapor inhibition, and dimensional stability, but lacks acoustic performance. Rigid insulation is more expensive to manufacture and install than fiberglass. The insulation can be applied and coated both inside and outside the frame

INSULATION 258

CHAPTER EIGHT

various surface treatments; may require protective boards during construction. All types of rigid foam insulation gradually lose their insulating value over time and the R-values used in thermal analysis should be based on 'aged' properties rather than initial ones. Polystyrene insulation comes in two types: expanded polystyrene ("beadboard"), an open-cell material sometimes specified for protection of perimeter foundations, and extruded polystyrene, which is used for wall and roof insulation. Extruded polystyrene, found in disposable coffee cups, among other things, has a closed-cell composition with better vapor-retardant properties than an open-cell variant. Expanded polystyrene is water permeable and for this reason is generally avoided in roofing applications. Due to the water permeability, however, it is suitable for some wall systems such as ETICS. Key benefits of extruded polystyrene include an R-value of 5.0 or slightly higher and a compressive strength of 30 to 40 lb/in2. The main disadvantage is the flammability. Polyisocyanurate insulation, essentially a modified urethane foam, has the great advantage of being fire resistant. It is thermally more efficient than polystyrene, with R values ranging from 5.8 to 7.2. However, it is also more compressible and brittle than styrofoam and less water resistant. Therefore, polyisocyanurates are typically covered with coating materials and depend heavily on them for stability and moisture resistance. If the top layers delaminate, the insulation is at risk. Unfortunately, this problem is very common, especially with thicker panels.5 There is also the question of the insulation's coefficient of thermal expansion and contraction, which appears to be quite high. So-called polyiso insulation is less expensive to install than extruded or expanded polystyrene, and because it's also more efficient, it has a much higher return on investment than polystyrene. This fact undoubtedly helps explain the fact that polyisocyanurate insulation is used in half of all new commercial construction.6 One of the most popular polyisocyanurate products is Celotex Corporation's Thermax*. The Thermax consists of a glass fiber reinforced polyisocyanurate foam core with aluminum foil on both sides. The sills are available in a variety of finishes: the visible sills are usually given an embossed white finish, while the hidden side typically has a reflective finish. Available insulation thicknesses range from 1Ⲑ2 to 3 inches; The standard panel width is 4 feet. Thermax was specifically designed to deaden noise generated by the movement of roofs with vertical seams.7 A new product called Nailboard, manufactured by NRG Barriers of Portland, Maine, uses polyiso insulation with bonded sheets of 7Ⲑ16 or 5Ⲑ8 gauge wire board . The product provides a presentable roof finish and eliminates insulation damage during handling and installation. The phenolic insulation has the highest R-value of any foam sheet, approximately 8.3 per inch5, as well as excellent fire resistance. It is more expensive than other products and reserved for the most demanding applications. Phenolic insulation may require a backing plate for protection. Recently, phenolic foam insulation has been associated with accelerated corrosion of steel roofs and should be specified with caution. 8.3.3 Foam core sandwich panels Pre-insulated panels in Chap. 7 may contain rigid or fiberglass insulation. Rigid insulation foam is typically sprayed between metal surfaces and is allowed to "blow" and expand, filling any indentations (Fig 8.1c). Glass fiber is simply inserted between the plates. In each case, the panels offer excellent R-values, but obviously lack the acoustic performance of exposed fiberglass. A typical 2 in. Urethane or isocyanurate foam cores can have a U-factor of 0.06 and weigh as little as 21Ⲑ2 lb/ft2; a 6 inch thick plate can have a U-factor of about 0.02.8 fluids (Fig. 8.1d). Low cost and good noise absorption are the only advantages. *Thermax is a registered trademark of Celotex Corporation.

INSULATION INSULATION

259

this material. Its disadvantages are numerous and include a limited R-value, lack of vapor lag, and a rough appearance; it can accumulate dust, absorb moisture and oily residue, and accelerate the corrosion of metal surfaces. Pulp spray must conform to ASTM D 1042, Type II, Class (a)9 and not contain asbestos, crack or lose bonding with the substrate. The skills of the applicator are critical to a successful installation. Recently, spray-on-site polyicene foam insulation has been introduced. This flexible foam insulation shows promise and significantly outperforms cellulose. Sprayed foam can be trimmed after curing to give a relatively flat internal surface.8 8.3.5 Type and Thickness Selection The insulation selection process starts with determining the U-values required by the regulations and moves through in connection with the wall and ceiling design. (A list of U-values for several common wall constructions is given in Section 8.6.) In addition to meeting the minimum requirements, the thickness of the insulation must be great enough to prevent condensation on the siding and to maintain its temperature above the melting point. Dew. Acoustic performance, appearance and cost must also be considered. The fire hazard class determined by a flame propagation test can exclude some insulation options from the outset. As previously mentioned, the stability of polyisocyanurate insulation is highly dependent on its cover layers, and if the cover layers delaminate for any reason, it loses its dimensional stability. In contrast, the slightly more expensive extruded polystyrene retains its properties when wet. Fiberglass batt insulation with an appropriate coating generally offers the best overall performance and is specified almost exclusively for metal building roofs. The choice of wall insulation is closely related to the construction of the wall system and fire safety requirements.

8.4 VAPOR RESISTANT As mentioned above, the primary function of a vapor barrier is to slow the flow of moisture through a roof or wall mount. (Another function is to reduce airflow, as explained below.) No known material is a true vapor barrier that completely prevents moisture penetration, but some vapor barriers are better at slowing the process than others due to their lower permeability or permeability. Permeability is measured in perms in the British system (a perm is 1 grain of water transferred per hour per square foot for 1 inch of mercury vapor pressure difference) or in nanograms per pascal per second per square meter in the SI. (One grain is 1⁄7000 pounds.) A perm is 57.2 nanograms per pascal per second per square meter. The water vapor flux rate is equal to the difference in vapor pressure on the two sides of the material times its vapor permeability. The lower the permeability, the more effective the vapor barrier. Perm ratings of commonly used vapor retardants range from 1.0 for basic vinyl to 0.02 for premium composites such as sheet/fabric/kraft, metallized polypropylene/fabric/kraft, vinyl/fabric/foil, polypropylene/fabric/foil, and vinyl /fabric /metallized polyester. In order to select a suitable cladding material, you need to estimate the amount of moisture likely to be generated in the building and compare it with the annual temperature and humidity conditions in the area. Then factors such as appearance, abuse resistance and cost are considered. Some uses may place additional requirements on vapor retarders, e.g. B. Noise reduction or low emissivity. Anywhere where chemical fumes are released - chemical plants, laboratories, poultry farms, steel mills and the like - a special chemical analysis may be required. For critical applications it is recommended to consult a manufacturer. Industry leaders in vapor control layers include Lamtec Corporation of Flanders, New Jersey; Vytech Industries, Inc. of Anderson, South Carolina; Thermal Design, Inc. of Madison, Nebraska; and Rexam Performance Products, Inc. of Stamford, Connecticut. White materials are popular for their high light reflectance, a desirable quality for ceilings. Some abuse-resistant insulating coatings, such as those made by Lamtec Corp. marketed, indicate their strength; These products are intended to compete with interior wall paneling and plasterboard.

INSULATION 260

CHAPTER EIGHT

In the past, pure vinyl vapor barriers were used almost exclusively, but the use of vinyl is decreasing. Despite its low cost, vinyl has very high permeability, low puncture resistance, and a tendency to yellow and crack over time (Fig. 8.2). Each of the new materials can do the job much better. When price is the only consideration, non-metallized polypropylene/scrim/kraft fairings, often marketed as a replacement for vinyl, are among the cheapest white reinforced fairings on the market. However, the high perm ratio (0.09 or more) is 4.5 times higher than the metallic version, which is still inexpensive. The use of a quality vapor retarder improves the build quality of the building and facilitates future occupancy changes. It's not uncommon for a 'dry' building to change use to one that produces a lot of moisture - while the same mediocre vapor barrier stays in place. This practically guarantees serious condensation problems later on, as explained in the following section. It is worth noting that the difference between the cheapest and most expensive product available can be as little as two tenths of 1 percent of the total construction cost.10

8.5 HOW TO MAXIMIZE THERMAL OUTPUT 8.5.1 Avoiding Condensation Condensation and rusting problems in metal building systems are common. William A. Lotz, a well-known moisture and building insulation consultant: “In my experience, 'traditional' metal building coatings and insulation do not protect the metal building from condensation and rust for very long when there is moisture in a building in a cold location Climate.”11 However, even the best insulation and vapor barrier can be seriously compromised if installed incorrectly. Specifying a low permeability vapor retarder is just the beginning. According to Ref. 12 and others, the diffusion of moisture through vapor retarders is a slow process that rarely causes real problems. Some

FIGURE 8.2 Cracked and torn vinyl vapor retarder.

INSULATION INSULATION

261

A more practical hazard is the leakage of moisture-laden air into the insulated space, caused by poorly sealed seams of vapor retardants and unprotected penetrations. Unless all liner joints and penetrations are properly sealed, moisture will penetrate the insulation and eventually condense on the underside of metal surfaces, saturating the insulation and ruining its thermal performance. Acceptable gasket materials include caulks and tapes formulated for use with certain vapor retardants. Ordinary tape is not suitable for long-term vapor sealing.13 Interestingly, according to several studies, the type of building insulation does not have a significant impact on the amount of air infiltration. The level of experience and care individual workers have in sealing seams and penetrations is likely to have a greater impact on the amount of air leakage than the type of insulation used. Some of the newer insulation jacketing products offer adhesive or peel-off paper seams, which can provide a better seal than the traditional fold and staple method, as shown in Fig. 8.3. So a mediocre vapor retarder with good seams can outperform a top product with hollow edges. It should be noted that incorrectly placed clamps significantly reduce the effectiveness of vapor barriers. According to Lotz13, a single miss staple can increase the perm ratio of a foil-kraft laminate from 0.02 to 0.34! The permeability of interior vapor control layers is generally greater than that of metal roofing and cladding. If the humidity level inside the building is high, the moisture will eventually penetrate the vapor retarder and condense on the metal. (For this reason, items such as insulation holders should be made of non-ferrous materials or have a plastic coating to prevent rust stains.) At this stage, venting the insulated space between the metal and the vapor retarder is the only solution that can prevent moisture build-up and covert corrosion . It is therefore advantageous to provide the cold side of the assembly - in this case the outside - with a mechanism for removing moisture, e.g. B. Bottom view and ventilation openings (Fig. 8.4). Unfortunately, the roof assembly of ventilated metal structures is still rare. Ventilation does not guarantee the absence of condensation and consequent damage to surfaces, only reducing their severity. If extremely high levels of indoor humidity are expected, particularly in cold climates, it may make sense to use a non-metallic roofing system that is better at dissipating moisture. In the words of Tobiasson and Buska:14 "For example, a vapor-retardant vinyl/fiberglass insulation/metal roofing system with vertical seams, vented or not, is unlikely to be suitable for a building with an indoor pool in Minnesota." Also, venting is sometimes a double-edged sword , which invites moisture into the roof in the summer months. (Remember that to keep a home's basement dry, it's best not to ventilate and use a dehumidifier in the summer. Unfortunately, dehumidifying an insulated attic void space still doesn't seem practical.) The bouncing Point in specifying fire Dampening retarders in hot climates is a tricky business. According to O'Brien and Condren15, more traditional analysis methods may not provide adequate answers when used in southern climates, but the authors provide some guidelines. A particular problem arises with composite metal panels which are sealed at the factory for the stated purpose of preventing water ingress. Unfortunately, no stamp is perfect. Moisture can penetrate this vapor trap much more easily than evaporate, as the owners of many permanently fogged insulating glass doors and windows can attest. The problem is that glass does not corrode and the seal breakage is clearly visible in a window, with metal panels it is exactly the opposite. The daunting task of simultaneously providing ventilation openings in the panel and keeping out rain must go beyond simply drilling drainage holes to drain condensate. For some practical tips on preventing condensation in metal buildings, see the MBMA Condensation Fact Sheet16

8.5.2 Minimizing Heat Loss Through Fiberglass Insulation A well-known phenomenon of thermal bridging occurs when a piece of highly conductive material, such as a sheet of metal, is bonded to it. B. Metal, outside and inside connects and causes a "short circuit" in the insulation. In cold weather, this leads to thermal bridges, loss of performance and condensation problems. A good example of thermal bridging occurs when metal roofing or siding with fiberglass insulation is attached to secondary structures.

INSULATION 262

CHAPTER EIGHT

FIGURE 8.3 Traditional folding and tacking method of sealing two exposed edges (flaps) of adjacent fiberglass insulation flaps. (Butler Manufacturing Co.)

INSULATION INSULATION

263

FIGURE 8.4 Details of trapezoidal cantilevered roof in a floating ventilated ridge. (MBMA.)

The insulation is compressed to less than 3Ⲑ16 inches at the supports and reaches its full thickness near the center of the span, resembling an hourglass shape (Fig. 8.5). This "hourglass" method of installing insulation is the easiest, cheapest, and least energy efficient. This is most likely what you will receive unless you specify a different project in the contract documentation. If you got a bird's-eye view of a snow-covered roof, you would immediately see where the Tuesdays were: where the snow had melted.

INSULATION 264

CHAPTER EIGHT

FIGURE 8.5 System NAIMA 1 - Insulation installed on purlins. (Courtesy of the NAIMA Steel Construction Committee.)

How to estimate the heat loss through the framing? Some just ignore it and pretend it doesn't exist - hardly an option for an enlightened designer. Others perform sophisticated parallel flow analysis. Still others rely on the results of actual "hot box" testing. A simplified but viable option is to increase the delivered R values by 25 to 40 percent over those required for the analysis. Federal government standards17 require R-values reported on design drawings to be one-third (or based on local experience) greater than calculated R-values. These R-values are determined at an average temperature of 75°F according to ASTM C 518.18. The thicker the insulation, the more efficiency it loses in an hourglass installation, since heat loss through a purlin or belt remains nearly constant. .19 Therefore, while insulation with an R-value of 10 would lose 25 percent of its efficiency in this type of installation, one with an R-value of 19 would lose 42 percent.8 There are other possibilities, realistic U-factors to be determined for steel walls, building and roofing, taking into account the effects of compressed insulation on joists and purlins and the effects of thermal shorting on clamps and fasteners. According to Crall20-22, the results of a three-dimensional finite element analysis that accounts for these factors form the basis of the assembly U-factor tables, which can be found in Appendix A of ASHRAE 90.1. In addition to putting insulation in an uninsulated space in the first place, adding more insulation leads to diminishing returns, as the ASHRAE tables clearly show. In addition to heat losses, the hourglass design encourages condensation on roof purlins and wall joists in cold weather because the vapor check is on the outside of the structure. Ironically, when the humidity in a building is high, the more effective the vapor retarder, the more condensation can become - in some cases so bad that it can be mistaken for roof leaks. In an attempt to fit more insulation between the purlins, another specific system for roofs was developed. Here the insulation fills most of the space between the purlins, resulting in a slightly better installed R-value (Fig. 8.6). Lining flaps on the sides of the vapor check are overlapped on the purlins for continuity. Nevertheless, the problems with thermal bridges and condensation are not solved. This system requires insulating support straps running between the purlins, an additional cost item.

INSULATION INSULATION

265

FIGURE 8.6 System NAIMA 2 – insulation installed between purlins. (Courtesy of the NAIMA Steel Construction Committee.)

The third insulation support system attempts to solve the thermal bridging problem by relying on thermal blocks of polystyrene foam insulation and running on purlins. Blocks must end in ceiling clips but this design is clearly a huge improvement over previous ones. To further increase the R-value of the system, a second layer of non-insulated filler can be applied between the thermoblocks (Fig. 8.7). This design was developed with roofs with vertical seams in mind, as fastened roofs would require longer fasteners that would penetrate the blocks and potentially break them. In addition, the use of foam strips to support the ceiling slab can reduce the effectiveness of the attachment, as discussed in Chap. 5. In any case, the system still does not solve the problem of third-octave compression. The fourth installation system combines the advantages of non-compacted bulk insulation and thermoblock (Fig. 8.8). Its installed R-value approaches that of the insulation alone, a major achievement. On the other hand, it is quite troublesome to install, still requires insulating support tapes, and still does not solve the condensation problem. The last system includes a new element: an insulated ceiling panel with an integrated vapour-retardant coating. The insulation panel not only provides support for fiberglass mats, but also finally solves the problem of condensation when laying purlins within the insulated area. This is a premium system, both in terms of performance and price. For even better performance, thermal blocks can be used on top resulting in installed R-values in excess of 34 at 1.5 inches. thickness can be used (Fig. 8.9). Which system to choose? The choice depends on the use of the building, the climate and the budget. Architects should familiarize themselves with new products and proprietary technologies that are coming to market. For example, Thermal Design, Inc.'s proprietary "Simple Saver System" is said to have improved on traditional technology by using a heavy-duty permanent vapor check under the purlins, eliminating the problem of purlin condensation. The lining is supported by

INSULATION 266

CHAPTER EIGHT

FIGURE 8.7 NAIMA 3 system - using thermal blocks. (Courtesy of the NAIMA Steel Construction Committee.)

FIGURE 8.8 System NAIMA 4 - Use of thermoblocks in combination with ceilings between purlins. (Courtesy of the NAIMA Steel Construction Committee.)

INSULATION INSULATION

267

FIGURE 8.9 A premium insulated ceiling tile system. (Courtesy of the NAIMA Steel Construction Committee.)

a series of strips attached to the underside of the frame. Similar to the fourth system, insulation is installed above the ceiling. Another proprietary product is Butler Manufacturing Co.'s "Sky-Web System," an open-mesh polyester screen with approximately 1Ⲑ2-inch openings designed to optically blend with white vapor retardants. In addition to providing insulating support, the net is designed to protect workers from falls and catch falling objects. Another proprietary insulation system, Finished R, by Owens Corning, consists of a 1-inch fiberglass blanket with a laminated vapor retarder supported by a grid of plastic extrusions suspended from the building's metal frame. Any insulation system that relies on materials trapped under or between purlins should be checked for deterioration of the purlin reinforcement described in Chap. 5.

8.6 UO VALUES OF DIFFERENT WALL SYSTEMS Determining the Uo values of the general conductance for roof attachments is relatively easy considering that metal roofing and fiberglass batt insulation are used almost exclusively and ASHRAE ratings are available. Our previous discussions have covered several fiber optic installation methods that make a difference in thermal performance more than anything else. The walls, however, are another matter. A variety of metallic and "rigid" materials (masonry or concrete) are available, as well as fiberglass or rigid insulation. Tables 8.1 and 8.2 have been included to facilitate comparison between the general guide values of the most common wall systems. The R values for the tables were taken from the ASHRAE Handbook of Fundamentals23 and Ref. 24. To simplify the comparison, the following values of R were assumed for all plants: outer air film at 15 km/h wind ⫽ 0.17; indoor air film ⫽ 0.68; an air gap 3Ⲑ4 to 4 in depth ⫽ 0.97. In order to calculate the total conductance Uo it is necessary to sum all the R values of the components and obtain an inverse function of the sum as shown in the example below. R-values for selected materials are given in Table 8.1. Uo is the number of Btu that will flow through 1 ft2 of wall in 1 hour when the temperature difference between the two sides is 1°F. Example Calculate the Uo value for brick veneer on CMU insulated assembly: part

bravery r

4" brick veneer exterior air foil 2" air gap 2" polystyrene insulation

0,17 0,44 0,97 10,8

INSULATION 268

CHAPTER EIGHT

Component (continued) 8-in. Air gap 3Ⲑ4 inch. Drywall 1Ⲑ2-in. Room air film R total ⫽

Bravery R 1.51 0.97 0.45 0.68 15.99

So Uo ⫽ 1/R ⫽ 0.0625. Table 8.2 contains the results of similar calculations for common wall systems, roughly in the range described in Chap. 7. Manufacturer hot box test data is used for some assemblies. Thermal efficiency losses from metal studs are accounted for by lowering the R-values of fiberglass insulation by a third. For those looking for an analysis method to analyze parallel flows through metal bands or pins and through completely isolated areas, the following procedure is offered: in clay bricks

0,44

4" Block (72% Solid) 6" Block (59% Solid) 8" Block (54% Solid) 10" Block (52% Solid)

1,19 1,34 1,51 1,61

12 inch block (48% solids) concrete, regular weight, in 5.5 6 8 10 12

1,72 1,30 1,33 1,49 1,64 1,82

Drywall 1Ⲑ2-in.

0,45 0,17 0,68

Outside Air Film (Winter) Inside Air Film Dead Air Gap (3Ⲑ4 to 4 inches) (Winter)

0,97 2,80

Air Gap on Board Surface Thickness of Insulation Type, in

1Ⲑ2

3Ⲑ4

11Ⲑ2

Polyisocyanurate (board surface) Extruded polystyrene

4,0 —

5,8 4,05

7,7 5,4

11,5 8,1

15,4 10,8

23,1 16,2

Fiberglass mat insulation (approx.), in 3 3.5

10 11

4 6

13 19

Source: Compiled from Refs. 19 and 23.

INSULATION INSULATION

269

TABLE 8.2 Uo values for selected wall mounts Wall mount

Illustration

3" fiberglass insulated steel side*

figs 7.1 and 7.7

0,13

4" fiberglass insulated steel side*

figs 7.1 and 7.7

0,12

Concealed zipper, 3" fiberglass ceiling, 1Ⲑ2"† metal and plasterboard ceiling

Abb. 7.4

0,112

Hidden mounting plate, 2" Thermax plate, wood paneling and

Coward. 7.4 (similar)

0,047

1Ⲑ2-inch plasterboard†

Factory insulated panel with 2-in. the thick‡

Abb. 7.6

0,069

8 inch non-isolated CMU

Abb. 7.21

0,424

8 inches. ⫹ 3Ⲑ4 inches. Furring ⫹ 1-in. of polystyrene ⫹ 1Ⲑ2-in. of bonded plasterboard 4-in. of bricks ⫹ 2-in. of airspace ⫹ 8 inches. from CMU ⫹ 1Ⲑ2 inches. Drywall over veneer

0,109 Abb. 7.28

4 in. brick ⫹ 2 in. air gap ⫹ 2 in. polystyrene insulation ⫹ 8 in. CMU ⫹ 1Ⲑ2 in. drywall (see example)

0,193 0,0625

4" brick ⫹ 2" air gap ⫹ 1Ⲑ2" cladding ⫹ 31Ⲑ2" fiberglass cladding on 35Ⲑ8" steel stud ⫹ 1Ⲑ2" drywall

Abb. 7.29

5.5 inches of concrete

Abb. 7.36

0,10 0,465

5.5 in. concrete ⫹ glued 1-in. Polystyrene Insulation ⫹ 1Ⲑ2 in. glued plasterboard

0,125

8 poles made of concrete ⫹ 3Ⲑ4 pol. Veneer ⫹ 1 pol. made of polystyrene ⫹ 1Ⲑ2 pol. cast plasterboard

0,069

ETICS (2 in styrofoam ⫹ 1Ⲑ2 housing, other materials neglected)

Abb. 7.37

(Video) Your SPIRITUAL AWAKENING Survival Kit – 35 Wisdom Practices For Grounding, Enlightening & Embodying

Insulating glass with 1Ⲑ4 to 1Ⲑ2 inch air gap

0,082 0,57

*As reported by Star Manufacturing Co. for Durarib or Starmark wall, plate for strap attachments a 12 in o.c., strap spacing 5 ft 0 in o.c. †As used by Star Manufacturing Co. for Star CFW panel, 24 in. ‡As used by Star Manufacturing Co. for STARTTHERM II panels. §Insulation reduced by a third to account for parallel flows.

Where from

S ⫽ Percentage of area occupied by Usteel steel structure ⫽ U-value for area occupied by Uinsul steel structure ⫽ U-value for isolated area between structures

For example, the S-factor for 16 inch center-to-center steel bolts is about 20 percent; with 24 pins on mids it is about 15 percent.

REFERENCES 1. ASHRAE Standard 90.1, Energy-Efficient Design of New Buildings Except Low-Rise Residential Buildings, ASHRAE, Atlanta, GA, 1999. 2. Metal Building Systems, 2nd Edition, Building Systems Institute, Inc., 1230 Keith Building , Cleveland, OH, 1990. 3. Understanding Insulation for Steel Buildings, NAIMA, Alexandria, VA, 1994. 4. ASTM E 553-92, Mineral Fiber Ceiling Insulation for Commercial and Industrial Applications, Philadelphia, PA, 1992. 5. Thomas Scharfe, "Specifying the Right Insulation for Roofs", Building Design and Construction, September 1988. 6. "Polyiso Insulation Outperforms Polystyrene in Tests", Metal Architecture, April 1995.

INSULATION 270

CHAPTER EIGHT

7. „Board Insulation Designed for Metal Buildings“, Metal Architecture, April 1994. 8. Bob Fittro, „More Than One to Insulate a Metal Building“, Metal Architecture, Dezember 1995. 9. ASTM E 1042-92, Akustisch absorbierende Materialien Angewandt durch Kelle oder Spray, Philadelphia, PA, 1992. 10. Bob Fittro, „Manufacturers Stress Importance of Proper Vapor Barrier Selection“, Metal Architecture, September 1995. 11. William A. Lotz, „Simple Rules Sidestep Insulation Failures“, Building Design and Construction, April 1996. 12. Wayne Tobiasson, „General Roofing Considerations“, ASTM Manual 18, Humidity Control in Buildings. Auch erhältlich als CRREL Misc. Dokument 3443. 13. William A. Lotz, „Specifying Vapor Barriers“, Building Design and Construction, November 1998. 14. Wayne Tobiasson und James Buska, „Stehfalz-Metalldachsysteme in kalten Regionen“, Proceedings of the 10th Conference on Technology of Roofs, 1993 (von CRREL Reprint). 15. Sean M. O'Brien und Stephen J. Condren, „Design Guidelines for Vapor Retarders in Warm Climates“, The Construction Specificationr, Dez. 2002, S. 42–51. 16. „MBMA Condensation Information Sheet“, Metal Building Manufacturers Association, Cleveland, OH, undatiert. 17. Abschnitt 13120A, Standard Metal Building Systems, Unified Facilities Guide Specifications, Huntsville, AL, Januar 2002. 18. ASTM C 518, Steady-State Heat Flux Measurements and Thermal Transmission Properties by Means of the Heat Flow Meter Apparatus, ASTM, Philadelphia , PA, 1998. 19. Facts about Metal Building Insulation and System Performance, Booklet Information from NAIMA, Alexandria, VA, 1994. 20. Chris P. Crall, „Metal Buildings: Specificationing to Meet Energy Standards“, The Construction Specificationr, Juli 2001, S. 55–58. 21. Chris P. Crall, „Die Wärmeleistung ist der Maßstab für den Erfolg bei den meisten Isolieranwendungen“, Metal Architecture, Juli 1998, S. 18–22. 22. Chris P. Crall, „Revised ASHRAE 90.1 Standard Addresses Thermal Performance of Metal Construction Envelopes“, Metal Construction News, Februar 2000, S. 58–61. 23. ASHRAE Handbook of Fundamentals, American Society of Heating, Refrigerating and Air Conditioning Engineers, Atlanta, GA, 1989. 24. Cavity Walls, Masonry Advisory Council, Park Ridge, IL, 1992. 25. John H. Callender (Hrsg.) , Time-Saving Patterns for Architectural Design Data, 6. Aufl., p. 4–49, McGraw Hill, New York, 1982.

REVIEW QUESTIONS 1. What type of insulation is most commonly specified for metal building systems? 2 Which standard sets requirements for energy efficiency in buildings made of metal? 3 What is the function of a vapor retarder? What are some of the challenges of working with vapor retarders in hot climates? 4 What advice would you give to an owner who wants to buy a building with a previously 'dry' use but wants to introduce a process that releases a lot of moisture? 5 Is it better to use a vapor control layer with the lowest Perm rating but installed by an inexperienced contractor, or a mediocre retarder installed by an experienced installer? Why? 6 What are the advantages of the vapor barrier below the purlins? 7 What are the advantages and disadvantages of ventilating a roof cavity in a warm, humid climate?

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 9

THE PROCESS OF PURCHASING A METALLIC BUILDING

9.1 GETTING STARTED 9.1.1 Before You Begin: A Note to Owners In this chapter we have talked about the many things that must happen between your decision to build and the handing over of the keys to a newly built facility. Once the decision to build or expand has been made, you need to decide on the dimensions, shape and headroom of the building. Just about anything goes, if these basics are not well thought out, the project will not be successful. If an important piece of machinery can only fit in by a couple of inches, what was the point of building it? We strongly recommend that you have the programming and pre-planning of your building carried out by an experienced architect: architects are trained to analyze the needs of the owner and to offer solutions. Many years of living under one roof with architects have convinced the author of the tremendous improvements these design professionals can make to their clients' original plans. An architect often comes up with a completely different and better building plan. Unless you need a small basic rectangle of a building or need to adjust a predetermined equipment layout, let designers, not contractors, help you with design decisions. (Of course, the architect you choose should have experience in specifying metal building systems, or at least have read this book....) The architect will help you determine your immediate and future space needs, provide a preliminary estimate, and propose a timeline for construction . For your part, you need to determine if there is adequate funding, an adequate or planned budget, and if your internal planning team members agree on what needs to be done.

9.1.2 Site selection After the programming phase, the project moves on to the schematic design. At this point, a potential site may already have been selected. If multiple locations are still being considered, it is best to focus on one or two options before proceeding as many building parameters such as height, size and building type can be affected by surrounding buildings and local zoning regulations. As is usual in such transactions, a prospective buyer will complete a title and easement research, zoning review, site survey, and environmental survey prior to purchase. The site must be large enough to allow for all necessary backlots, parking, driveways, and future expansion needs. If time is of the essence, it's best to stay away from sheltered areas like swamps. Ideally, the location already has all the necessary supply connections or can be supplied economically. Drainage requirements can be tightly controlled by the community and must be specifically investigated and any site drainage issues should be addressed.

271 Conveyed from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

THE PROCESS OF PURCHASING A METALLIC BUILDING 272

CHAPTER NINE

Now it's time to hire a reputable local engineer and soil drilling contractor to conduct a soil survey. Is the soil good enough to allow for economic shallow foundations? The question is not in vain, as in many industrial areas the best lots have long since been developed, leaving only the least desirable lots - often those that others consider uneconomic to build on. Poor soils can require deep and expensive foundations such as piles and caissons; the additional costs can push the budget beyond acceptable limits. Some other site preparation costs include demolition of all existing structures, site clearance, excavation, backfilling and paving.

9.2 THE ROLE OF THE DESIGN PROFESSIONAL 9.2.1 The basic responsibilities Once the schematic design is completed, either by the owner's internal staff or an external architect, and the location is chosen, the owner can start to think about delivery methods for the project. Building. With a set of schematic plans and specifications in hand, an owner can pursue one of three basic building methods: conventional (design and bid or negotiate), design-build, or contract directly with a preselected building systems manufacturer. In both conventional and prefabricated construction, the owner can be represented by an independent architect; In design-build mode, the architect is part of the client's team. Because we are specifically interested in metallic building systems and the fabricator's team rarely includes architects, it is best for the homeowner to hire an independent design team. One of the design team's first priorities is to develop a site plan package for review and approval by a local planning and zoning authority. The package shows how the owner intends to comply with federal, state, and local regulations. It can address issues such as wetland protection, increased traffic, pollution, sewage flow, parking, and appearance. While some locations are conducive to development, others may not be; Occasionally, obtaining all permits may take longer than the combined design and construction time. To prepare the site package, the design team conducts an extensive code review. (Needless to say, intimate knowledge of the law's complex requirements is reason enough to even hire an architect!) By filing a set of documents that comply with local codes and all local codes, the homeowner can save a great deal of valuable time and lower mortgage interest rates. Design development and final design can continue while site package is reviewed. The goal is to create a set of contractual documents that adequately communicate the intent of the project without being overly specific and prescriptive. In general, the design professional is responsible for selecting the design criteria, for all tasks not normally performed by the fabricator, and for overall coordination. Items not normally available from manufacturers are listed in the MBMA's Common Industry Practices and include foundations, insulation, fire protection, siding, cranes, electrical and mechanical equipment, overhead doors and miscellaneous iron. The practices state explicitly that ventilation, condensation and energy saving issues are outside the responsibility of the manufacturer and should therefore be included within the scope of the design professional. The design team should study the impact of the proposed building on adjacent structures, e.g. B. the possibility of snow falling on a lower existing roof. The manufacturer should not be expected to take on this purely technical task, as some smaller manufacturers may not even have a full-time engineer - just a technician entering the numbers into a computer program. (Most homeowners are unaware of this fact, as the term prefabricated building implies the presence of an engineer.) Homeowners, in turn, should help planning professionals determine the appropriate project design criteria by providing you with adequate data that describes the details of the current problems and

THE PROCESS OF PURCHASING A METALLIC BUILDING THE PROCESS OF PURCHASING A METALLIC BUILDING

273

sustainable operation. This data could include, for example, the dimensions and weight of the equipment to be housed in a metal construction system and any crane requirements. Large industrial and government customers should also provide copies of their internal design and construction standards and other relevant design materials.

9.2.2 What to specify Some people still think that the architect's role in specifying metallic building systems is to choose the colors of the coatings. It's not like that. Although the construction documentation prepared for a prefabricated building may not be as extensive and detailed as for conventional designs, it still needs to convey a great deal of information. Some of the points that manufacturers are looking for when preparing quotations are:1 ●

●

● ●

●

Current building code information, including the edition. Avoid listing too many codes that may contain conflicting design criteria. While reputable manufacturers use the most conservative criteria in such cases of conflict, some hungry newcomers may decide to take a different approach. Design loads to be used such as collateral, snow, tension, wind and earthquake. Some recurring problems when specifying snow and roof traffic loads are discussed in Chap. 10. The tolerance of a persistent (overlapping) collateral load should be carefully assessed and its nature preferably identified. Rooftop HVAC equipment must be located on the roof plan, specifying its weight and required openings. All other concentrated loads, e.g. B. from a catwalk, deserve a separate mention. It is important to research local regulations that may have higher design load requirements than model regulations. For example, the forecast wind speed may be specified by a local code as 110 miles per hour, while a national code only requires 70 miles per hour. (Of course, the opposite could also be true: local code could be based on an outdated edit of the template code.) Load combinations. In addition to the combinations listed in the regulatory code, designers may wish to include some others, as discussed in chap. 3. The static scheme assumed in the design (e.g. rigid multi-span frame with pin supports). Building dimensions, including length, width (remember that building width for a manufacturer is the distance between the outside flanges of the wall studs, not between the center lines of the studs), eaves height and headroom. Exterior wall materials, finishes and insulation. Some designers choose to omit doors and windows from a metal construction package: by purchasing these items locally it is often possible to purchase sturdier products with better hardware and avoid damage in transit. In this case, however, the design wind pressures to be used for these important components of the building envelope must be communicated to their suppliers. Locations where wall bracing should be avoided for aesthetic or functional reasons and where bracing may be desirable. Also all open wall positions. anti-corrosion requirements. Designers are advised to note the presence of existing facilities within a 1Ⲑ2 mile radius that emit corrosive chemicals, proximity to salt water areas, and any other potential sources of corrosion. You should also assess any corrosive or moisture generating potential of the operation within the building itself. In metal buildings it is difficult to protect against corrosion from the inside. While exterior cladding is very good at fighting corrosion, the inner steel structure is usually protected only by a primer (Fig 9.1). Many manufacturers lack facilities for quality surface preparation and application of premium coatings; They ship the steel to specialty shops when those coatings are mandated, which drives up the cost. For the main structure, it is preferable to use a high quality, on-site applied paint rather than to specify an electroplated finish: hot-dip galvanizing tends to promote warping and warping of structural members made from thin engineered panels. Some manufacturers offer galvanized C or Z purlins and straps.

THE PROCESS OF PURCHASING A METALLIC BUILDING 274

CHAPTER NINE

FIGURE 9.1 Shop primer is applied to a segment of the primary structure.

●

Any restrictions on frame sizes. The drawings should indicate the maximum column depth that the foundations can accommodate. The taper ratio of the columns must also be controlled when equipment or internal walls are near the columns. Lateral drift and vertical deflection criteria for both primary and secondary framing. (This topic is critical enough to merit a chapter of its own; see Chapter 11.) Crane requirements, if any, including service levels, as explained in more detail in Chapter. fifteen.

The author's practice for most projects, including those using metalwork systems, is to provide a drawing with general structural annotations. Annotations summarize design loads, material specifications, concrete strength, etc. A sample section dealing with prefabricated buildings is shown in Fig. 9.2. Project loads, including incidental loads, would be reported in a separate notes section (usually Section I). A set of typical details showing areas where specific performance is required (rather than the manufacturer's standard details) may also be included. This may include purlin reinforcement details, anchor bolt construction, support plates on sloping sheaves, frames around overhead doors and other features deemed necessary.

9.3 MANUFACTURER'S RESPONSIBILITIES It is the manufacturer's responsibility to design and manufacture the metal structure, excluding the above parts, to the bottom of the column base plates. Using the ones provided by the owner

THE PROCESS OF PURCHASING A METALLIC BUILDING THE PROCESS OF PURCHASING A METALLIC BUILDING

FIGURE 9.2 Example set of notes for specification of metal construction systems.

275

THE PROCESS OF PURCHASING A METALLIC BUILDING 276

CHAPTER NINE

By design criteria, the manufacturer selects a "ready-made" frame from a standard product catalog or custom designs it. While the industry started with the first approach (hence buildings are called “prefabricated”), the latter is now the norm. Computers have revolutionized the design of metal building systems, blurring the line between "standard" and "custom" options. Today, the vast majority of metal building systems are custom designed for a specific project. Many major manufacturers have developed extensive CAD libraries of details and connections that can help quickly put together a computer aided frame design. While each company tends to develop their own software with slightly different features, all programs perform similar functions. Typically, computers assist in the creation of anchor bolt plans and details, frame views, structural calculations for each member, and cost data. The latest graphics software can create impressive looking documents useful for presentations to rental companies and licensing agencies. Large fabricators see their advanced software capabilities as technical and marketing tools that differentiate them from less-equipped competitors whose staff may not even include registered engineers.

9.4 THE ROLE OF THE MANUFACTURER Normally, the manufacturer does not enter into direct contracts with the owner. The company that does this is a local chartered construction company. Builders can act as the general contractor for the project with full responsibility for the project or just as suppliers of the prefabricated building. In either case, they can outsource the construction of the building to another company. Large manufacturers are very selective about the types of people they allow as builders and look for contractors who are financially stable, experienced and committed to quality workmanship. A prospective builder is typically required to complete a manufacturer-sponsored course and obtain a renewable certificate. The builder doesn't just take a bunch of owner contract documents and mail them to the fabricator; Many manufacturers don't want to go through a lot of plans and specifications to make a suggestion. Instead, the client interprets the documents and distils them into so-called order documents, a standard bid form accompanied by graph paper sketches and other supporting data (Fig. 9.3). Architectural contract documents are only referenced in an agreement between builder and owner; a contract between the client and the manufacturer results exclusively from the order documents. Given that builders often lack in-house engineering expertise, the potential for misinterpreting some complex design requirements is very real. In fact, the task of condensing hundreds of pages of information into a simple form can be overwhelming even for the specifiers themselves. It's easy to see how some design details get lost in translation and never make it to the manufacturer. A close examination of the manufacturer's design certification letters (Section 9.5) and workshop drawings is extremely important to reassure the owner that the building will be constructed as planned. Some large manufacturers have established "National Account" departments, reflecting a desire to provide personalized service and better communication with large interstate regular customers who would otherwise deal with a multitude of local contractors.

9.5 BIDDING AND SELECTION Except in cases of captive relationships, the project will likely be bidding or negotiated. Several manufacturers present their proposals through their distributors. Which one to choose? Of course, the cheapest proposal can have the best chance of being accepted if it agrees with the others on all points. It is not easy to make this statement. Sometimes a contractor does not even mention which manufacturer will supply the building, or whether the metal roof and wall panels will be fabricated in the workshop or on site. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

THE PROCESS OF PURCHASING A METALLIC BUILDING THE PROCESS OF PURCHASING A METALLIC BUILDING

FIGURE 9.3 Sample contract proposal. (Association of Metal Builders and Contractors, formerly Association of Systems Builders.)

277

THE PROCESS OF PURCHASING A METALLIC BUILDING 278

CHAPTER NINE

FIGURE 9.3 (continued)

THE PROCESS OF PURCHASING A METALLIC BUILDING THE PROCESS OF PURCHASING A METALLIC BUILDING

279

One way to ensure a level playing field is to require all manufacturers to submit design certification letters with their proposals. Such a letter must clearly state the name of the client, the configuration of the building, the codes and standards to be complied with, and any loads and combinations of loads for which the building is to be designed. The letter must bear the seal and signature of an engineer employed by the manufacturer. Another critical point to review in the letter is the snow or payload construction of the roof, a common target of manipulation by some manufacturers seeking an edge over the competition. (More on this in the next chapter.) Insist on seeing the actual roof load, not the "ground snow load" which is just a starting point for further calculations. If the payload of the roof rather than snow drives the design, the chart should reflect the actual load used to construct various elements and any reductions in payload made by the manufacturer. In addition, the seismic and wind loads of the project, and particularly the collateral loads, should be clearly indicated on the map. Verify that the correct criteria for lateral deflection and vertical deflection are being used: a manufacturer who "ignores" these criteria has a major cost advantage over others. If the letter provides too little or too vague information, do not hesitate to write to ask for clarification if anything is unclear. An example of a design certification letter is shown in Fig. 9.4 of Ref. 2. Note that the letter gives the specific release dates of the relevant standards and codes, a fact that is particularly important for the AISI specification, as explained in chap. 5. This and other points often lacking in clarity in design certification letters are highlighted in bold. To ensure the design and manufacturing quality of a critical project, the owner can choose to only work with manufacturers whose facilities have passed the AISC quality certification program, category MB. Although there are many high-performance manufacturers who are not certified, the restriction of the group of bidders to the providers marked with the designation simplifies the comparison considerably. Compare guarantees carefully. Is the guarantee required in the contract documents provided? The standard warranty on frame components is 1 year from the date of shipment. Depending on the material, metal roofs are guaranteed for up to 20 years. Also check the qualifications of the builder. Since the builder concludes the contract with the owner, this check is at least as important as the manufacturer comparison. Have the builders worked on similar projects? Do your references validate your ability to deliver on time and on budget? Are you satisfied with the quality of your work? Are they financially stable and committed? (Who hasn't heard of contractors going bankrupt mid-job?) Are they members of the Metal Building Contractors & Erectors Association? Are they certified by the manufacturer? Examine how the builder tends to deal with elements that are out of scope. Are the "extras" cheap or do they become an asset? Find out if the builder (or the president of a large construction company) is personally involved in the projects: the best builders are. For example, the President of Span Construction and Engineering, Metal Construction News' "1994 Top Metal Builder," personally inspects all large metal roofs completed by the company and invests up to 1Ⲑ2 days in each of these "walks." In his words, this test lends credibility to the 20-year weather guarantee.3 The selection process ends with the signing of the contract documents by the owner and builder. The contract documents may include the contract, general and additional conditions, drawings and specifications. The contract should clearly assign responsibility for various aspects of design, manufacture, assembly, compliance, and approval. Except for very small and simple buildings, we recommend using EIA contract forms instead of a one-page offer and contract form similar to Fig. 9.3. (Incidentally, the MBCEA contract stipulates that owners will be fined 25% of the contract price if they do not proceed with work after the contract is signed.) The contract may reference MBMA Industry Common Practices to establish a scope of work. Some contract clauses can lead to lengthy negotiations. If discussions are at an impasse on a really important provision, treatment of the lowest bidder can be waived where permitted by law, and not. 2 bidders invited to the negotiation. In this tense situation, some homeowners may rely on verbal promises rather than written agreements in their eagerness to build, forgetting Samuel Goldwyn's joke about a verbal contract not being worth the paper it's written on. Manufacturers may be more open to negotiation if contacted during the slowest months of the year - November and December.

THE PROCESS OF PURCHASING A METALLIC BUILDING 280

CHAPTER NINE

FIGURE 9.4 Sample attestation letter. (Butler Manufacturing Co.)

THE PROCESS OF PURCHASING A METALLIC BUILDING THE PROCESS OF PURCHASING A METALLIC BUILDING

281

9.6 BEARING AND BUILDING DESIGNS One of the first pieces of information that the manufacturer must present to the owner's designers is the value of the column reactions. As already mentioned and in Chap. 12 The foundations of metallic structures are often designed by the manufacturer prior to receiving this data and then rechecked against the actual responses. This process is on the critical path and this submission is needed as early as possible, preferably in the bidding phase. In any case, the manufacturer should not need more than 2 weeks to process the order plus 1 week to generate the responses. Thus, the reaction report as shown in Fig. 9.5 and perhaps even a full approval set could be ready in 3 weeks. The permit set can include an assembly plan, frame views, anchor bolt plans, wall views and some details. (What the set includes should have been stated in the contract documents and hopefully in the job documents prepared by the builder.) An example of a gantry lift is in Fig 9.6; You can find an anchor plan in Chap. 12. The permit set may not be drawn to scale, but is still a source of valuable information; it should be carefully examined for any indication of a misunderstanding of design intent. For example, on one project, the first item submitted for review by the architect engineer was a plan for anchor bolts. The cartoon was accompanied by a broadcast that read "RUSH!!". However, instead of quickly stamping the plan 'Approved', the reviewing engineer took some time to think about it and noticed that each column had 8 anchor bolts. The engineer surmised that the manufacturer intended to use fixed foot columns instead of pins as advertised. The suspicion was investigated and proved to be correct. By this time the foundations were in place and a serious problem was avoided. The submission must also include detailed static calculations sealed by the manufacturer's engineer. Some landlords insist that the engineer be registered in the state where the building is located and include this requirement in the contract documentation. Standard industry practice for the MBMA only requires the engineer to be registered in the manufacturer's home state. While reputable manufacturers tend to present calculations with assumptions and clearly labeled input data, others may try to overwhelm verifiers with reams of incomprehensible computer data. These inputs can, in Tom Clancy's words, appear written by computers to be read by calculators. If you find anything suspicious, ask written questions and insist on strict adherence to project requirements. (In at least one design, a contractor has indicated that the building meets the design's stringent lateral deviation criteria. Calculations indicate otherwise.) Don't be surprised to find that all comments identified in the permit record are interpreted as changes and be received with manufacturer changes order. Some complex three-way negotiations may occur. The reader is invited to review Secs. 2.2 and 3.3.3 of the MBMA Common Industry Practices dealing with such changes. In some unfortunate circumstances, the project can stall right there and get into a fight. The approval set, as schematic as it may be, is the first and usually last opportunity to review the manufacturer's manufacturing drawings. Subsequent detailed work represents production drawings, which are generally not made available by the manufacturer, unless this is expressly stipulated in the contract. After all the shop design issues are solved, the construction can finally begin. Some "warning signs" that you should pay attention to during the construction phase are given in Chap. 16

REFERENZEN 1. Alexander Newman, „Engineering Pre-engineered Buildings“, Civil Engineering, September 1992. 2. Duane Miller und David Evers, „Loads and Codes“, The Construction Specificationr, November 1992. 3. Shawn Zuver, „Span Construction & Engineering gewinnt die fünfte Top-Builder-Ehrung“, Metal Construction News, Mai 1995.

THE PROCESS OF PURCHASING A METALLIC BUILDING 282

CHAPTER NINE

FIGURE 9.5 Response Report. (VP building.)

FIGURE 9.6 Factory drawing of a manufacturer. (VP building.)

THE PROCESS OF PURCHASING A METALLIC BUILDING

283 Conveyed from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

THE PROCESS OF PURCHASING A METALLIC BUILDING 284

CHAPTER NINE

VERIFICATION QUESTIONS 1. What document must the manufacturer provide to demonstrate compliance with the contractual documents? 2 Name at least five points that must be included in the contract documents. 3 Explain how the manufacturer typically learns about specific contract requirements. 4 What could be some of the problems that await the owner who decides not to hire an outside architect-engineer for the project? 5 What third-party certification program is available to fabricators? 6 Is building insulation normally part of the manufacturer's scope of services?

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 10

SOME COMMON PROBLEMS AND ERRORS

Time and time again, those who specify metalwork systems are faced with the same pesky problems that cause more than their share of problems. These issues deserve special attention from planners. Most problems are rooted in misunderstandings or miscommunications between the owner and the owner's design team on the one hand, and the manufacturer and builder on the other. Each of the vignettes in this chapter was inspired by a not-so-pleasant real-life event. A brief overview of flaws in metal construction - the ultimate problems - rounds out our discussion.

10.1 SPECIFICATION OF BUILDINGS WITH COMPLICATED SHAPES AND WALL MATERIALS 10.1.1 Very Small Buildings Metal building systems are sometimes specified for inappropriate applications where their full benefits cannot be realised. The systems are best suited for large, low-rise rectangular buildings, particularly those that can benefit from metal plate walls and roofs (see Figure 1.3 in Chapter 1). Yet time and time again, prefabricated structures are provided for small buildings with irregular floorplans, complex roof shapes, and disparate wall materials—with mixed results at best. When such conditions apply, a rule of thumb places a minimum footprint of buildings suitable for prefabricated construction at about 3,000 square feet. Although some smaller buildings were successfully built, they could be compared with some of the others in chap. 3. Certainly there are manufacturers who specialize in producing simple rectangular freestanding metal buildings at very reasonable prices. However, our focus is on customer-specific constructions, perhaps with architectural features. Manufacturers often find that small buildings with complex layouts require careful construction and no less effort than large simple boxes. This engineering and detailing time as well as the mobilization and transport costs are difficult to recoup in small structures (Fig. 10.1). For all these reasons, major manufacturers rarely pursue small buildings in distant areas, leaving such buildings to their smaller competitors who, unfortunately for homeowners, lack experience with custom applications of metal building systems. A common result is a series of design and coordination problems faced by the owner and builder, problems that a more experienced fabricator could easily have solved. 10.1.2 Complex Configuration Large buildings with complex shapes can also pose a problem. Many manufacturers' construction programs are designed for rectangular structures. For example, a C-shaped building can be destroyed. All rights reserved. All use is subject to the terms of use provided on the website.

SOME COMMON PROBLEMS AND ERRORS 286

CHAPTER TEN

FIGURE 10.1 Although having a metal roof, this small but complex building with masonry walls would not be an ideal application for metal building systems.

divided into three rectangular "units" by the manufacturer, each of which could assign the structural design to a different engineer or technician. Not surprisingly, each unit with its own set of columns can end up with a double row of columns at the unit interface.1 Unless the owner's design team anticipated this turn by providing expansion joints at the interface—with a double set of columns and foundations - it's a shock when the workshop drawings arrive. The unexpected double set of columns can have devastating effects on the proposed column sizes and require new expansion joints in the exterior walls and roof, complicating the building's appearance. It will also require much larger foundations than shown in the contract drawings. In this construction phase, the foundation order may already have been awarded or, in the horror scenario, concreted. In either case, the switch will be neither easy nor cheap. We recommend that builders divide L, C and Z shaped buildings into rectangles in advance or warn in the contract drawings not to introduce columns not shown there. The fabricator can avoid additional supports by using transfer beams, a solution that complicates the frame construction somewhat, but is certainly within the capabilities of most metal fabricators. Metal roofing with ridges and valleys should be avoided in metal building systems, especially when using structural roofs with vertical seams. As in chap. 6, these roof configurations present conceptual difficulties and impede expansion and contraction of the metal roof. We add here that the details of the purlin supports on inclined hip and collar beams are quite complicated (Fig. 10.2) and the stability of the purlin on supports is difficult to ensure. The complexity of the design increases the likelihood of leaks and structural problems in these critical areas. 10.1.3 Hard wall materials according to Chap. As demonstrated, metal building systems are increasingly being constructed with masonry, concrete and other hard wall materials. These applications extend the acceptance limits of systems and

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

FIGURE 10.2 Details of load-bearing purlins at ridges and valleys: (a) cut up the ridge; (b) Section in valley looking down. (A&S Building Systems.)

287

SOME COMMON PROBLEMS AND ERRORS 288

CHAPTER TEN

will surely increase in the future. But to paraphrase Calvin Coolidge's famous saying, fabricators are in the metal business. The industry has accumulated considerable know-how in the design and construction of metal-clad structures, but the same level of expertise cannot necessarily be found when dealing with masonry or concrete, let alone GFRC or ETICS. To ensure a successful project, the owner's design team would be wise to provide as much detail of the rigid exterior wall as they would with a conventionally framed structure. Interaction between exterior walls and prefabricated structures is a common source of problems. Individually. 7 already mentioned and chap. 11 will further explore an inherent conflict between hard wall materials and flexible metal frames. As for example in chap. 7, a fairly substantial eaves element is required for lateral support of vertical masonry and concrete walls, or for gravity support of heavy parapets and siding panels. And yet, prefabricated buildings typically do not have the large wrap-around roof beam found in traditional framed steel structures. Instead, the usual feature is a lightweight eaves support that works well with metal roofing and siding but is almost useless for hard walls. Most major manufacturers are able to modify your standard frame for hard wall conditions, but some others may not be as knowledgeable. If the wall construction requires a hot rolled steel section in the eaves or elsewhere, it is best to design it in-house and indicate it on the contract drawings. To demarcate responsibility, such items designed by the owner's team can be excluded from the manufacturer's project scope, if necessary. Likewise, the architect-engineer team must decide how to support part-height rigid walls. The in chap. 7 is to “fix” these walls to the ground and give up any lateral support of a flexible metal structure. If this solution weren't shown, manufacturers would probably think in Z-beams as in metal cladding. It is often expected that the CMU and the concrete walls act as load-bearing or shear-resistant elements. While this use is certainly rational and economical, it presents a different set of problems. Just like building foundations, walls must be designed by the architect-engineer team, but for what loads? In theory, the fabricator could supply the horizontal and vertical loads on the wall after the frame construction is complete. In practice, the wall design must first be based on the designer's own analysis and later be confronted with the manufacturer numbers. The manufacturer can even expect the architect-engineer team to design the connections between the wall and the metal structure. An often overlooked point: the manufacturer's standard wall mounts are designed for metal paneling and are not necessarily adaptable to non-metal walls. Items such as doors, windows and shutters that may be available from the manufacturer may need to be purchased offsite. As we alluded to in the previous chapter, it can definitely be a good idea.

10.2 FIXED BASE COLUMNS VERSUS STACKED COLUMNS Most experienced structural engineers agree that it is quite difficult to achieve complete column base fixation in metal building systems; some of the reasons are in chap. 4 and more in chap. 12. Therefore, 'external' designers routinely assume that the bases of the columns are pinned, ie no bending moments are transmitted to the foundations. Unfortunately, not everyone remembers to put this assumption in writing - in the contract documents. Without being "pinned," the assumption of the base of the pen exists only in the minds of the engineers who specify it. Manufacturers are welcome to propose a mast-type system with a fixed base construction, which they consider to be a more cost-effective solution. Since manufacturers don't see the additional foundation costs, they can reasonably expect the solid foundation solution to save money. However, in a public tender situation, when the manufacturer's design is available, foundations may already have been designed with studs in mind and may even be built by, or at least subcontracted to, a concrete contractor. Because fixed-base column construction results in an unforeseen bending moment being applied to the foundations, new construction is almost always required.

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

289

At this point, the homeowner has a choice between accepting a major change order from the contractor or engaging in a protracted battle with the fabricator. Both of these could easily have been avoided by including a sentence in the drawings and specifications of the contract: "The prefabricated columns of the building must have solid bases and must not transfer moments to the foundations".

10.3 ANCHOR BOLTS Anchor bolts, or as AISC prefers to call them, anchor rods (Fig. 10.3) cross a line of demarcation between the design responsibility of the steel fabricator and the registered structural engineer who is responsible for the entire project. An inexperienced designer usually assumes that anchor bolts are supplied by the manufacturer, who eventually has factory drawings that include a bolt schedule clearly showing bolt sizes and locations. In addition, factory drawings for column base plates also include screw sizes; The calculations shown indicate how many and what type of bolts are needed. In reality, the manufacturer usually does not supply the screws. This fact is clearly stated in the MBMA Industry Common Practices. The manufacturer does not even determine the length of the bolt, which depends more on the design of the foundation than on the parameters of the metal structure. As in chap. 12 the strength of anchor bolts is determined by one of two factors: tensile strength of the steel section and strength of the concrete. The former is determined by the manufacturer, the latter by the design professional. For example, anchor bolts embedded in an isolated pier are likely to be longer than those in a large, thick foundation. This is because the tensile strength of the pier concrete can be less than that of the bolts, which may require an increase in bolt length to engage the pier reinforcement. Since many column studs are actually placed in narrow concrete piers, the registered structural engineer may want to specify the minimum length of embedded studs and minimum edge distance. The engineer should then check the size and location of the anchor bolts supplied by the manufacturer for compliance with the specified minimum distance from the edge. Often the manufacturer's standard baseplate details show that the anchor bolts are placed too close to the concrete edge to provide the values required by the designer. To avoid disputes, the structural engineer is responsible

FIGURE 10.3 Anchor Bolts.

SOME COMMON PROBLEMS AND ERRORS 290

CHAPTER TEN

will take a much stronger position if he has clearly indicated the edge distances in the contract documents. It is more difficult to try to change the manufacturer's specifications after the order has been placed. In many cases where significant anchor capacity is required it is not possible to use flush mortise slings as the posts would be placed too close to the edge and the anchor bolts would not have the desired edge spacing. A belt bypass may be the solution. This makes it clear once again how the registered structural engineer has to influence those areas of the metal construction that are supposed to be the responsibility of the manufacturer.

10.4 DOOR FRAMES 10.4.1 Construction requirements Many commercial and industrial buildings have large sectional doors and other wall openings (Fig. 10.4). Fabricators of prefabricated buildings share common procedures for framing door and window openings in metal-clad exterior walls. Extended mullions of cold-formed section for the next horizontal joist or eaves brace usually frame an opening, with a similar header at the top (Fig. 10.5). This detail can work for small to medium sized doors and windows, but not for large industrial sectional doors. With wide sectional doors, the slats that are subjected to high wind loads are usually very flat and flexible to act as beams to bridge the gap between the frames. Instead, high wind doors behave like elastic membranes that rest on the jambs, providing additional support at the top and bottom edges. These membranes are severely deformed under stress, which can cause the doors to pull out of their tracks. Unfortunately, that's exactly what often happens. Hurricanes are usually accompanied by torrential downpours, and windblown rain that seeps inside the building ruins its contents and results in a large financial loss, even if the rest of the outer skin remains intact. For example, if you are examining damage from two 1970 hurricanes, ref. 2 notes that the wind caused the light frames around the sectional doors to deviate excessively, leading to derailment of the roller supports and subsequent door failure.

FIGURE 10.4 Large sectional doors are an integral part of metal building systems. (Photo: Maguire Group Inc.)

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

291

FIGURE 10.5 Typical framed opening for a sliding door. (Star building systems.)

To avoid this, large swing doors are usually fitted with windbreaks – devices that positively attach the slats to the guides. Now the door cannot be easily inflated because it is attached to the supports, but these supports must be essentially immobile under load. The horizontal reactions exerted by the membrane on the supports are quite substantial and the abutments must be designed to withstand them without overstressing, excessive deflection or rotation. Otherwise, both the brackets and the door will be blown out, or wind may enter the building. The manufacturer of revolving doors determines the size of the reactions depending on the door size, wind load and other parameters. The Doors and Access Systems Manufacturers Association explains the various components of wind load response in its Technical Bulletin No. 251, “Architects and Engineers Need to Understand the Loads Exerted by Overhead Doors”.

SOME COMMON PROBLEMS AND ERRORS 292

CHAPTER TEN

the problems of installing sectional doors in hard walls. The DASMA website (www.dasma.com) is a valuable resource for sectional door designers. Bulletin #251 states that laminated channel mullions "may rotate under wind loads and the gate curtain may be torn from the tracks". In addition to wind load, the structure around overhead doors must be able to support the weight of the door when lifted, as well as the weight of siding and other wall materials. With roller shutters, the weight hangs eccentrically on the mullions, transmitting the reactions to the eaves support - or a horizontal member of structural steel that extends between the mullions.

10.4.2 Recommended construction details We recommend that a conceptual construction for the frame around overhead doors is made by the designer's engineers. The fabrication fabricator can use the jamb reactions and door weight provided by the door supplier to determine actual frame sizes. For critical applications, the entire system can be designed by the designer and potentially excluded from the building fabricator's design scope. A possible framework is shown in Fig. 10.6. This construction uses hollow structural steel members (HSS), i. H. Tubes that have excellent torsional properties and are uniquely suited to withstand horizontal loads from all directions.

10.4.3 Manufacturer alternatives to tube designs Some manufacturers do not use structural tubes because tubes are outside their usual repertoire of engineered plates, wide flanges and U-sections. These manufacturers may propose a framing system of hot-rolled profiled studs bounded at the bottom by the eaves brace at the top and supported laterally by the wall joists (Fig. 10.7). The eaves brace would be braced laterally to prevent torsional failure under wind loads (Fig. 10.8). We should mention here that unlike pipe and pipe

FIGURE 10.6 Suggested structure for large sectional doors.

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

293

that have "closed" sections, the C- and Z-shaped light meter elements have "open" sections with low torsional strength. Note that the eave reinforcement in Fig. 10.8 is achieved with cold-formed channel sections similar to our recommended purlin reinforcement design discussed in Chap. 5. If it is necessary to use this detail, we recommend continuing the purlin bond behind the eaves bond to spread the load over as many purlins as possible. Another common, but less effective, method of bracing the eaves brace to the posts is to have an elbow that extends to the nearest purlin - and with no purlin reinforcement behind it. Obviously, a single weak-direction curved left side is unlikely to provide lateral reinforcement for the eaves brace.

10.4.4 Specify frames in contract documents Regardless of the detail used, it is important that the contractor provide a frame around the doors. MBMA Industry Common Practices consider such a frame an accessory and state that it is not provided unless specifically required. The port itself is also not included in the system and must be specified separately; must be designed for at least the same wind load as the building walls. Despite our preference for tubular door frames and headboards, any rationally designed structure is preferable around overhead doors in a situation where the issue does not receive technical attention. Leaving the project in the hands of field workers can be disastrous. Consider the jambs of the sectional door in Fig. 10.9: they are supported at the top only by the remnants of the beams cut to install the jambs.

10.5 ROOF HVAC UNIT MOUNTS 10.5.1 Rooftop units Roof mounted or suspended HVAC units can include anything from small fans and heaters to large air handling units. Although mechanical equipment is not part of metal construction systems

FIGURE 10.7 Hot rolled rail abutment attached to wall studs. [Manufacturer recommends the use of (4) 1Ⲑ2 in. ASTM A307 screws] (Nucor Building Systems.)

SOME COMMON PROBLEMS AND ERRORS 294

CHAPTER TEN

FIGURE 10.8 Connection of doorstop to braced overhang bracket. [Manufacturer recommends the use of (4) 1Ⲑ2 in. ASTM A307 screws] (Nucor Building Systems.)

and are not provided by the client, device consoles must be integrated into the roof construction. Unless expressly stated in the contract documents, liability for this support tends to be disregarded. Of course, the design profession is well aware that equipment does not belong on metal roofs: equipment restricts the movement of roof panels and encourages leakage through penetrations. Many roofs have been ruined due to careless installation and maintenance of equipment. Unfortunately, there are few inexpensive alternatives to the rooftop location. Mechanical roofs, common in skyscraper construction, are a rarity among prefabricated buildings. It looks like roof-mounted HVAC equipment will continue to degrade the look and function of metal roofs for years to come. There are two basic methods of supporting roofing equipment: a continuous curb and a steel frame raised on legs. A properly designed and installed burred curb may be less prone to leaks than discrete penetrations in the frame legs.

10.5.2 Metal Curb Curbs are custom made to suit the particular roof profile and fit perfectly into the ridges of the panels (Fig 10.10). Curbs for vertical roofs can consist of two parts: a light curb pre-assembled to a base plate of the same thickness and configuration as the roof, and a heavier curb

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

295

FIGURE 10.9 These fold-down door frames are supported at the top by nothing other than the struts, which were cut to allow the frame to be installed.

(e.g. 10-gauge) unitary support structure directly connected to roof purlins. The two parts move independently; any difference in their movement is absorbed by the blink. Butler Manufacturing Co.4 recommends that the two piece design be used for equipment weights from 2000 to 4000 lbs. For lighter equipment, a one-piece curb, typically made of 14-gauge steel, may be sufficient. The manufacturer recommends having the curb fitting done by a curb specialist. Still, many planners include curbs in a work scope for the fabricator to increase the chances of a better fit and coordination. Despite its sophistication, even this curb system can be prone to leaks if simply placed on top of the roof and relying only on intermediate caulking. For best results, the curb and roof should be installed at the same time, allowing the curb shingles to be placed under the sloping roof slab and over the sloping roof slab to prevent water from entering the roof. In addition, Buchinger5 recommends that the curb bottom plate be large enough to allow at least 1 foot of space between the end of a set plate and the nearest edge of the curb or your cricket to prevent water accumulation at this critical point. The curb must be supported on all four sides by purlins or additional frame members. The actual information depends on the manufacturer, the type of metal roofing and the load. Two exemplary details are shown in FIGS. 2 and 3. FIG. 10.11 and 10.12. In either detail, if the opening is to interrupt a purlin, a properly designed head and curved purlin (or heavier frame) on each side is required.

SOME COMMON PROBLEMS AND ERRORS 296

CHAPTER TEN

FIGURE 10.10 Vertical Spliced Roof Metal Curb Frame. (Butler Manufacturing Co.)

FIGURE 10.11 Support structure for floating/structural curb. According to the manufacturer, this detail is suitable if the weight of the roof unit does not exceed 1200 lb and the maximum load on a purlin line does not exceed 600 lb. (Nucor Building Systems.)

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

297

FIGURE 10.12 Support structure suitable for roof units between 1200 and 6000 lb or where each row of purlins would support more than 600 lb. (Nucor Building Systems.)

Figure 10.11 can be used for a moderately loaded floating curb. Here the curb supports purlins and elbows. Figure 10.12, intended for heavier loads, uses structural steel sections between primary frames parallel to purlins. Light channels frame the gap between the hot-rolled profiles. The manufacturer may request that this light skirting be placed slightly above the other purlins. It goes without saying that the curb materials must be compatible with the roof. For Galvalume roofs, consider Galvalume, aluminized steel, aluminum, or stainless steel curbs.6

10.5.3 Elevated Structure on Legs A raised steel structure concentrates the weight of the equipment on four legs (Fig. 10.13). These point loads often exceed the carrying capacity of purlins and require wide-flanged steel girders for support. For this reason, it is best to rest your legs on the main structures of the building whenever possible. As with metal curbs, the problem of tightness arises whenever a frame support leg penetrates the roof. A common solution is to use an elastomeric sleeve pipe liner (also known as a ceiling bushing) to cover the penetration. The sleeve liner is available in a variety of sizes to accommodate different pipe or column diameters and must be able to accommodate the differential movement between the metal cover and the underlying structural supports. Otherwise, it is certain that there will be leakage. Boots are commonly, but incorrectly, installed over roof ridges, inviting water to leak through difficult to seal panel seams, unprotected from intrusion. A better detail is to place the roof penetrations on the flat part of the panel where a more effective seal can be made.5 The detail would be as in Fig. 10.14 showing the hood flashing on a flue outlet. Who designs the structural roof frames? Some architecture and engineering firms prefer to design the structures in-house and ship them separately from the prefabricated structure. Some others prefer to make coordination easier by requiring the manufacturer to design and supply the frames. Therefore, roof openings are likely to be incorporated into the structural steelwork, all interferences between purlins, purlin braces and support beams noted and corrected, unit weights incorporated into the structural loading on the roof and provided for all penetrations.

SOME COMMON PROBLEMS AND ERRORS 298

CHAPTER TEN

Fig. 10.13 Elevated frame on legs.

FIGURE 10.14 Sealing the pipe penetration with a flexible cover placed in the flat area of the plate. Sealant is applied between the panel and the metal flange of the hood prior to attachment. (Central.)

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

299

In any case, the location and weight of the HVAC equipment must be closely coordinated with the building fabricator to ensure adequate structural support. It is easy to see how a lack of such coordination can easily lead to "extras" when additional structure needs to be provided after the roof is in place.

10.6 ROOF SNOW AND PROJECTING LOADS As the reader may understand from Chap. 3, there are two equally distributed roof loads applied from the outside: snow load and roof traffic load. The nature of these loads and the differences between them have already been explained; our current discussion addresses problems of their specification. In the last two decades, the application criteria for these fillers have changed dramatically, confusing both the manufacturer and the designer; even building authorities are sometimes unclear on this matter. The owner needs to be sure that all manufacturers interested in the project will apply the same design loads and deliver buildings of similar strength. Otherwise, a party that claims to meet the project's design criteria but uses sleight of hand to achieve smaller design payloads than those used by others will gain a strong competitive advantage. As Miller and Evers7 noted, competition can lead some companies to look for a way to gain price advantage. For the hard-pressed competitor, this could mean employing questionable code interpretation methods and project loads to get the lowest price. Typically, this is less outright cheating and more a clever interpretation of what is sometimes permissible damage control or a failure to consider a recent code update. Experienced structural designers know that the magnitude of structural loads is one of the most important factors affecting construction costs. As Ruddy8 found, for single-story steel-framed buildings, the cost of the structure increases by 2 cents per square foot for each additional pound per square foot of superimposed load. For example, if the roof structure, columns, and foundation cost $5 per square foot when the structural load is 20 lb/ft2, the same building could cost $5.20 per square foot when designed for a roof load of 30 lb/ft2. A common problem is the designer's failure to distinguish between traffic loads and snow loads on the roof. Designers need to compare the design values listed in the local building code for both and clearly understand which controls the design for the design site. For the northern regions it is usually snow, for the south - live load from the roof. The code can specify the snow load value as "snow on the ground" or "snow on the roof". Often the ground or "base" snow load can be converted to the roof snow load using a multiplication factor of 0.7 (and other factors). To avoid confusion, contract documents should list a snow load value for the project roof (if snow dominates the project) and clearly identify it as snow. A careful review of the project certification letters (see Chapter 9) should be carried out to ensure that no bidder has incorrectly assumed that the project load is a ground snow load that could be reduced by a further 30 percent. That reduction alone could reduce the amount of steel used by as much as 5 percent.7 Traffic loads on roofs present another complication: the reduction in traffic loads allowed by regulations for large areas—generally over 200 ft2—carried by a structural member of the roof to be carried. In many codes, the working load of low-slope roofs is assumed to be 20 lb/ft2 for sub-areas up to 200 ft2, 16 lb/ft2 for sub-areas between 201 and 600 ft2, and 12 lb/ft2 for sub-areas over 600 ft2. Therefore, where the contract documents refer to a "live roof load" of 20 lb/ft2, all bidders will reduce that load equally. However, when the documentation states a "living roof load" of 30 lb/ft2 when it actually means a roof snow load of 30 lb/ft2, a costly problem can arise. Some manufacturers understand what snow load refers to and will design all roof panels to 30lb/ft2, while others can scale them down to the roof payload numbers listed above. A load reduction from 30 to 18 lb/ft2 can save up to 8 percent in primary structure costs.7 The manufacturer that accepts this reduction will clearly be able to win - unfairly compared to the competition - due to the ambiguous contract documents. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

SOME COMMON PROBLEMS AND ERRORS 300

CHAPTER TEN

10.7 GUARANTEED FREIGHT SPECIFICATION 10.7.1 Challenges As discussed in Chap. 3, Collateral load is a subset of permanent load, which includes the weight of all materials except permanent construction. These materials may include pipes, sprinklers, mechanical wiring, electrical wiring, ceilings, and trim. Some typical weights of these components are given in Chap. 3, where we concluded that a commercial or industrial building with sprinklers, lighting and mechanical ducts - but no ceiling - could in many cases be rated for a collateral load of 5 psf. It is easy to demonstrate that a uniformly distributed load of 5 psf applied to a simply supported purlin of 25 feet (the typical span of the purlin) spaced 5 feet on center will produce a slightly larger bending moment than a 300 lb concentrated load applied to purlin mid-span. So the design load of 5 psf seems substantial enough to account for the weight of many hanging objects and even small ceiling fans. However, the allowable collateral load of 5 psf may not be sufficient to accommodate the weight of some common building elements, e.g. B. the sprinkler system, which is usually 8 inches. or larger. An 8 inch pipe filled with water weighs about 50 pounds per linear foot. These pipes are typically suspended by sprinkler installers in center purlins every 10 feet, while center purlins are typically 5 feet apart. Therefore, all other Tuesdays would likely carry a projected suspended load of 500 pounds while their neighbors would carry none. Although the average collateral load on the roof is theoretically not exceeded, the purlins supporting the sprinkler mains carry more than their share of the collateral load; These purlins can become overloaded under full design snow or roof traffic loads.

10.7.2 Two Ways to Account for Heavy Pipe There are two ways to solve the problem: (a) add a third directly over each heavy pipe and support the sprinkler mesh on each third instead of alternate thirds, or (b) elevate the level of the collateral fee every Tuesday. The first solution—adding purlins in just the right places on the pipes, or dictating where to support the sprinkler network—is probably the most economical, but not the most practical. It requires an unusual degree of coordination between the fabricator, the building erector, the sprinkler system designer and the sprinkler system installer. Far from such coordination being the norm, the homeowner sometimes chooses to design and install fire protection after the building is constructed. Unfortunately, the opportunity to involve the structural engineer is then lost and the pipes may be installed in a less than desirable manner. Figure 10.15 shows a sprinkler pipe (fortunately not the main pipe) hanging from a small diameter prefabricated truss diagonal. The second solution - increasing the collateral burden every Tuesday - is more likely to succeed, but obviously increases costs. If we accept the fact that the structural action of a 300 lb concentrated load is approximately equal to that of a 5 psf uniform collateral load, a 500 lb concentrated load would require a uniform tolerance of approximately 8 psf. So what level of security should be declared? The author's practice is to specify at least a uniform collateral load of 5 psf, increasing to 8 psf or even more if sprinkler lines or other heavy overhead loads are expected. Other sources suggest similarly high collateral charges. For example, Westervelt9 recommends using 8 psf for “a moderate amount of mechanical and electrical items”. Miller and Evers7 state that 5 or 10 psf is a typical value for collateral stress. It is not recommended to increase the collateral load above 10psf as this load level indicates the presence of some supported heavy objects. It is generally very difficult to support heavy loads from light steel members and this situation is best avoided by providing special support structures. Some manufacturers use lower collateral loads for primary frame construction than for purlins. On the surface, this approach seems reasonable, but it is important to remember that building regulations do not provide criteria for such load reduction.

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

301

Figure 10.15 This is not a good way to support collateral load. The tube should have been hung from a grid plate point.

10.7.3 Complaints about high warranty fees High warranty fees sometimes cause controversy and complaints. Prefabricated building manufacturers tend to compete on price with suppliers of traditional structural systems and with each other, and there is pressure to use the least amount of collateral load - the only type of structural load that is not firmly established by regulations. In fact, relatively small changes in collateral load can result in significant price changes for metal roof construction. An additional permanent superimposed load of 5 psf contributes relatively little to the total load in a concrete or mild steel frame building, but is significant in metal building systems where roofs typically weigh only 2-3 psf. Also, as in chap. 9, the task of specifying the design loads is usually the responsibility of the owner, who may or may not be technically savvy or swayed by the sharp advice of a contractor trying to stay on a tight budget. As a result, the specified amount of project security was not used or not used at all in some projects - a situation full of dangers. Another argument against high collateral loads is: why would all purlins be oversized to compensate for a single line or a few lines of purlins that really carry the heavy load? According to the author, the real behavior of metal construction systems under high loads is not fully understood. A few extra pounds of construction collateral load can mean the difference between survival and collapse for a structure nearing its breaking point. As we discuss below, the failure of a single row of purlins can bring down the entire metal building.

SOME COMMON PROBLEMS AND ERRORS 302

CHAPTER TEN

10.8 DEFINITION OF OPTIONAL STRUCTURES Some "romantic high-tech" architects like to incorporate exposed steel structures into the exterior of buildings. The exposed framing may look dramatic on sketch paper, but the real-world structure may not be as good for a variety of reasons. First, as mentioned in the previous chapter, there can be a problem getting a high quality factory paint or colored zinc finish on the primary structure. Exposed steel with a poor coating may not survive long in a corrosive atmosphere. Second, the steel frame is used more efficiently when column and beam flange supports are available; Such an orthosis, as well as bolted-on limb splices, may not look particularly attractive when exposed to the eye. Third, some architects forget that the prefabricated structure is built from relatively thin panels. Web-to-flange welding usually occurs on only one side of the web. Aside from raising conceptual concerns for some structural engineers, this type of welding does not approach a known smooth fillet line that hot rolled steel offers and does not provide a good weather barrier. Fourth, frame-to-wall connections are difficult to weather; The integrity of the wall depends solely on the sealants. Due to manufacturing and assembly tolerances, the gaps between the frame and fairing are seldom the same width and may require large amounts of sealant. The results are seldom pretty. On a project that used exposed steel, the only party who was completely satisfied with the appearance of the building was the gasket seller.

10.9 FAILURES OF PRE-FRAMED BUILDINGS 10.9.1 Main causes of failure in metal buildings Like any other type of structure, pre-fabricated buildings can and do fail. Some of the failures were quite dramatic as they involved complete building collapses. Consider the recent experiences of clients of commercial and industrial property insurer FM Global: “Over the past eight years, approximately 60% (dollar loss) of FM Global clients' roof break-ins have involved the construction of metal roof systems (MRS). This consisted of 74 collapses that caused nearly $221 million in damage -- an average loss of nearly $3 million per incident. Damage typically accounted for about two-thirds of the entire structure, versus about a quarter of the entire structure for other types of construction.”16 Among the many possible causes of failure, the most important are 1. 2. 3. 4. 5 .

Overloading (Fig. 10.16) Improper design practices Faulty design Degradation Others, such as B. improper modifications or the lack of an original design

These causes are discussed separately in the remainder of this chapter. A combination of two or more of these causes is often responsible for building failure.

10.9.2 Failures Due to Overloading Many metalwork system failures can be traced to overloading - a condition in which the structure is subjected to more stress than it should be able to withstand. For example, a building designed to withstand a 30 psf snow load may receive a documented snowfall record of 80 psf, or a building designed to withstand

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

303

FIGURE 10.16 Building collapse during heavy snowfall.

80 mph winds are hit by a 150 mph hurricane. A building collapse in either of these two scenarios can be attributed to simple overloading. In fact, even a perfectly designed and constructed building will collapse if the overload is severe enough. But what if the building fails under a load only slightly higher than the design load, for example a measured snowfall of 32 psf on a roof rated for 30 psf? It could be argued that such a roof should be able to withstand at least the design level of snow load times a safety factor. Therefore, for a projected snow load of 30 psf and a safety factor of 1.67, the predicted failure load would be 51 (30 ⫻ 1.67) psf. In theory, the building should be able to support an even greater load, as the actual strength of steel is likely to be greater than its nominal rating. In practice, design and construction irregularities can greatly reduce the theoretical breaking load. (And there are those who argue that buildings need not be able to support a gram above design load.) As discussed in Chap. 4, sec 4.12, prefabricated buildings are designed for near full efficiency, leaving little 'fat' to compensate for random local overloads caused by some common factors. For example, overloading can occur on frame columns that have not been designed to the minimum eccentricities resulting from MBMA-allowed manufacturing tolerances, particularly curvature. The column sweep is a constructed cut measured in the direction perpendicular to the web. According to Table 9.2 of the MBMA Handbook10 Common Industry Practices, allowable sweep in inches is 1Ⲑ40 equal to column length in feet. A 30 foot high column can have up to 3Ⲑ4 inch sweep. Unanticipated weak axis bending moments resulting from the eccentricity of the design load can overwhelm a column designed for purely axial loading. When it comes to wind overload, the worst damage tends to occur at the corners, eaves and ridges of buildings – the areas where modern codes can dictate much higher local pressures than some older codes. The provisions of the Wind Code are continually being refined and reflect the historical performance of buildings designed under previous editions of the Code.

SOME COMMON PROBLEMS AND ERRORS 304

CHAPTER TEN

Figure 10.17 illustrates building damage from localized wind overload. Investigators traced it to a microburst — a small but powerful gust of wind. Microexplosions have been blamed for many plane crashes during takeoffs and landings. Note that the end wall beams and supports and the roof purlins are curved in the first outer span - and that the purlins are curved upwards. The sectional door also failed, consistent with our discussion in Section 10.4.

10.9.3 Effects of temperature stress on bus congestion Another example of possible overvoltages under common but sometimes unplanned conditions concerns additional stresses and displacements resulting from temperature stress. As in chap. 5, when the primary structures cannot move laterally under thermal stress, the purlins, constrained by free expansion and contraction, experience a significant build-up of axial compressive stresses. At the other extreme, when the frames are completely free to move, the expansion and contraction matrices push and pull the frames out of plane and move them away from their original positions. This leads to an additional torsional load on the frame members. To some extent, a properly designed and installed flange mount can mitigate this torsion. Regardless of the design assumptions, additional stresses on purlins or additional beams must be considered. Most likely there will be some of both.

FIGURE 10.17 This Southern California building partially failed under wind loads. Note the buckling of the bulkhead beams and roof purlins and the failure of the sectional door. (Photo: J.R. Miller & Associates.)

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

305

In a cold climate, the worst-case scenario tends to affect primary structures in freezing, snowy winters, when the structure's added torsion (if not unloaded) from its internal motion is combined with the bending stress of heavy snow. In a hot climate in summer with high moving loads on the roof, the worst-case scenario could occur: the bending compressive stresses in the purlins due to the changing load would be combined with axial compressive stresses due to the temperature limitation. To understand the numbers involved, consider a 500 foot long uninsulated warehouse with purlins bolted together with no expansion devices. Roof purlins are subject to a temperature change of 100°. When free to move, the purlins will expand and contract a total of nearly 4 inches - meaning the primary end structures will be displaced 2 inches on each side from their original position. When purlin movement is restricted by adjacent structures, the purlins build up over 19,000 psi of axial stress!

10.9.4 Wrong Construction Damage The causes of some metal construction damage are due to poor projects. Sometimes non-conservative and unrealistic design assumptions are made. The load is sometimes misreported, as discussed above in the sections on collateral loads and the difference between snow and roof loads. These two factors are sometimes aggravated by improper maintenance. Consider the hypothetical case of heavy snow accumulation on a metal roof. The building is still able to withstand an even and balanced snow load (perhaps hardly), but the situation changes when the owner decides to start shoveling snow. Without proper guidance, workers begin by thoroughly cleaning the tailstocks. This leads to a partial load condition which proves more critical than the previously evenly applied load and leads to overstressing and purlin failure. This is a combination of design flaws (the building was not designed for part loads) and improper snow removal techniques (which led to part loads in the first place). Incidentally, Appendix A8 of the MBMA manual10 provides an overview of good snow clearing techniques. He recommends consulting the building fabricator or a structural engineer before clearing snow, and suggests, among other things, gradually removing snow "in layers from the entire roof" rather than an entire bay at a time. Common areas of questionable design include frames around overhead doors, as discussed in Section 10.4, and in particular the design of cold-formed beams and purlins. Secondary members deserve special discussion.

10.9.5 Purlin and transom failure There is much controversy in the design of secondary roof and wall panels. As we did in Chap. 5, the structural design of cold formed C and Z profiles is quite complex and their actual behavior is not fully understood. It is not surprising that the designers of these sections use different design assumptions, some of which are controversial, and use different construction techniques. Among the controversial issues discussed in Chap. 5 are: ●

●

Use of prismatic design (reduced rigidity) versus non-prismatic design. While the prismatic design is easier to use and acceptable, it can result in some over-stressing of the elements in the negative areas and at the seams compared to the more realistic non-prismatic (full rigidity) design. Forcing heavy Z-profiles on columns into each other, which can lead to some built-in twists in purlins (see fig. 5.8 in chapter 5). Consider the imaginary tipping point as the fulcrum (this particular assumption has several well-known proponents). Neglect of partial loading.

SOME COMMON PROBLEMS AND ERRORS 306

CHAPTER TEN

●

The use of very low roof pitches (e.g. 1Ⲑ4:12) in cold regions, which lead to a depression in the first inner rafter line near the eaves and can lead to leakage and ice formation there, as discussed in Chap. 11. This accumulation, combined with overhead loading of the sprinkler network (discussed in Section 10.7), can result in overloading of purlins near eaves. Inadequate purlin bracing - or no bracing at all - coupled with poor purlin stability on the supports.

The last point is by far the most serious and deserves a separate space.

10.9.6

Failure due to missing purlin and girt bracing As in Chap. 5, purlins and flanges without lateral bracing only have a fraction of the load-bearing capacity of fully braced profiles. To recall this discussion, the bracing must be able to perform three tasks: 1. Lateral bracing of the compression flange 2. Securing of purlins or beams against rotation 3. Securing of the entire set of purlins and covers against lateral displacement Effective Date Abutment To achieve this For this reason, the bracing system must provide stability to the entire C or Z profile, not just one wing of it. For this reason, metal roofs, even in the fastening variant, cannot fulfill all three tasks. Of course, the element must be stable between and on the supports, which means that the discrete bracing must be complemented by some sort of anti-roll device. However, based on the author's experience of examining damage to steel structures, correctly designed purlins and orthoses are still relatively rare. In many older buildings there is no support at all. Beams and purlins that are not laterally stiffened tend to fail in lateral-torsional buckling (Fig. 10.18) long before their full bending capacity is reached. This may help explain why some prefabricated buildings fail under heavy but not extreme snow and wind loads. Note that the buildings in Figs. 10.16, 10.17 and 10.18 do not have discrete secondary bar bracing. Many other engineers involved with metalwork systems have confirmed these observations verbally and in writing. For example, Zamecnik11 examined several prefabricated buildings with obvious damage to purlins that suffered from snow loads well below design values. Some of these roofs partially collapsed. He blames much of the blame on inadequate third reinforcement. (The MBMA disputes these conclusions, insisting that the failed buildings were old, possibly misdesigned, and therefore did not represent modern practice.) Peraza12 describes his investigations of multiple metal building collapses in the 1990s that provide full purlin support, investigators agree concluded that the actual degree of lateral support was only about 60%. Peraza points out that for roofs with vertical seams, “it was undoubtedly known at the time that 100% bracing was an unrealistic expectation”. In another fancy roof design, investigators concluded that purlins could only support about 59 percent of the load that fully reinforced purlins could. This building also contained an interesting bracing system that was considered questionable at best. The percentages given above are consistent with the results of independent baseline tests in which the author participated. The degree of lateral bracing of the structure cover with trapezoidal profile in the positive third range was only 52 percent in these tests, even with the mildest interpretation of the results. The test was aborted when the purlins twisted so much under load that the roof construction was a generous distance from the test frame. This brings up a very important point. The strength and stiffness of distorted (rotated) C and Z profiles decreases as the degree of rotation increases. When these sections eventually flatten out, they are as strong as their weak axis section properties will allow. The more the purlins rotate under constant load, the weaker they become and the more they deviate vertically. In absence from

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

307

FIGURE 10.18 Buckling of unbraced side members and top support under wind action. (Photo: J.R. Miller & Associates.)

effective bracing of purlins or other external factors that can stop the rotation, purlins can continue to rotate and their carrying capacity will continue to decrease, until the purlin strength is insufficient to support the load. Figure 10.16 is fairly representative of the large deflections in purlins that tend to accompany roof collapse. The independent baseline tests mentioned above identified another problem with unbraced or lightly braced purlins: much of their rotation under load, as well as vertical and lateral deflection, can be irreversible. The thirds may already be weakened when the next heavy load is applied. The benefits of Tuesday bracing can be seen from the experience of FM Global customers. During the severe winter of 1995–1996, there were virtually no collapse losses at facilities that implemented FM Global's technical recommendations.

10.9.7 The Collapse Scenarios A question can be asked: Why does excessive purlin rotation and failure under heavy snow loads tend to collapse the entire building? Couldn't the purlin and roof set hang like a membrane from the primary inner frame? Unfortunately, the membrane analogy does not work for end walls: they generally cannot withstand the enormous horizontal catenary forces generated by membrane action (see the discussion of swing door behavior in Section 10.4). There is also an issue with the presence of flange clamps (“kickers”) on the lower flanges of primary structures.

SOME COMMON PROBLEMS AND ERRORS 308

CHAPTER TEN

Chapter 4 explains that flange reinforcement is typically used to improve the efficiency of frame members under loading conditions that compress their lower flanges. However, a flange support that is only mounted on one side of the frame can become a problem in heavy snowfall: sagging purlins press the kicker angles and move the lower frame flange sideways (Fig. 10.19a). This puts additional torsional stress on the frame. Because the structure may already be fully loaded from heavy snowfall, any unanticipated torsion can lead to overstressing and failure. When that happens, the entire building can collapse - if the stress is strong enough. (To avoid additional torsion, flange reinforcement should be installed on both sides of the frame, as shown in Fig. 10.19b, so that the forces from the sagging purlins partially or completely cancel each other out.) These failure scenarios go against each other. the desire of many modern building codes to prevent the progressive collapse of buildings. In a progressive collapse, a single, or even a single, overloaded or damaged structural member will collapse the entire structure. As ASCE 713 puts it, buildings and other structures must be designed to withstand localized damage while maintaining the integrity of the structural system as a whole and not damaging it to an extent disproportionate to the original localized damage. In a report by Murtha-Smith et al. In the proposed model for the progressive collapse of steel buildings14, failure begins with some overloaded purlins, perhaps in the final spans, failing and suffering major damage.

FIGURE 10.19 The number of struts on the main frame flange affects frame stability under high loads: (a) the struts on one side shift the lower frame flange when the purlins yield; (b) Braces on either side balance the opposing forces of the flaccid thirds.

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

309

deviations. Then the metal roofs extend in a chain, pulling on the adjacent purlins with an immense force they cannot withstand. As a result, purlins in adjacent spans also rotate and fail. When there is one-sided flange support, sagging purlins push out the lower flanges of the beams as just discussed. The rafters twist and fail, pulling the outer columns and walls inward. As soon as the purlins of one field are screwed to those of neighboring fields, the purlin break spreads throughout the building. The author's experience studying collapses of metallic buildings is consistent with this model. Another suspected failure mode in cold regions is water freezing in purlin supported gutters. The resulting overload of the first or second purlin causes it to rotate to a nearly horizontal position. The roof then pulls the remaining purlins with it, which also “lie down” and fail. A suitable eaves and ridge set, shown in Chap. 5, can help prevent this type of progressive collapse. The insurance industry is aware of the fact that some metal buildings can fail very quickly under heavy snow loads. Such outages occurred in Northwest Georgia in March 1993 and in eastern Pennsylvania in 1994. A magazine published by Factory Mutual notes that "Metal structural systems fared worse in the '93 Storm, perhaps because of their design... When a breakdown occurs, it's usually total." , from start to finish. Again, the obvious cause of this failure was a large amount of snow. These observations are supported by the statistics cited in Section 10.9.116 and other sources. For example, in a recent discussion among representatives of civil engineering organizations in 14 states, as reported in Structure17 magazine, a question was asked, "What types of buildings appear to have had the most problems in recent snowstorms?" The response was, "The vast majority of responses referred on roof constructions made of wood-coated lattice girders and low building systems made of metal". Most respondents "believed that the actual loads were consistent with the design loads and acceptable safety factors required by the code". The most common causes of these failures? "[I] inadequate installation, inadequate detailing, inadequate bracing, web flexing, lack of roll-over protection on supports, and failure to account for unbalanced loads on continuous members." Another common response was failure to account for snow accumulation. Observations made by experienced practitioners across the country agree with those in this book. What about wind damage? Damage from hurricanes can be local, as in Fig. 10:18, or involve destruction of the entire building. As a result of widespread building damage in the aftermath of Hurricane Andrew (August 24, 1992), the South Florida Building Code's wind resistance requirements—then already the most stringent hurricane code in the country—were further tightened. . Saffir18 states that some of the new requirements for prefabricated buildings include a ban on the use of cable for tension members (in line with our suggestions in Chapter 3), the use of metal cladding with a minimum thickness of 24 gauge, and reductions in allowable deflection criteria and anchorage of the doors on the building structure. After Hurricane Iniki struck the Hawaiian Islands in September 1992, the Hawaii Association of Structural Engineers prepared a damage assessment. The survey found that in addition to residential buildings, some prefabricated buildings were also affected. The report's findings include: "Prefabricated metal buildings appeared to suffer proportionately more damage than other types of engineering structures."19 Again, some building collapses were complete (Fig. 10.20). The troubling aspect of such metal construction failures is not that some of the buildings were overloaded - any structure can be - but that there were so few reserves of ductility and strength when the overload came. Of particular concern are reports of structural failure at sub-design loads. In general, however, properly designed metalwork systems must provide secure, solid protection that can withstand all structural loads required by the code without undue deformation. It should also be noted that metal building systems have good earthquake resistance. Problems are not likely to arise in well-designed metalwork systems, but rather in structures composed of poorly designed metalwork components.

10.9.8 Bad Design Bugs Some bugs have been attributed to improper design techniques. As in chap. 16 A collapse can occur right at the beginning of the erection of metallic buildings—the erection of the primary structure—when the

SOME COMMON PROBLEMS AND ERRORS 310

CHAPTER TEN

FIGURE 10.20 Total collapse of a metal building. (Photo: Association of Structural Engineers of Hawaii.)

Fitters are inexperienced or careless. Completed buildings can collapse during heavy snowfall or wind overload if their support systems are compromised in some way or improperly assembled. Construction-related construction defects can be missing, loose or improperly installed purlin supports, missing or loose ceiling and wall supports as well as missing connection elements (high-strength frame screws, wall screws, concealed clip fasteners, etc.) .). If there are extra parts left over after building the building, it's not a good sign. Damage to vital parts of the structure during construction, as shown in Fig. 10.9, facilitates future failure. Relatively common deficiencies related to the substructure are omitted, misplaced, too short tie rods or the wrong size and type. L-shaped tie rods used instead of prescribed head anchors are known to be made of concrete, as the author has seen. Tie rods were installed into the foundation piers with insufficient hook engagement and then pulled under load, allowing the building columns to spread under load. Only the vigilance of the supervisor and inspector limits the possibilities for design errors.

10.9.9 Damage from Aging As with any other type of construction, aging of structural elements leads to damage ranging from leaks in the roof to complete collapse. Damage to building components is obviously the most serious as it affects the strength and rigidity of the building. Corrosion of steel is perhaps the most well-known mode of deterioration of metallic structures. Corrosion can result from roof leaks, condensation and even groundwater intrusion. In a prefabricated industrial building by the sea, the author found the supporting foundations at the beginning of the harvest

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

311

Wall joists supporting light trusses completely rusted (Fig. 10.21). The deterioration was determined to be the result of repeated flooding during various storms and hurricanes that occurred over the building's long lifespan. The integrity of the steel was virtually nonexistent - to the point where the wall cladding would likely support the weight of the building! Structural damage caused by fire typically results in deformed primary structures and twisted secondary members. Since the thickness of the metal frame in prefabricated buildings is quite small, even a small fire can cause deformation and other serious problems. A small fire in Fig. 10.22 barely burned the roof and purlins but caused severe, albeit localized, damage to the main structure. As in Fig. 10.23, the heat deformed both rafter wings and caused the upper wing to detach from the web. Degradation of non-structural elements can have less dramatic but no less noticeable effects. In Fig. 10.24 the roof leak has been attributed to a damaged ridge which is as shown in Fig. 10.14. Where is the rubber part of the boot? Reportedly pecked by seagulls who thought it was delicious.

10.9.10 Other Causes of Failure Failures in the building can be due to poor maintenance of the building or even careless modifications made during its useful life. It is not uncommon for plumbers to remove primary structures.

FIGURE 10.21 The bases of the cold-formed wall joists that support the trusses in this early-season prefabricated building have been completely eroded by repeated saltwater intrusion.

SOME COMMON PROBLEMS AND ERRORS 312

CHAPTER TEN

FIGURE 10.22 Effects of a small fire on a metal building system. The left area is in Fig. 10.23. Also note the method of attaching the bar to the frame. (Photo: Chabot Engineering.)

Flange support located in the pipe run. (Also, remember our discussion in Section 10.7 about the dangers of hiring the sprinkler contractor without checking with the structural engineers.) Some homeowners wouldn't think twice about moving the wall mount from one enclosure to another—or even removing it entirely— if you want to add a door or window. In an older prefab studied by the author, someone cut off the entire underside of a curved internal rebar, presumably because it was in the way of a piece of equipment or piping (Fig. 10.25). This, of course, severely affected the lateral load-bearing capacity of the building.

LITERATUR 1. Alexander Newman, „Engineering Pre-engineered Buildings“, Civil Engineering, September 1992. 2. Joseph E. Minor et al., „Ausfälle von Strukturen aufgrund extremer Winde“, ASCE Journal of the Structural Division, vol. 98, ST11, November 1972. 3. „Architects and Engineers Must Understand Loads Exerted by Overhead Doors“, DASMA Technical Data Sheet #251, International Association of Door and Access Systems Manufacturers, Cleveland, OH, 2001. 4 Butler Roof Systems Design / Specifications Manual, Butler Manufacturing Co., Kansas City, MO, 1993. 5. Ken Buchinger, „Proper Design of Roof Penetrations in Trapezoidal Standing Roofs“, Metal Architecture, Juni 1994. 6. Krista Hovis, „Common Questions about Roof Penetrations Find eine Antwort“, Metal Construction News, August 2002, S. 84–85.

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

FIGURE 10.23 Primary structure damaged by fire. (Photo: Chabot Engineering.)

FIGURE 10.24 Damaged pipe cap allowing water to enter the building.

313

SOME COMMON PROBLEMS AND ERRORS 314

CHAPTER TEN

FIGURE 10.25 The lower part of this inner bracing curve was partially removed during previous modifications. This condition affects the lateral strength of the building.

7. Duane Miller und David Evers, „Loads and Codes“, The Construction Specificationr, November 1992. 8. John L. Ruddy, „Evaluation of Structural Concepts for Buildings: Low-Rise Buildings“, BSCE/ASCE Structural Group Lecture Series at MIT, 1985. 9. Kevin Westervelt, „Tips for Engineers Working on Prefabricated Metal Building Projects“, Structural Engineer, Juli 2001, p. 24. 10. Steel Building Systems Handbook, Steel Building Manufacturers Association, Cleveland, OH, 2002. 11. Frank Zamecnik, „Errors in the Use of Cold Formed Steel Members“, ASCE Journal of the Structural Division, Bd. . 106, Nr. ST12, Dezember 1980. 12. David B. Peraza, „Lessons from Recent Collapses of Metal Buildings“, Proceedings, 15th International Specialty Conference on Cold-Formed Steel Structures, St. Louis, MO, 19.–20. Oktober 2000. 13. Minimale Auslegungslasten für Gebäude und andere Strukturen, ASCE 7, American Society of Civil Engineers, Reston, VA, 1998.

SOME COMMON PROBLEMS AND FAULTS SOME COMMON PROBLEMS AND FAULTS

315

14. Erling Murtha-Smith, Howard I. Epstein, and Jason D. Mitchell, "Improved System Response to Abnormal Snow Loads," Final Report of MBMA Research Project 96-04, April 30, 1998. 15. "Prevention of Roof Collapse of Snow Loading ", Record, P.O. Box 9102, Norwood, MA, Fall 1994. 16. Source: Correspondence between the author and FM Global PR staff. 17. Wesley G. Britson and Emile W.J. Troup, "Structural Forum: Specialty Structural Engineers ' Structure, May 2003, p. 22. 18. Herbert S. Saffir, Discussion of article 'Residential Building Failures Caused by Hurricane Andrew', Journal of Performance of Constructed Facilities, August 1996, p. 137. 19. A Survey of Structural Damage Caused by Hurricane Iniki, September 11, 1992, Hawaii Association of Structural Engineers, Honolulu, HI.

REVIEW QUESTIONS 1 What are the concerns about specifying exposed structures in metal building systems? 2 Name the most common reasons for failure of metallic structures. 3 What are the challenges faced by designers of prefabricated buildings with complex configurations? 4 What are the two basic approaches to dealing with the weight of heavy sprinkler networks when the exact piping layout is not known in advance? 5 What is the possible side effect of a one-sided main frame bracing? 6 The fabrication supplier suggests using a single 8 inch cold formed channel. on each frame of a swing door measuring 20 feet by 20 feet. Is that acceptable? Why or why not? 7 What could be the problem with adding a roof frame after the roof is in place? 8 Explain what progressive collapse is. Why is it desirable to avoid?

SOME COMMON PROBLEMS AND ERRORS

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 11

LATERAL DEVIATION AND VERTICAL DEFLECTIONS

11.1 THE MAIN POINTS The discussion in Chap. 7 and 9 have already highlighted the importance of specifying correct design criteria for lateral drift and vertical deviation. This chapter specifically addresses this critical and controversial issue that worries many structural engineers specifying metalwork systems. First the definitions: lateral (soil) deflection is the lateral deflection between two adjacent floors of a building caused by lateral loads (wind and earthquakes) (Fig. 11.1). For a one-story building, the lateral displacement is equal to the amount of horizontal displacement of the roof. The horizontal deflection of a wall refers to its horizontal movement between columns under wind or seismic action. The vertical deflection of a floor or roof member is the amount of sagging under gravity or other vertical loading. Why is this controversial? Deflection and deflection criteria, as well as some other issues such as vibration, deal with the maintenance or functionality of buildings under load. Model building codes traditionally dictate desired levels of strength and safety, leaving the nebulous issues of satisfying occupant perceptions of comfort and solidity to the designers. Designers' criteria for achieving these goals are necessarily subjective, since a building that seems fragile to one may seem comfortable to another. Different design firms tend to set similar, though not identical, limits on the horizontal and vertical displacement of buildings for medium and tall structures. On the other hand, the rigidity requirements for prefabricated low-rise structures remain a mystery to many engineers. These squat structures were traditionally clad with flexible metal roofs and roofing that could tolerate large movements of the structure, and their maintenance was rarely a problem. Designers only became concerned when brittle wall materials such as masonry and concrete found their way into metal building systems. For these cases, some engineers continued to specify the same stringent deflection and deflection criteria used in conventional construction - only to be rejected by many manufacturers who denounced such rigidity requirements for metal buildings as unnecessarily expensive and impractical. Should metal structures with hard walls be given special privileges? Similarly, when a metal construction system touches masonry (Fig. 11.2), its lateral vibration must be controlled in order not to damage the fragile masonry. Alternatively, the two structures can be separated by the amount of combined lateral displacements expected by the building, but the most common approach seems to be to simply connect the two structures together, or to use the front solid wall for the metal building at the rear. . What about vertical deflections? Should a structure that supports flexible metal roofs have more lenient vertical deflection criteria than other lightweight structures – such as wood beams with asphalt shingles? As we shall see, the answer is not as simple as it might seem at first glance. Our journey through these emotional waters begins with the themes of transverse floor drift and horizontal wall displacement, and then continues with the theme of vertical deflection criteria. 317 Conveyed from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

LATERAL DEVIATION AND VERTICAL DEVIATIONS 318

CHAPTER ELEVEN

FIGURE 11.1 Ground Deviation from Lateral Loads.

FIGURE 11.2 In this common design, a masonry facade backs against an all-metal building system, raising concerns about appropriate lateral deviation criteria.

11.2 LATERAL DEVIATION AND HORIZONTAL WALL SHIFT 11.2.1 Should this be required by the regulations? Lateral drift and deflection limits are usually expressed relative to floor or wall height (H) as a floor deviation of H/400. Another commonly used term is the deflection or drift index, an inverse of the drift limit; For example, the deflection index for a deflection limit of H/400 is 0.0025.

LATERAL DEVIATION AND VERTICAL DEVIATION LATERAL DEVIATION AND VERTICAL DEVIATION

319

Drift limits characterize the solidity and robustness of the building. A stable building will cost more to construct than a flimsy one, but can have a higher resale value. Two buildings designed for the same structural loads but used for different purposes may have different stiffness requirements. For example, a high-tech research lab or hospital would likely impose much tighter limits on the building's allowable movement than a typical office. Many designers believe that horizontal deflection and drift criteria are related to the quality of building construction and should not be mandated by law. However, regulations may set a maximum drift limit to preserve the structural integrity of buildings and their fragile components. In fact, excessive lateral displacements due to the P-Delta effect can lead to unforeseen overloading of building columns; large movements can also damage the outer casing and inner lining. Because of this, the regulations are more concerned with controlling the building sway caused by violent earthquakes than by severe hurricanes. Strong seismic vibrations affect the structural and non-structural elements of the building, while wind forces act mainly on the outer shell. The extent of seismic movements generally exceeds those caused by wind. A building that has suffered an earthquake may suffer more internal damage than one that has been hit by a hurricane, directly affecting its usability. For example, after the 1995 Kobe earthquake, some buildings appeared relatively intact from the outside, but were still unusable due to warped partitions and doors stuck due to violent movement. Drift limits related to seismic loads are much softer than those related to wind. 11.2.2 Model Code Provisions for Seismic Loading Deviation Limits The deviation criteria listed below are found in the code sections dealing with seismic loading. The International Construction Code1, which is intended to eventually replace the three traditional model codes, lists seismic drift limits in Table 1604.3. Allowable historical deviations depend on building construction and seismic use group. For non-masonry buildings, the drift index ranges from 0.020 hsx for Group I seismic exploitation to 0.010 hsx for Group III seismic exploitation, where hsx is the height of the ground below level x. For buildings less than four stories without masonry walls, "with interior walls, partitions, ceilings and exterior wall systems designed to accommodate the variances of history". As noted below, such an arrangement may require an unusual level of detail and coordination between business and project teams. The 1997 Uniform Building Code, 2 Section 1630.10 requires that pavement displacements be calculated using the maximum inelastic response displacement ⌬M. The code limits the calculated floor drift to 0.025 h (where h is the floor height) for buildings with fundamental vibration less than 0.7 s and to 0.020 h for other buildings. The Code makes two important exceptions to this rule. The first states that the limit can be exceeded if it is shown “that greater deviations from either can be tolerated . . . and non-structural elements that may affect the safety of life”. The second exemption specifically exempts from any drift restrictions single-story steel-framed buildings used for factories, manufacturing, storage, commercial workshops and certain other occupations. To qualify for this exemption, a building must not have any equipment attached to the frame unless it is sufficiently detailed to account for the exemption. To avoid damage from large frame movements, the code also requires that walls laterally supported by the steel frame be designed to accommodate deflection. This objective is to be achieved through a stretch tolerance analysis and compliance with certain prescribed anchoring and wall connection requirements. The BOCA National Building Code of 19993 and the Standard Building Code4 contain similar but not identical provisions on seismic drift. Note that in previous editions of the code, seismic loading was usually expressed in terms of operational loads, but more recent codes treat it as a factored load. Under seismic loading, of course, none of the model codes attempt to impose any flexibility limitations on all-metal clad metalwork systems. Such conditions occur mainly in industrial buildings and warehouses. For other structures, the drift limit under seismic loading can be specified in the applicable building code. In any case, seismic loading and its drift limits rarely determine the design of prefabricated houses.

LATERAL DEVIATION AND VERTICAL DEVIATION 320

CHAPTER ELEVEN

11.2.3 Drift Limits for Wind Loads In light metal building systems, seismic loads rarely dictate the design of structures that resist lateral loads - wind loads usually do. Model building codes are silent on lateral wind drift limits, a fact that may reflect a lack of consensus on the subject and an understanding that such limits are related to build quality and should not be required by the codes. However, guidelines are available elsewhere. A lateral deflection limit of H/500 has been recommended for tall buildings since the 1940s.5 The authoritative Structural Engineering Handbook6 states that the commonly used range of deflection index is 0.0015 to 0.0035 (resulting in a deviation limit in the range of H /666 leads to H/286). It includes a Weiskopf & Pickworth Deflection Index guide that plots index values against the magnitude of wind loads and wind exposure. The Handbook notes that engineering judgment must recognize the economic values involved and that a speculative office building may be constructed with a less stringent tolerance limit than a corporate or prestige single-occupancy building. The Building Structural Design Handbook7 states that even in a 25-year storm, a deviation index of 0.0025 (H/400) “may be appropriate for a speculative commercial building. On the other hand, it could be totally unsuitable for a hospital, library or other high-value building project.” He goes on to suggest that the problem can be solved by setting a strict drift limit, say H/500, but that the Drift is calculated using a design wind load less than that imposed by a 50 year storm. For example, the load from a 10-year or 25-year storm can be used. A national survey of structural engineers from the ASCE Steel Structures Drift Control Working Committee of the ASCE Steel Structures Committee8 found that design practices related to wind drift vary significantly. However, most designers give drift indices of 0.0015 to 0.003 (corresponding to the limits of H/666 to H/333) caused by an average wind return interval of 50 years for all wind types. Again, the most commonly used wind drift limit for low structures is 0.0025 (H/400), caused by a 50-year wind. Incidentally, the Working Committee was of the opinion that wind drift limits should not be codified. A comment to Section B1.2 of ASCE 7-989 summarizes the Task Committee's finding that drift limits commonly used for building design are on the order of H/600 to H/400. ASCE 7 indicates that lower drift limits may be appropriate for brittle coatings. This suggests that an absolute limit of deviation may be required as some bulkheads, trim and glass can be damaged by deviations greater than 3Ⲑ8 inches unless special details are used to accommodate the movement. To calculate drift, the Commentary suggests using 70% of the operating wind load calculated by ASCE 7 methods.

11.2.4 Deviation Limits in AISC Design Guide #3 Recognizing the lack of maintenance criteria for metalwork systems subjected to wind loading, MBMA and AISC have published a design guide entitled Maintenance Design Considerations for Low-Slage Buildings.10 The guide's respected authors, James M. Fisher and Michael A. West have gone to great lengths to stimulate discussion on a variety of maintenance issues, including drift and deflection. The guide should be read by anyone involved in low-rise structural design. Reflecting the subjective nature of maintenance criteria, the guide's authors base many of their recommendations on their own judgment and experience. They acknowledge that the criteria are controversial and view the guide as a catalyst for debate rather than the final word in discussion. (However, some fabrication fabricators seem to think just the opposite - that there is no longer any doubt.) The guide uses a wind speed load with a mean repetition interval of 10 years for its drift limit criterion instead of the 50 annual charge used for strength calculations. The reasoning is that 50-year storms are rare events that have little to do with the everyday experience of buildings. Additionally, the consequences of maintenance outages are “non-catastrophic” and must be weighed against the high initial costs required to prevent outages. The guide states that 10-year wind pressure can be reasonably approximated using 75% of 50-year wind pressure values. (Some other sources have also questioned the common practice of basing wind drift calculations on wind loads likely to occur only once in 50 years. Galambos and Ellingwood11 for example, downloaded from the Digital Engineering Library @ McGraw-Hill (www .digitalengineeringlibrary .com ) Copyright © 2004 The McGraw-Hill Companies All rights reserved All use is subject to the terms of use posted on the website.

LATERAL DEVIATION AND VERTICAL DEVIATION LATERAL DEVIATION AND VERTICAL DEVIATION

321

advocate the use of a reference period of 8 years, which is the average length of an office building lease.) For different types of walls, the Guide proposes certain upper limits for the amount of lateral movement of the barframe, horizontal deflection and shelving (lateral movement parallel to the wall). Below are the criteria for primed coatings; The guide also considers criteria for column-supported panels and timpani. In the following expressions, H stands for the height of the wall and L for the length of a supporting steel element. The recommended maximum deviation for different materials is H/60 to H/100 for metal panels H/100 for precast concrete H/200 for reinforced masonry (can be reduced to H/100 with proper detailing) If using interior walls, bare frame Story- Drift is capped at H/500. The maximum recommended horizontal deflection of joists or wind columns supporting metal or masonry walls is L/120 for metal panels and L/240, but no more than 1.5 inches for masonry walls. A limit of H/ 500 is recommended for columns and parapets Curtain walls. Again, all of these criteria apply to a 10-year wind load. The lateral drift and horizontal deflection constraints suggested by the guide are more generous than those suggested by other sources. Some engineers find it counterintuitive that the guide appears to provide a greater level of protection for interior drywall partitions than for fragile exterior walls. The guideline drift limits are reprinted in the MBMA Metal Building Systems Manual.12

11.2.5 How lateral deflection is calculated Before discussing the various criteria listed above, it is necessary to briefly review how lateral and horizontal deflection are calculated and what the numbers actually mean. The total course deviation is a sum of two components - the frame deviation and the displacement of the diaphragm between frames (Fig. 11.3). For a typical prefabricated building with rigid structures spaced 20 to 30 feet apart and a horizontal pole roof support, the deflection component of the membrane can be negligible. At the other extreme, in buildings without roof reinforcement, where the wind load is distributed to the frames via eaves braces, the membrane deflections can be greater than the frame displacement. Unfortunately, membrane deflection calculations are occasionally overlooked by some metal structure designers. The actual frame deviation can be easily determined by most ready-made design software. Any general statics computer program can be used for preliminary calculations. The approximate formula in Fig. 11.4 can be useful for rough checks of double-hinged frames with constant bar sections. Of course, in the case of rigid frames with conical supports and beams, the process is much more complex, then computers are required. 11.2.6 Lateral Deflection Under Gravity Loading A discussion that focuses only on lateral deflection due to wind or seismic loading misses an important point: Lateral displacement of the frame can be caused not only by lateral loads, but also by gravity loads. Many structural engineers accustomed to conventional building designs do not realize that a gable structure can have significant "ejection" at roof level when loaded with snow or live load from the roof (Fig. 11.5). Lateral displacements at the knees of the structure due to large snow loads can exceed wind-induced ground deviations. The codes do not address the problem, probably because gable frames are largely endemic to metal building systems. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

LATERAL DEVIATION AND VERTICAL DEVIATION 322

CHAPTER ELEVEN

Figure 11.3 Story drift components.

FIGURE 11.4 Approximate formula for calculating the frame deviation for a double-hinged frame with a constant cross-section of the beam.

We recommend that for all service load combinations that include traffic, snow and wind loads, there is a specified building tolerance limit. It could be argued that moderate and permanent collateral loads applied prior to the installation of fairings, equipment and finishes attached to the structure need not be included in the drift calculations. Seismic loading drift can be assessed separately using the relatively mild Code criteria mentioned in Section 11.2.2. Excessive structural distortion from snow is real and should not be ignored. Ruddy13 tells of two prefabricated buildings that were used as a school and office and had false ceilings. after a few, after some

LATERAL DEVIATION AND VERTICAL DEVIATION LATERAL DEVIATION AND VERTICAL DEVIATION

323

FIGURE 11.5 Lateral displacement of rigid frame under gravity loading.

Accumulated snow on the roofs, the roofs of both buildings were noticeably shifted. Local builders called in insisted that both buildings be cleared, the snow removed and the roofs re-tested for strength. After such an incident, building designers tend to use fairly strict displacement limits for future designs. Lateral movement caused by gravity is less pronounced in multi-span rigid frames than in their single-span brethren, and is virtually absent in truss and cone beam systems.

11.2.7 Lateral Deviation and Deviation: The Discussion Are major medical history deviations always harmful? As far as designers are concerned, not only is an absolute value of historical drift, but also an angle of curvature that the wall takes on. In the theory of elasticity, the greater the slope of a curved shape of a flexure, the greater the stresses. Brittle materials such as unreinforced masonry and glass can easily break under high loads; ductile materials such as reinforced masonry and concrete can tolerate some cracking without failure. However, large cracks, particularly in plain masonry and concrete walls, are likely to become avenues for water ingress, which can damage interior finishes and accelerate freeze-thaw wall deterioration. The total degree of wall curvature depends on the magnitude of three deflection components (Fig. 11.6): 1. Floor deflection, a sum of structure deflection (Df) and membrane deflection Ddiaph 2. Horizontal deflection of deflection beams support and wind columns, if present, Dchord 3. Horizontal deflection of the wall itself Dw Of these, the first two depend on the stiffness of the metallic structure and the third is a function of the stiffness of the wall. The second component, dgirt, occurs only when the intermediate chords provide lateral support to the wall, as is often required for tall steel beam and brick veneer walls. In this case, the points where the curvature of the wall changes - turning points - are located near the circumferential locations (Fig. 11.6a). Apparently, the horizontal deflections of walls running from the foundation to the roof with no intermediate beams, such as B. Full-height CMU or prefabricated CMU, no dgirt (Fig. 11.6b). The critical question in this discussion is whether the wall functions as a simply supported or a continuous element. Stresses and deflections in simply supported beams are not affected by the movement of the columns. Tension bars, on the other hand, are statically indeterminate and are statically influenced by elastic supports. The ends of each simply supported element rotate freely. Therefore, a wall can only be considered simply supported if its ends are free to pivot at the base and ceiling. The maximum horizontal deflection of a simply supported wall can be taken as Dw and not Dmax as the movement of the columns is irrelevant. However, if the wall is attached to the ground - an example is a CMU wall attached to the foundation - it will prevent ultimate rotation at the base, but the single span model will not

LATERAL DEVIATION AND VERTICAL DEVIATION 324

CHAPTER ELEVEN

FIGURE 11.6 Components of total horizontal wall displacement: (a) with struts; (b) without straps.

does not apply and Dmax should be taken as the maximum displacement. (Walls are rarely fixed at the top, of course.) It is evident that the wall's susceptibility to cracking is influenced more by its own rigidity and detail at the base than by the magnitude of the deviation from history. But is it possible to make CMU walls behave like simple beams with dowels? For reinforced CMU, AISC Guide No. 3 suggests that instead of a common CMU basic detail of Fig. 11.7a, a detail similar to Fig. 11.7b is used to facilitate the final rotation. In Fig. 11.7b, at the base, a continuous transverse sheet metal skin with mastic is installed around the vertical bars to introduce a plane of weakness into the wall and allow for rotation of the end. However, it seems that the CMU with vertical bars and pins still develops a significant moment of fixation at the base (Fig. 11.7c), which challenges the whole theory of fixed base CMU. A base rotation can be allowed if the dowels are omitted but the ridge is retained. Unfortunately, such a detail is likely to result in a wall without adequate "hold" to the foundation, a wall that can shift under lateral loading. More specifically, the dowels are retained, but the length of the dowel above the ridge is encapsulated in a knockout sleeve that allows the dowel to slide into the wall (Fig. 11.7d). In this case, the dowels do not withstand tensile forces, but are able to transmit shear. Another important topic in this discussion concerns the construction of corners. A wind-exposed wall whose apex is attached to the eaves moves with the frame, while the vertical wall does not (Fig. 11.8). Placing a control joint in the wall near the corner is intended to prevent the wall from cracking. However, the hinge cannot survive large rotations of the wall without failure and leakage. A problem with excessive deflections of the outer walls perpendicular to the direction of lateral loads is important, but so is stacking of walls parallel to the load (Fig. 11.9). For example, shelving affects internal drywall partitions that are attached at their top to the main structure of the building. A drywall partition can undergo significant deflection perpendicular to its surface, but is susceptible to displacement along its plane. Although it is possible to overcome this problem using special "sliding" connections above

LATERAL DEVIATION AND VERTICAL DEVIATION LATERAL DEVIATION AND VERTICAL DEVIATION

325

FIGURE 11.7 Overcoming the fixation at the bottom of the masonry wall with pegs: (a) a common construction detail; (b) introducing a plane of weakness to facilitate rotation; (c) forces resisting rotation and providing strength; (d) a possible detail of the true "pin" connection.

FIGURE 11.8 Sagging shape of the wall near the corner of the building.

B. the partition, such connections are best designed by structural engineers, who unfortunately rarely deal with architectural details. Exterior masonry and concrete walls in metal buildings also suffer from shelving. These "hard" walls are much more rigid than any wall bracing that might be along the same column line; They tend to act as shear walls, rendering bracing ineffective. Unless they are completely isolated from the movement of the structure - a rare scenario - these walls should be intentionally designed and reinforced as shear walls rather than wall braces.

LATERAL DEVIATION AND VERTICAL DEVIATIONS 326

CHAPTER ELEVEN

FIGURE 11.9 Wall shelf.

11.2.8 Choosing a Lateral Deflection Limit Of course, the lateral deflection and deflection criteria depend on many project-specific factors, including the type of occupancy; presence of partitions, cranes and equipment supported by the structure; type of exterior wall materials; and owner expectations. For “all” metal buildings with no bulkheads, frangible exterior walls, or frame-mounted equipment, floor displacement limits in the range of H/60 to H/120 may be appropriate. For exterior masonry or concrete walls attached to steel structures, it makes sense to use traditional building deck drift controls, such as e.g. B. H/200 for seismic loading and H/400 for wind. Some cases require tighter drift and transmission limits. Figure 11.10 shows a building with a glass brick facade—one of the most fragile building materials—for which a strict tolerance limit has been set. Note that to minimize shelving build-up, the designers specified wall reinforcements made of steel angles instead of rods or cables. Slightly more liberal limits for wind induced drift (H/200 to H/300) may be warranted where excellent collaboration between architects and engineers allows for the development of custom construction details for the base, top and corners of exterior and interior walls. The specialist knowledge, experience and quality of the supervision of the contractor also play a major role. The presence of drywall without custom connections - with any type of exterior wall - as in Fig. 11.11 limits the project variance to H/500. Some custom details of partition-purlin connections are shown in Fig. 11.12. These details include a custom-made oversized "idler bar" (runner) that accommodates the separator pins. The pins are inserted but not directly attached to the deflection rail, so that an expansion space remains at the top, the size of which is determined by the vertical deflection criterion of the purlins. So that the cleats do not rotate within the deflection path, they are reinforced on the side with a solid bar. The bridge is installed at close intervals, e.g. B. 5 feet on center, in areas where the beams are not reinforced by cladding on both sides (e.g. above the ceiling). If the bulkhead runs parallel to the purlins, the diverter rail is attached to spaced thin steel pins or similar elements that run between the purlins (Fig. 11.12a). Note that in this case the corners of the custom deflection rail are not square to account for the roof pitch. If the partition wall is at right angles to the purlins, a corner rail with square corners can be used. The running rail can be installed solely via the purlin distance or, in the case of higher strength requirements, a light steel pin can be added (Fig. 11.12b). An example of a custom detail for the partition running under a primary frame is shown in Fig. 11.13. Here, too, the special rail allows the frame to deflect under vertical load without load transfer to the dividers. The detail shows how to fix the plaster covering on the upper part: it must not be fixed on the deflection track, so as not to fill the room for movement and impair its function. Figures 11.12 and 11.13 largely decouple the partition from roof movement parallel to its plane, thereby drastically reducing, if not eliminating, the shear forces acting on the plasterboard. Therefore they could allow the use of a milder deviation threshold than H/500.

LATERAL DEVIATION AND VERTICAL DEVIATION LATERAL DEVIATION AND VERTICAL DEVIATION

327

FIGURE 11.10 The presence of the glass block in this metallic structural system justifies the building's stringent drift and transmission limits. Note the stiffening of the inner wall from steel angles. (The visible external support on the right is purely cosmetic.) (Photo: Maguire Group Inc.)

The decision to base these drift limits on 50, 20 or 10 year wind loads is at the discretion of the specifying engineer. While recommendations favoring a 10-year wind load make sense, we hesitate to adopt them for all conditions. It may be easier to adjust the deviation threshold so that calculations are only performed once. Incidentally, the H/500 drift limit calculated from a 10-year wind corresponds to the known H/400 drift limit from a 50-year wind.

11.2.9 Selection of a horizontal deflection limit Horizontal deflection limits for some common exterior wall systems are given in Chap. 7. Building regulations may include requirements for maximum allowable horizontal deflections of walls. For example, the International Building Code1 contains a table of deflection limits. It requires exterior and interior soft-faced walls to be designed for a maximum horizontal deflection of L/240, while soft-faced walls are double that – L/120. A footnote treats the secondary wall elements that support the metal cladding to be milder: L/90. All deflections in the IBC table can be calculated using a wind load equal to 70% of the declared load for "Components and Paneling", avoiding the debate on which wind load to use. Where exterior walls rely on intermediate beams for lateral support, horizontal deflection criteria are influenced by wall construction details. As previously shown, a wall that is free to pivot on its base can tolerate greater deflection than a wall with a fixed base. For the former, the provisions of AISC Guide n. 3, which limit chord deflections to L/120 for metal panels and L/240 (but not more than 1.5 inches) for masonry, seems reasonable. For the latter, a stricter limit is justified.

LATERAL DEVIATION AND VERTICAL DEVIATION 328

CHAPTER ELEVEN

FIGURE 11.11 Very common practice: the internal bulkhead is attached directly to the purlins with no provisions to accommodate vertical deflections. The partition wall must behave like a load-bearing wall and can fail when the vertical roof load is applied.

11.3 VERTICAL DEVIATIONS 11.3.1 Model Rules The criteria for vertical deflection are less controversial than those for lateral drift: everyone knows that the sight of a slanting bar overhead is not a good one. In addition to this visual effect, uncontrolled vertical deviations can cause damage to interior walls, windows and drywall ceilings. Large deflections of low pitched roofs can be dangerous. Examine the in Fig. 11.14 where a small roof pitch (1Ⲑ4:12) is not sufficient to compensate for purlin deflection under high snow load. Deflection of the inner first third caused by snow and possibly compounded by the impact of heavy overhead pipes or other objects can cause localized flooding. Look at the numbers. Assume the purlins are designed for a common vertical deflection limit of L/150, the first interior purlin is 5 feet from the vertical immovable outline, and the purlins are 25 feet long. The maximum allowable purlin deflection is (25 ft) (12 in.) ⌬max ⫽ ᎏᎏ ⫽ 2 in 150 Note that this figure does not include purlin deflection of overhead pipes or other overhead members, nor does roof deflection between purlins. Also, if the thirds are not correct

LATERAL DEVIATION AND VERTICAL DEVIATION LATERAL DEVIATION AND VERTICAL DEVIATION

FIGURE 11.12 Examples of custom details on steel stud partitions: (a) partition parallel to purlins; (b) Division perpendicular to thirds.

329

LATERAL DEVIATION AND VERTICAL DEVIATION 330

CHAPTER ELEVEN

FIGURE 11.13 A custom detail for a drywall that runs under the main structure allows for vertical deflection of the structure.

FIGURE 11.14 A flat roof pitch, such as 1Ⲑ4:12, may be insufficient for adequate drainage when purlins deviate under load.

supported, they tend to rotate under load. As in chap. 10, their actual vertical deflections will be greater than those predicted by calculations that neglect any reduction in purlin stiffness due to torsion. On the other hand, the change in roof height is only ⌬pitch ⫽ (5 ft) (1⁄4 in/ft) ⫽ 1.25 in ⬍ 2 in. Therefore, the roof pitch is insufficient to prevent local lake formation. As the reader can easily understand, to make ⌬slope at least equal to ⌬max, the roof slope has to be increased to 1Ⲑ2:12 or the strictest purlin deflection limit of L/240 has to be used. We prefer to consider both the additional deflection of the suspended pipe purlins and the deflection of the roof between the purlins.

LATERAL DEVIATION AND VERTICAL DEVIATION LATERAL DEVIATION AND VERTICAL DEVIATION

331

Why is this topic so important? The accumulation of meltwater and refreezing in this area can not only cause leaks, but also significant ice loads that can overwhelm purlins. As in chap. 10, in some severe cases this can lead to the collapse of the entire building. The author studied collapses of two metal buildings where this phenomenon was identified as one of the main causes of their failure. Over the centuries, builders and designers have come to the conclusion that a deflection limit of 1Ⲑ360° of bar length (L/360) is sufficient to prevent cracking in plastered ceilings. The deflection limit applies to live loads, snow or other superimposed loads acting after the roof structure. This criterion has been largely adopted by the building regulations. In the absence of plastered ceilings, less stringent limits such as L/180 have traditionally been applied. The deflection provisions of the International Building Code1 are representative. According to Table IBC 1604.3, roof panels supporting gypsum ceilings must comply with the L/360 limit; those that carry unplastered ceilings, L/240; and those not wearing blankets, L/180. These limits apply under traffic, snow or wind loads (equivalent to 70 percent of the specified load for "Components and Panels"). An exception is secondary beams supporting metal roofs formed without any other roofing material - these purlins need only meet the L/150 criterion under live load. Its deflection under snow load is probably still limited to L/180. What about combined dead and live stress deflections? Table IBC 1604.3 specifies the L/240 limit for roof panels supporting gypsum ceilings; L/180 for ceilings without plaster; and L/120 for those not wearing blankets.

11.3.2 Other recommended criteria The AISC14 specification limits the maximum allowable overload deflection of plaster-bearing roof and ceiling panels to L/360. The MBMA Metal Building Systems Manual12, in its Maintenance section, reproduces some of the provisions of ASSC Design Guide No. 3. The guide lists deflection limits for various elements of roof construction, including those required to meet drainage and puddle considerations. For example, it recommends the well-known deflection criteria of L/360 for roofs supporting gypsum ceilings and L/240 for roofs supporting other roofs. The guide notes that some “maximum absolute value consistent with ceiling and bulkhead detailing should also be used,” and suggests a range of 3Ⲑ8 inches to 1 inch. The guide also recommends purlin deflection testing under a combination of dead snow and half the design load (or a minimum of 5 lb/ft2) to ensure positive drainage is still present when members deflect under load. These guide criteria are based on the payload of the project or 50 years of snow. The guide states that the above deflection criteria are most important along the perimeter of the building and that the maximum purlin deflection in the roof pitch can be limited to L/150 due to snow loading. Presumably the last number only applies where there are no ceilings or partitions. The guide makes an important point that localized deflections from concentrated loads are likely to be of greater concern than those from uniform loads. In fact, a common complaint from users of prefabricated buildings is that a light fixture or suspended pipe on a purlin deviates the purlin too much from its neighbors. In our opinion, the best protection against such high localized deflections, apart from sizing each purlin for each minute of loading - an impractical task - is to use stiffer purlins throughout (and a generous side load allowance). This means that deflection limits tighter than L/150 are used throughout the roof. ASCE 79 includes Appendix B with comments addressing maintenance issues. Section CB.1.1 states: deflections of approximately 1/300 of the span. . . are visible and can cause general architectural damage or leaks. Deflection greater than 1/200th of the span can affect the function of moving components such as sliding doors, windows and partitions.

LATERAL DEVIATION AND VERTICAL DEVIATION 332

CHAPTER ELEVEN

As AISC Design Guide #3, ASCE 7 suggests that damage to unloaded bulkheads can occur when vertical deflections exceed approximately 3Ⲑ8 inches unless special detail is used. Otherwise, ASCE 7 does not cover the complexity of establishing vertical deflection criteria on metal roofs. What vertical deflection criteria do we recommend? In our opinion, the L/150 criterion is too generous for finished surfaces. At a typical span of 25 feet, that translates to 2 inches of deflection under the design's snow load — a noticeable drop that can alarm occupants. We also believe that in northern climates this criterion should not be used even for roofs without decks but with very low pitches to avoid ice accumulation near the eaves. Traditional deflection criteria L360 and L240 for plastered and bare ceilings respectively should be adequate for most finished spaces. As in Sect. 11.3.1, the vertical deflection limit of L/240 must also be sufficient for uncovered metal roofs if the roof pitch is very small and/or the building is in the north. In addition, the use of this boundary increases the value of the building by allowing later completion without incident. For buildings where all three of the following conditions are met, the L/180 limit may be acceptable: (a) the snow load does not dictate the design, (b) the roof pitch exceeds 1Ⲑ2:12, and (c) the future A ceiling mount is extremely unlikely. To avoid misunderstandings with current and future homeowners, it may be a good idea to write to homeowners advising that the design of the building will not allow for future drop ceiling installations unless future designers use special details to accommodate vertical deflections. It should be noted that the L/240 or L/180 deflection limit is more stringent than most manufacturer standards: some still design purlins with an L/120 limit for snow or overload. If desired, strict deflection limits (and those for lateral deflection) should be specified in the contract documents and verified by checking the manufacturer's specifications.

LITERATUR 1. Internationaler Baukodex, International Code Council, Falls Church, VA, 2000. 2. Einheitlicher Baukodex, Internationale Konferenz der Baubeamten, Whittier, CA, 1997. 3. Der BOCA National Building Code, Building Officials & Code Administrators International , Country Club Hills, IL, 1999. 4. Standard Building Code, Southern Building Code Congress International, Birmingham, AL, 1996. 5. „Wind Bracing in Steel Buildings“, Final Report, Subcommittee 31, Committee on Steel, Structural Division, ASCE-Transaktionen, Bd. 105, S. 1713–1739, 1940. 6. Edwin H. Gaylord, Jr., Charles N. Gaylord und James. E. Stallmeyer (Hrsg.), Structural Engineering Handbook, 4. Aufl., S. 25–28, McGraw-Hill, New York, 1997. 7. Richard N. White und Charles G. Salmon (Hrsg.), Building Structural Design Manual, p. 583, Wiley, New York, 1987. 8. „Steel Drift Design of Steel-Framed Buildings: State-of-the-Art Report“, vom ASCE Task Committee on Drift Control of Steel Building Structures des Design Committee of Steel Building Structures , ASCE Journal of Structural Engineering, vol. 114, Nr. 9, September 1988. 9. Minimum Design Loads for Buildings and Other Structures, ASCE 7-98, American Society of Civil Engineers, New York, 1998. 10. Functionality Design Considerations for Low-Slender Buildings, Steel Design Guide Series No. 3, AISC, Chicago, IL, 1990. 11. Theodore V. Galambos und Bruce Ellingwood, „Funktionsgrenzzustände: Durchbiegung“, ASCE Journal of Structural Engineering, vol. 112, Nr. 1, Januar 1986. 12. Metal Building Systems Manual, MBMA, Cleveland, OH, 2002. 13. John L. Ruddy, „Evaluation of Structural Concepts for Buildings: Low-Rise Buildings“, BSCE/ASCE Structural Group Lecture Series am MIT , Cambridge, MA, 1985. 14. Specification for Structural Steel Buildings, Allowable Stress Design and Plastic Design, AISC, Chicago, IL, 1989.

LATERAL DEVIATION AND VERTICAL DEVIATION LATERAL DEVIATION AND VERTICAL DEVIATION

333

QUESTIONS FOR CONTROL 1. Explain the differences between floor deflection, horizontal wall deflection and vertical deflection. 2 Why have building codes historically avoided imposing drift limits on buildings subjected to wind loads? 3 List at least two types of exterior wall materials that can be considered base-fixed. 4 What are the dangers of applying generous vertical deflection criteria to buildings with suspended ceilings in snowy regions? 5 How high is the AISC Design Guide No. 3 for buildings with internal walls attached to the structure? With which wind load should the drift be calculated? 6 Can a Reinforced Concrete Masonry (CMU) wall that is fixed at the base to the foundation wall be considered pinned at the bottom? Explain why or why not. 7 Explain the potential problem of setting a 1L/150 vertical deflection limit for a building in a snowy region with a 1/4:12 pitched metal roof.

LATERAL DEVIATION AND VERTICAL DEVIATIONS

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 12

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS

12.1 INTRODUCTION Contrary to the dreams of builders, foundations do not come pre-packaged with metallic building systems. The concept of one-stop responsibility for prefabricated buildings is put into perspective by the fact that foundations are usually designed by external engineers. In this chapter we examine the differences between foundations for metal construction systems and those for conventional construction and explore some common solutions. We will not cover the basics of foundation design, a subject that should be familiar to every civil engineer and hopefully most architects. Likewise, we will not delve into the complex issue of determining allowable bearing pressures for different soils, a task best left to geotechnical engineers. As in chap. 9, the poor soils at the site can require deep foundations, an expensive item that can break the project budget and suddenly make a competing site much more attractive. On the positive side, an experienced geotechnical engineer may be able to justify a much higher allowable ground support value than could be deduced from the necessarily conservative tables of assumed support pressures in the building codes. This recommendation can result in significant cost savings. As Ruddy1 noted, “Increasing the allowable bearing pressure from 3 ksf to 6 ksf can result in savings of $0.08/s.f. to lead. for a shallow foundation system in a one-story installation.”

12.2 SOIL RESEARCH PROGRAM The results of a geotechnical survey program are of interest to all parties to the construction project. The building owner and local official must have reasonable assurance that the proposed project can be carried out safely without endangering the occupants of the building and adjacent property. The survey engineer needs to know the soil type, stratification and groundwater location to determine the most appropriate foundation type and support depth. The contractor needs much of the same information when selecting a ditch support system as when determining drainage needs. All participants are curious to see whether unsuitable material such as organic sludge or peat was found. A ground reconnaissance program can reveal the presence of abandoned foundations, buried utilities, and occasionally an archaeological site. Any of these "findings" could adversely affect the cost and schedule of the project. Exploration of the subsoil usually includes several soil boreholes or test pits, the number, type and location of which are determined by local regulations and experience. The BOCA National Building Code2, for example, requires at least one auger per 2500 ft2 of building area for buildings over 40 ft tall or more than three stories supported by mats or deep foundations. Common practice requires drilling holes in each corner of the building, one in the center and the rest nearby if necessary.

335 Conveyed from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 336

CHAPTER TWELVE

critically burdened foundations. For large buildings founded on poor soil, the burrows should be no more than 50 feet apart and perhaps much closer together. How deep should you carry the holes? The BOCA Code requires holes to be drilled to rock or to "a reasonable depth below bearing strata." For low-rise buildings, some engineers specify a hole depth of 20 feet below the expected foundation level, with at least one hole continuing deeper, perhaps even less than 100 feet, the smallest building dimension or denial. If the longer drill does not find unsuitable materials at the bottom, the remaining drills can be stopped at the originally planned depth. A hole should never be drilled in an unsuitable material. Because the length – and cost – of the program can change dramatically during field operations, it's a good idea to have a competent technician on site to observe the process and modify it if necessary. Instead of drilling holes in or above the ground, test pits are sometimes dug. Test pits are particularly useful for lightly loaded foundations supported by good soil at some depth but questionable material near the surface. A test well can provide a clear visual picture of the bottom condition at a shallow depth of up to about 10 feet. Test pits are relatively inexpensive as no special equipment is required and are often useful in supplementing the information provided by drilling the ground. For example, a rejection encountered by the rig could be indicative of a ledge or large boulder. A test well can quickly give an answer. The end result of the subsoil investigation is a subsoil investigation report prepared by the geotechnical engineer. The report describes the soil conditions at the site and recommends the maximum allowable bearing pressure and other relevant foundation properties.

12.3 WHAT MAKES THESE PRINCIPLES DIFFERENT? Three main factors distinguish prefabricated building foundations from others: significant horizontal column reactions, large column uplift, and a common need to design foundations before column reactions are determined. Experienced structural engineers can easily spot improper foundation design, as the first two problems are often overlooked by the uninitiated.

12.3.1 Horizontal Column Reactions Lateral forces such as hurricanes and earthquakes act on all buildings and result in vertical and horizontal column reactions. In "conventional" buildings with moderate floor area and relatively closely spaced columns, horizontal reactions are distributed across multiple column and wall foundations. It is a rare case when the columnar foundation has to withstand large horizontal loads. The situation is quite different with prefabricated buildings. The rigid frame, a cornerstone of metal structural systems, generates large horizontal shear from gravity loads (Fig. 12.1a) as well as horizontal reactions from lateral loads (Fig. 12.1b). Assuming that the frame columns are pinned, the column reactions in a typical foundation are shown in Fig. 12.2a. In the case of fixed base columns, the clamping moment M (Fig. 12.2b) is added. Horizontal column reactions tend to produce two types of foundation failures - overturning and sliding - which are detailed in Section 12.5.3.

12.3.2 Buoyancy Buoyancy – an upward force – is the natural result of wind action on gabled buildings (Fig. 12.3). In conventional two-story and multi-story buildings, wind uplift rarely exceeds the combined permanent roof and floor loads and therefore almost never dictates foundation design. One story metal building

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

FIGURE 12.1 Column reactions of a rigid frame structure: (a) from gravity loads; (b) lateral loads.

FIGURE 12.2 Forces acting on foundations supporting rigid frame columns: (a) pin base columns; (b) Columns with a solid base.

337

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 338

CHAPTER TWELVE

FIGURE 12.3 Height and reactions of horizontal columns caused by wind action in gabled buildings.

Systems, on the other hand, have extremely light roofs, totaling only 3 to 5 psf, which are often not nearly heavy enough to prevent lifting. Fortunately, not only the constant roof load but also the weight of the foundation and the overlying soil can counteract the erection of the column (Fig. 12.4). So, rather than adding weight to the roof, it is better to weigh down the foundation, or even better, the ground above - the most economical way to do this is to lower the foundation. While this solution requires additional excavation and backfill and is therefore not “very cheap”, it is typically more cost-effective than increasing the foundation footprint. In Fig. 12.4 the heave U is counteracted by the weight of the soil W1 and W2 and the weight of the foundation W3. To lift a cohesive soil like clay, the lifting force must first overcome its shear strength; The earth fault plane is generally inclined from the vertical. In soils without cohesion, e.g. B. sand, the fault plane is close to the vertical line. Most engineers use a conservative approach, neglecting the inclined soil segment and any soil shear strength.3 If included, the inclination angle may be assumed to be 30° for cohesive soils and 20° for non-cohesive soils according to Department of the Navy criteria. 4 Previous editions of model building regulations required a minimum safety factor of 1.5 against wind uplift, but modern regulations are less clear on the need for a safety factor. In areas at risk of flooding, the permanent "beneficial" load on the foundation is reduced by the rising water pressure. It is not unthinkable that a flood and a hurricane could occur simultaneously, although the likelihood of such an event must be carefully assessed. In the case of deep foundations, additional heave resistance can be mobilized using friction piles; Foundations at the edge can be anchored in the rock with drill rods.

12.3.3 Foundations Designed Before Construction In conventional construction, including single-story buildings, the structural design process typically follows a load path from the roof to the foundations. The load reactions determined for the upper structure are transferred to the lower bars and finally to the foundations, which are among the last elements designed. In the case of prefabricated buildings, the situation is reversed. Unless the owner deals directly with the fabricator in a captive relationship, the project will likely require the creation of a full set of contractual documents. This applies in particular to projects with public funds. Contractual documents usually contain information about the metal structure and its foundations, which means that the foundation design must be prepared before the fabricator performs an analysis of the structure and develops column reactions. To add fuel to the fire, some developers are insisting that a set of "founding treaty" drawings be released before the rest of the documents are ready.

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

339

FIGURE 12.4 Development of lifting resistance.

One way or another, foundations may need to be designed before final building responses are obtained from the selected manufacturer. This unfortunate situation is the bane of structural engineers specifying metalwork systems, who often must design foundations based on mere estimates of column reactions. To mitigate the impact of a possible foundation redesign, some engineers choose to include a notice in the contract drawings stating that the foundation design is provided for tendering purposes only and that the actual foundation design is to be provided by the contractor with similar detail. This, of course, introduces another part of the project.

12.4 HOW TO ESTIMATE THE SIZE OF THE COLUMN REACTIONS 12.4.1 Manufacturer's Tables The values of the column reactions for a symmetrical building of standard width, roof pitch and eaves height can be taken directly from the manufacturers' design manuals. The manuals usually provide column reaction tables depending on the primary frame type, dimensions and building load. Typical tables for buildings with one, two, three and four equal span gable frames are given in Appendix D. They all assume a roof pitch of 1:12, which largely corresponds to our recommendations. Load tables for a conical truss system are shown in Figure 12.5 but note that they are based on the outdated 1986 edition of the MBMA manual and should be used with caution. (In general, manufacturer tables are inevitably based on previous releases of code.) For other types of framing, other sources should be examined, some of which are described below in descending order of the accuracy of the results they can provide. Those who follow an approximate method should be warned that the final responses given by the manufacturer are based on the actual sizes of the tapered rods and therefore differ from the results of the simplified analysis, which assumes that the cross-sections of the rod are constant. In fact, even the column responses provided by different manufacturers for identical structures are not always the same, reflecting slight variations in the software, design assumptions, and the actual construction of the structure.

12.4.2 Specialized Software A design firm that is often involved in estimating responses may consider investing in specialized software for metal building design and analysis. This type of software essentially duplicates the manufacturers' design process; It is particularly useful when column responses and element sizes need to be known in advance. Any of the many programs available on the market will do

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 340

CHAPTER TWELVE

FIGURE 12.5 Typical column responses for the conical beam and straight column system. (Constructive appendix systems.)

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

341

be more than sufficient for this purpose. However, there is a good chance that the responses provided by the manufacturer will differ from those determined by the software due to slight variations in rod sizes and construction details.

12.4.3 General Frame Analysis Software and Frame Formulas Most frame analysis software programs are suitable for determining approximate column response values, especially when dealing with rigid multi-span frames with unequal spans and the tables cannot be used. The reactions of statically determined, but relatively rare, three-hinged gable constructions can be easily calculated using static equations. Two-hinged frames that are statically indeterminate in one degree of freedom are much more common. The vertical reactions of a double-jointed frame are the same as those of a single-jointed beam. The horizontal forces of a single-span rigid frame with atypical pitch, not covered by manufacturer tables, can be estimated by standard framing formulas found in Kleinlogel5 and elsewhere.

12.4.4 Altitude Check Wind altitude often controls foundation size for metalwork systems rather than down load. A height check takes the minor area of a column, multiplies it by the vertical component of the wind lift force, and compares the result to the offset weight of the roof and foundations. For multi-span rigid frames, the calculated lifting capacity can be increased by 10 to 20 percent to account for continuity effects. If the sustained load does not provide the required safety, the foundation size or depth is increased.

12.4.5 Other Scenarios As an alternative to the response estimation methods described above, it may be worth establishing a good working relationship with some fabricators of prefabricated buildings. Many of these companies would like to run a proposed framing scheme on their computers and print out the column responses (and maybe even provide some preliminary element sizes). An additional advantage of this participation can be sound advice on the feasibility of the project. Occasionally, despite the engineers' best efforts to estimate column responses, the final numbers reported by the manufacturer differ significantly from the assumed values. Smaller numbers are of course acceptable, but larger column responses can lead to foundation redesign. If a time-bound "foundation advance package" has already been awarded - or worse - built, a change of foundation order will certainly follow. Such experience is usually sufficient to open the engineer's mind to the dangers of such assumptions, however educated, and to the advantages of using wide safety margins in such circumstances.

12.5 METHODS OF RESISTING LATERAL REACTIONS Once the column reactions of various loads are determined, they should be combined into load combinations required by the applicable building code to arrive at the most critical values for the internal and external loads (Fig. 12.6) . Once the worst combination of reactions is known, a method must be chosen to withstand the forces. There are several foundation designs that can withstand horizontal loads, some of which are discussed below.

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 342

CHAPTER TWELVE

12.5.1 Tie Rods A single-span rigid frame, under equal gravitational loading, produces two equal horizontal reactions acting in opposite directions (Fig. 12.1a). The most direct way to "erase" both is to connect opposite columns of the frame with a tie rod. Tie bars are best suited for large horizontal forces (over 20 kip) and are generally not economical for smaller loads. The required cross-sectional area of a tie rod is determined by dividing the tensile force by an allowable tensile stress in the bar. Some designers consider the maximum allowable tensile stress for this purpose to be 60% of the rod's yield strength, but this approach is fraught with danger. The elongation of a highly stressed bar under load can be significant, as can be easily shown by standard formulas. When a rod of length L, area A, and Young's modulus E is subjected to a force P, its length changes according to the length of the rod: PL Rod AE To get a feel for the numbers involved, assume that L is 120 ft, P is 36 kip and Fy of tie rod rebar is 60 ksi. If the allowable tensile stress Ft is assumed to be 0.6 Fy, then Ft 0.6 60 36 ksi. Loads are adjusted according to applicable building codes: (a) dead ground snow (or payload on roof) wind from the right (also check this combination with no wind; produces maximum external reaction on left foundation); (b) Wind from left - dead (creates maximum lift and internal load on left foundation).

and the required area of steel Arq is 36 Arq 1.00 square inches or a #9 bar 36. The elongation of this bar under load would be 36 120 12 bar 1.79 inches 1.00 29,000 If the columns are evenly spaced, each can differ by 1 .79 0.895 inches 2 move

A tie rod that allows the frame posts to spread nearly 1 inch under load can damage the frame, and for this reason it is best to keep the tie rod tension low. Apparently, halving the allowable stresses reduces strain under load by 50 percent. Alternatively, some designs allow the bolts to be re-tensioned after the concrete has been installed and cured, as explained below. In one of the older tie rod designs, a mild steel rod is attached directly to the column base with a yoke and pin (Fig. 12.7a). Obviously, the base of the column must be sunk below ground for this approach to work. If the base is at floor level or higher, the tie rod can be hooked onto the pillar (Fig. 12.7b). In this case, the most suitable anchor material is deformed rebar. Hidden but exposed to soil moisture, steel rods must of course be protected from corrosion. Tie rods should be galvanized or epoxy coated and encased in a grout filled plastic sleeve for added protection. Because building codes do not allow lap splices for tensile members, mechanical splices are required. For multi-bar tie rods, the seams must be staggered by at least 30 inches. One of the two main disadvantages of both designs is that the bolts tend to sag under their own weight when simply set into the ground. It is certainly possible to eliminate the slack

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

343

FIGURE 12.7 Foundations with tie rods: (a) tie rod connected to base plate; (b) Tie rods embedded in the columnar pier.

Install turnbuckles, but can be tricky to incorporate into the hems. Another associated disadvantage is the elongation of the bar under load mentioned above. For the design of Fig. 12.7a, this strain cannot be resolved by post-stressing as explained below, and for the design of Fig. 12.7b post-stressing can be difficult to achieve. Tensioning the rods in Fig. 12.7a does not make them elongate less under future loading. In fact, it can overload the bars and their connections. Why? For post-tensioning to work, the bars must be anchored in the concrete after stretching and all future tensile stresses must be applied to the concrete rather than directly to the bars. The anchoring of the tension bars compresses the concrete and any future tensile forces acting on the concrete must first neutralize these compressive stresses. On the other hand, when the tie is attached to the base plate, the external reactions of the column are applied directly to the rod and this strain is added to any tensile stresses introduced during tensioning of the rod. The foregoing suggests that for post-tensioning to work, the tie rods must be fitted into columnar concrete grade beams designed to withstand the appropriate compressive forces. 1

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 344

CHAPTER TWELVE

such a design is shown in Fig. 12.8. Note that the cable is centered on a flat beam to eliminate any bending stresses from eccentricity. Also note that the joist and pier are placed together to minimize the amount of dowels and bracing that would be required to transfer external forces from the pier to the pier. As with any underground post-tensioning, protection of the anchors and tendons is critical to the long-term durability of the system. Concrete beams are located some distance below the floor (typically 12 to 16 inches) and are reinforced with at least four bars plus post-tensioned tendon in the middle.

FIGURE 12.8 Post-tensioned tie anchor enclosed in concrete beam.

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

345

The bars are usually surrounded by two-piece stirrups. To protect these bars from corrosion, the concrete cover should be at least 2 inches on the top and sides and 3 inches on the bottom. A corrosion-resistant coating is also helpful. These "tie rods" can not only work in tension, but also in compression and act as seismic foundation anchors. Common class bar sizes range from 14″14″ to 24″24″ depending on column behavior needs. Stresses are small: a bar placed in a 16-16 degree beam and tensioned to the aforementioned 36 kip force level exerts a compressive stress of only 0.141 ksi. Some engineers are concerned about the potentially high upward pressures exerted on grading bars framed in open foundations. As the foundations settle slightly under load, they pull the ends of the sleeper beam with them while the rest of the sleeper beam is supported by the ground at the original height. To reduce this upward pressure and allow some settlement under the horizontal joists, they can be placed on loose, non-locally compacted subsoil (or perhaps on a layer of biodegradable material such as cardboard). This rather complex system has to compete with the cheaper tie rod construction, where the rods are simply embedded in a thick plate (Fig. 12.9). These built-in bars are normally not in tension and are therefore subject to the loosening and stretching problems mentioned above. But there is an even more serious potential hazard - that of stem cutting during future installation of underground pipelines and utility lines. Both a sheathing and a massive reinforced beam have a better chance of surviving because workers are trained to avoid cutting reinforced underground pipes and wires. A thick slab is not that much of a concern. Tie rods are very effective at resisting opposite column reactions, but unless the canted beam construction is used, they are not as effective at resisting reactions acting in the same direction as a wind load. In addition, tie bars obviously cannot be used in buildings with deep trenches, large tool pits, and similar discontinuities in the ground.

12.5.2 Hairpin Rebar Hairpin rebar uses the same principle as tie rods, but instead of connecting two opposing supports with a steel rod, the hairpin system relies on the floor slab acting as a tie rod. Of course, the concrete itself cannot withstand large loads, but the steel reinforcement in the slab can. The function of the hairpin bars is to transmit the reactions of the horizontal support to the rebars of the slab or welded wire mesh, essentially by overlapping seams. The required area of slab reinforcement - and clip bars - is determined by dividing the horizontal column reaction by an allowable tensile stress of the reinforcement - 24 ksi for Class 60 rebar and 20 ksi for welded wire mesh. The length of the hairpin bars depends on the amount of slab reinforcement to be attached. Usually, hairpin rods are hooked around the external anchor bolts and inserted into the slab at 45° (Fig. 12.10). The hairpins should be long enough so that the assumed fault plane intersects the desired number of plate rebars or wires and allows for proper development. Hairpin rebar works best when embedded in slabs containing properly spliced, deformed bar rebar, which is common for structural slabs but not planar slabs. The use of hairpins on flat panels raises several troubling questions about the panel's ability to transfer stress. First, at-grade slabs are often unreinforced or contain only short fiberglass or steel fibers that are clearly unsuitable for stress transmission. Second, the plate, even if reinforced, is usually with welded wire.

FIGURE 12.9 Tie rods in thickened plate.

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 346

CHAPTER TWELVE

FIGURE 12.10 Hairpin reinforcement.

Fabric that tends to receive less attention than deformed rods in terms of placement and connection practices. Third, in-plane slabs often have construction and control joints where all or most of their reinforcement is interrupted – along with stress transfer. (A conscientious engineer, determined to preserve the load path and choosing to continue the welded wire mesh across the joints, can pay dearly for a cracked floor - hardly a good alternative.) Even if the slab contains no joints, as it can There is a fourth and biggest problem with shrinkage-compensating cement – the possibility that the panel will be cut in the future to replace an underground cable or pipe. The building's transverse load protection can be destroyed in one fell swoop – unnoticed! Such a catastrophe is much less likely with a structural panel. Also, a caveat regarding pits and ditches mentioned for risers applies equally to hairpins. Some engineers contend that a flat slab need not be continuous between opposite ends of the building because lateral loads can be transmitted directly to the ground through friction under the slab. others counter that the usual use of polyethylene vapor barriers would greatly reduce any potential for friction. In either case, reliance solely on friction to transfer lateral loads is objectionable. Yet another unanswered question is: how does the slab, weakened by joints, behave in compression against internal horizontal loads? Will the plate hold or will it bend like a sheet of ice near a bridge pier? As the author has recommended in the past,6 it is best not to rely solely on stacked rebars to transfer stresses in graded slabs. This suggestion goes against the persistent recommendations of many in the metal fabrication industry who see cheap hair clips as a simple solution to a complex problem. However, until some realistic testing of this system under various conditions has been performed, it is best to restrict the use of clip bars to slabs with continuously spliced deformed bar reinforcement.

12.5.3 Moment-resistant foundations Pillar foundations can be designed to withstand vertical and lateral loads in a cantilever retaining wall, completely independent of ground anchors. The design methodology is well developed and widely used; a good source is the CRSI Design Handbook.7 Unlike retaining walls

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

347

However, moment-resistant foundations are subjected to significant vertical loads, and it is often advantageous to design your shoes with a toe that is longer than the heel, rather than the other way around. In this configuration, the downward reaction of the column helps counteract the external horizontal load (Fig. 12.11). This solution offers many advantages: it allows future cutting or even complete removal of the slab on the ground without compromising the integrity of the foundation; can accommodate any number of pits, trenches and slab depressions; it can withstand internal and external lateral loads. However, this method usually results in foundations that are larger than those designed using either of the previous two methods, although in either case some additional weight may be required to prevent wind uplift. Designing moment-resistant foundations is time-consuming. Horizontal column reactions can fail a moment-resistant foundation by overturning, sliding, or both. A minimum safety factor against tipping and slipping due to transient loads shall be at least 1.5; should be increased to 2.0 if the loading is caused by gravity. While the buoyancy of the support is counteracted with the usual means such as correct "ballast", the resistance to lateral loads is achieved through a combination of ground friction and passive ground pressure. The slip resistance can be built up by ground friction and, if necessary, by a concrete shear spring projecting under the foundation floor in the undisturbed ground (Fig. 12.11). Moment resistant foundations, like most cantilever retaining walls, rely on some degree of wall rotation under load to mobilize active and passive soil pressure. The slope arises because the soil pressure under the foundation is not uniform (Fig. 12.12). This twisting movement can endanger the fragile materials of the outer walls. Such rotation can be avoided by a bottom plate as described below, but in this case a much higher 'resting' bottom pressure coefficient should be used instead of the active pressure. For example, basement walls are usually designed for static pressures. How much rotation must occur to allow the use of active and passive floor presses instead of resting presses? Current practice allows the use of active pressure coefficients for wall or column motions as small as one-tenth of 1 percent of their height.8

FIGURE 12.11 Moment-fixed foundation.

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 348

CHAPTER TWELVE

FIGURE 12.12 Load transfer in the moment-bearing foundation.

12.5.4 A Combination of Methods Some methods of resisting lateral loads can be combined to create an economical and practical foundation design. A common system are floor slab anchors in combination with a moment-bearing foundation or tie rods. This system is often inadvertently provided by engineers who routinely specify penetrations that extend from the foundations to the slabs (Fig. 12.13). The design function of the dowels and foundation seat is to support the slab to allow it to pass through poorly compacted areas near the walls; An unintended function is to constrain the movement of the column wall and column under horizontal loads. In fact, the dowels look like neatly arranged hairpins – at no extra cost. The main advantage of this "plate and strut" system is its redundancy: the moment-bearing foundations or tie bars serve as the primary means of withstanding horizontal reactions at extreme loads; The slab helps by acting as a confined horizontal membrane, supporting the tops of the foundations at all other times. If the continuity of the slab were broken at any point, the foundation would not lose all of its lateral bearing capacity; In the worst case, the structure may suffer some minor maintenance issues, but not a complete load path collapse. Likewise, the foundation of a tie-rod building may rely on slab pins to transmit internal forces to the slab against which tie-rods are ineffective (unless enclosed in horizontal beams). In fact, the design of bolts and foundations resisting moments against inwardly acting horizontal forces with buoyancy (Fig. 12.6b) would be quite difficult without dependence on the plate. While the objections in Sec. 12.5.2, it seems easier to justify the plate's contribution when placed under compression than under tension. Example 12.1 Design a typical moment-bearing foundation for a single-span, bolt-ended rigid frame building 84 feet long and 25 feet apart. The structures have a roof pitch of 4:12 and an eaves height of 16 feet (Fig. 12.14). The ice depth spacing is 3.5 feet and the bottom of the column is 6 inches above the adjacent floor. The following load combinations have been found to produce the most unfavorable structural responses: 1. Dead snow load as secondary load: vertical, 37 kip (down); horizontal, 30 kip (off). 2. Right dead wind load (to right side foundation): vertical, 14 kip (height); horizontal, 11 kip (inwards). Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

349

FIGURE 12.13 Moment-bearing foundation combined with slab dowels.

FIGURE 12.14 Rigid frame model used in Example 12.1.

The direction of column reactions on foundations caused by gravity loads are shown in Fig. 12.15a and by wind loads in Fig. 12.15b. Assuming the vertical load on the column is applied 6 inches from the right edge of the pier, the soil weight is 120 lb/ft3 and the concrete unit weight is 150 lb/ft3. The slab on the level covers the inner part of the foundation. The permissible floor pressure is 3 ksf. The desired safety factors against rollover and skidding are 1.5 and against lift off 1.1. Use fc′ 4000psi. Determine the total resistance of the foundation to tipping, sliding and lifting. Case 1: Dead load due to accompanying snow. Foundation sizes are determined by trial and error. Experience a 9ft long, 4ft wide, 2ft thick base with a 2ft, 2ft columnar pier. The weights and recovery times are as follows (Fig. 12.16): Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 350

CHAPTER TWELVE

FIGURE 12.15 Column reactions for example 12.1: (a) gravity loads; (b) from wind loads.

Figure 12.16 Weights and forces for case 1 of example 12.1.

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

Distance to point A

Weight W1 (0.50 0.15 1.5 0.12) 2 4 W2 2 2 2 (0.15 0.12) W3 9 4 2 0.15 W4 5 1.5 4 0.12 PΣW

2.04 chicken 2.16 10.8 3.6 37 55.6 chicken

8 feet 6 4.5 2.5 5.5

351

MR 16,32 kip-ft 12,96 kip-ft 48,60 kip-ft 9,00 kip-ft 203,5 kip-ft ΣMR 290,38 kip-ft

Roll over moment MOT 30 kip 4 ft 120 kip-ft ΣMR 290.38 roll over safety factor 2.42 1.5 120

it's all right

Find the position of the resultant measured from point A: 290.38 kip-ft -x 5.22 ft c.g. 55.6 kip The resultant of the vertical loads acts with an eccentricity to the centerline of the foundation of: e 5.22 4.5 0.72 (ft) left of the centerline of the foundation. So the total eccentricity of the load is MOT 120 eo e 0.72 1.44 (ft ) 55.6 ΣW The core limit of the shoe is 9 ft 1.5 1.44 (ft) 6 Therefore the resultant is within the core limit of the shoe, which means that ground pressure can be determined by the formula: P M fp, max, min ± A S with P 55.6 kip (ΣW) A 9 ft 4 ft 36 square feet (footprint) M 55.6 kip 1.44 ft 4 92 S 54 ft 3 (shoe profile module) 6 fps, max 3.02 ksf fp, min 0.06 ksf The shoe is designed below. Slip resistance is achieved through a combination of floor friction and passive shoe pressure (see Fig. 12.12). Assume

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 352

CHAPTER TWELVE

Ka (active pressure coefficient) 0.33 Kp (passive pressure coefficient) 3.00 (dynamic friction coefficient) 0.55 So for a 4 foot wide strip of soil FR 55.6 0.55 1 2 (3.00 0.33) 0.12 3 .52 4(ft) 34.9 kip 30 kip However, the safety factor of 34.9/30 1.16 is not sufficient. At this point, either a prominent shear cut can be made in the ground (see Fig. 12.11) or the passive pressure of the foundation walls running horizontally between the column foundations can be used. The construction of shear springs is explained in many references such as e.g. B. Ref. FIG. 7 then shows the second approach. Determine the required length of wall that, as a horizontal cantilever, encloses enough floor to provide a 1.5 slip rating. The total resistance required is Freq'd 30 1.5 45 kip The amount of skid resistance added by a linear foot wall is fw 1 2 (3.00 0.33) 0.12 3.52 1.96 kip/ feet The required wall length on each side of a column is (45 34.9) Lw,rq 2.58 (ft) 2 1.96 Check the horizontal curvature of a 12 inch wall, 3.5 feet deep with at least three #4 horizontal bars placed in inner layers behind the #4 vertical bars: wu 1.96 1.7 3.33 kip/ft 3.33 2.582 mu 11.05 kip-ft 2 d 12 2 1 2 1 4 9.25 (in) A 0.58 in2 0.58ρ 0.0015 → au 4.44 9.25 3.5 12 Mn 0.58 9.25 4.44 23.82 kip-ft Mu

Case 2: dead wind uplift. Check the experimentally selected foundation for case 1 for wind uplift and permanent load. The forces acting on the foundation (see Fig. 12.17) are: U 14 kip, H 11 kip, ΣW 18.6 kip, calculated for case 1 (55.6 37 18.6 kip). Also from Case 1, minus the column reaction restoring moment: ΣMR 290.38 203.5 86.88 (kip-ft) 86.88 x–c.g. from right edge 4.67 (ft) or 18.6 0.17 ft left of foundation centerline (l 4.33 ft) Capture and restore moments about point B:

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

353

FIGURE 12.17 Forces acting on the foundation for case 2 of example 12.1.

MOT 11 4 14 3.5 93 (kick toe) MR 18.6 4.33 80.5 (kick toe) MOT

Again we must enlist the help of the foundations shown in Fig. 12.17. The walls are 12 inches deep with foundations 2 feet wide, also 12 inches deep. The weight of the walls, foundations and floor above the foundation edges is Wwall [(3′ 1 2′ 1) 0.15 (0.5 2.5 0.5 3) 0.12] (25 4) 22.68 ( kip) Then ΣW 18.6 22.68 41.28 (kip) MR 18.6 4.33 22.68 3.5 159.9 (kip-ft) 93 (kip-ft) 159.9 F.S. 1.72 1.5 93

Gliding is ok after observation. Construction of the Concrete Pillar The maximum working load moment at each pillar is Mmax 30 kip 2 ft 60 kip-ft As we do not know what percentage of this is the permanent load, conservatively use a total load factor of 1.7: Mu 60 1.7 102 kip - ft Try a 2- and 2-foot pier (Fig. 12.18) with three No. 7 poles on each side, Nos. 4 and 3 anchorages in transparent cover: d 24 3 1 2 7 16 20.06 in. AS 1.80 inch2

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 354

CHAPTER TWELVE

1.80 0.00374 20.06 24 au 4.35 Mn 1.80 4.35 20.06 157 kip-ft 102 Control cut: Vu 30 1.7 51 (kip) Vn 0.107 20.06 24 5.15 kip from

Provide loops #4 with the maximum spacing required by the code, d/2 or 10 in o.c. FIGURE 12.18 Pillars for Example 12.1.

Shoe Design Column Case 1 governs the design. The pressures acting on each shoe are shown in Fig. 12.19. As determined above, fp,max 3.02 ksf fp,min 0.06 ksf

The pressure on the right side of the pillar is 1.38 ksf (by interpolation). (3.02 1.38) 5 5 2 1.38 52 Mmax 30.95 kip-ft 23 2 Mu 30.95 1.7 52.62 kip-ft/ft For a 24-inch. thick with bar No. 7 and 3 in. coverage, d 24 3 7 16 20.56 in. Try six bars No. 7 on a 48-in. basis: 3.61 .0036 20.56 48 au 4.35 mn 3.61 4.35 20.56 322.9 kip-ft Mu 52.62 4 210.44 kip-ft Mn

For short direction bars the minimum gain is Amin 0.0018 24 9 12 4.66 (in2) Use 11 #6 bars (Ao 4.84 in2). Check the shear (carefully on the support surface): 1.64 5 Vu 1.38 5 1.7 18.7 (tilt per foot width) 2

Monte

冣

CV 0.107 20.56 12 26.4 From

Finally, check for negative deflection of the foundation under wind uplift, when the foundation must support its own weight and that of the upper soil since it is essentially floating in the air. The final downward load is

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

355

FIGURE 12.19 Ground pressures under the column base in case 1 of example 12.1.

Wn 1.4 (1.5 0.12 2.0 0.15) 0.672 kip/pe l 5 pe 2

Mu 0.672 5 2 8.4 kip-ft 100,800 lb-in per ft width Design foundation in negative pressure as normal concrete, based on ACI 318.9 Chap. 22. Assuming effective thickness is 2 inches less than actual thickness, 12 222 S 968 (in3) 6 100,800 ft.u 104 psi 968

0.55 for plain concrete, not ACI 318-02

Ft,u 0,55 5 兹4000 · 174 psi ft,u Ft,u

Check beam shear:

Monte

冣

22 Vu 0.675 5 2.14 kip/ft 12 4 12 22 Vn 0.55 兹4000 12.24 kip/ft Vu 3 1000

If the decisive load combinations contain a case of lateral column reactions on the building, the foundation must also be checked in this case. The final design of the pier foundation is shown in Fig. 12.20. The section through the outer foundation wall is shown in Fig. 21.12.

FIGURE 12.20 Moment-bearing column foundation according to example 12.1.

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS

356 Conveyed from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

FIGURE 12.21 Wall foundation between column foundations in example 12.1.

357

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 358

CHAPTER TWELVE

12.5.5 Track foundations Track foundations may be prescribed to a greater depth for locations with even but poor soils. Such a situation usually requires deep foundations, such as piles, unless the loads are so small that, when distributed over a large area, the soil pressures are justifiably low. A prefabricated lightweight building with rigid multi-span structures and flexible walls is usually suitable. Carpet foundations are considered economical when the sum of the individual pedestal areas accounts for more than half of the total building floor area. The greatest savings are achieved when the top of the ceiling is at floor level, eliminating the need for a slab on the floor. A special type of carpet underlay that is occasionally found in low-rise buildings with a basement is the so-called floating carpet. A floating mat is located well below the surface, usually at basement level, and is designed to put no more pressure on the ground than it did before excavation. Even some very weak soils (except peat and organic matter of course) can support a floating mat. A rigorous design of carpet foundations takes into account carpet stiffness, boundary conditions and variability in column loads. Most accepted methods follow the model of a slab on an elastic foundation subjected to point loads and bending moments. One source of information on carpet design is ACI Committee Report 436.10. Another excellent method of carpet analysis, complete with formulas and graphs, can be found in Ref 4. There are also several computer programs available that deal with carpet design. For buildings of modest size and decent floor plan, it may be possible to simply locate the geometric center of the mat at the center of gravity of all vertical column loads and assume that the soil pressure is uniform under the mat.2 This approach can present problems when considering the properties are variable enough to cause significant local pressure fluctuations, the only safeguard against this is a generous and conservative factor of safety used in the design. In fact, it is recommended to provide more reinforcement than required for the analysis and to specify the same reinforcement at the top and bottom of the mesh.11 Local stiffening of the mesh may be required to cover all known isolated areas with unsuitable material. For mat foundations, horizontal column reactions are resisted by external action within the mat, while internal action reactions are resisted by friction of the mat with the ground. Wind uplift is rarely a problem because of the large mass of concrete. With proper design, differential settlements between multiple supports are minimized. These advantages of track foundations are offset by disadvantages, which include design complexity, the need for heavy reinforcement, difficulties in accommodating deep pits and trenches, and in northern climates potential frost damage to track on the level. In fact, according to BOCA Code Section 1806.1.2, carpet foundations must be transported to a level below the freeze line. This presents a problem for flat mats. Some engineers have found that recessing the mat edges below the freeze line adequately meets this requirement.

12.5.6 Pile foundations Pile foundations are commonly used for weak and insufficient soils supported by high quality material. Piles made of wood, precast concrete, steel tubing and steel H-shapes are driven into the ground; Concrete piles can also be erected on site. Pile engineering is a complex subject well beyond the scope of this book; Whenever piles are involved, geotechnical support is mandatory. Our main interest is mainly how piles withstand lateral loads and elevations. Piles supporting building columns are usually installed in groups of three or more, regardless of column loading. The "tripod" is stable, even if the piles aren't driven with perfect accuracy - normal driving tolerances can easily result in a pile being 3 inches or more from the intended position - and the column ends up being eccentric to the pile's center of gravity Group. One and two-post clusters require other methods of relieving unplanned eccentricities, such as B. Rigid beams at surrender level. A stake acts as a laterally braced column that extends through weak soil to suitable soil. Column loads are transmitted to the ground through end bearings, skin friction, or both. Friction stakes offer superior quality

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

359

Lifting strength and equivalent compressive and tensile capabilities when stress transmission is provided by appropriate splicing and stress reinforcement. (Some engineers limit the pile's tensile capacity to two-thirds of its compressive capacity.) End support piles, on the other hand, can only offer their own weight against lifting forces. The bearing capacity of piles resisting loads from a combination of surface friction and end support is calculated as the friction capacity of the pile plus the weight of the pile; can also be determined by testing. The piles resist lateral bending loads. A simple general model assumes cantilever behavior with a point of attachment some distance below the surface (Fig. 12.22a); The stiffer the stake and ground, the shallower the depth for strength. A more sophisticated model assumes that piles behave like beams in elastic foundations. Regardless of the calculation, the lateral strength of vertical piles is often 5 to 10 kips per pile. A problem with both approaches is that the piles must undergo significant displacement at their tips to engage the bending mechanism; such movements can damage buildings with brittle surfaces. In the case of large lateral loads, damaged piles may be appropriate (Fig. 12.22b). Most seismic codes require blocks to be connected by links capable of carrying tensile or compressive forces equal to 10% of the column load. This bracing can be provided by a height reinforced slab or by tie beams attached to foundation walls or height beams, allowing load transfer to members with significant surface areas and passive bearing capacities. Passive pressure on the piles themselves is often overlooked during pile driving because of their small contact area and ground disturbance. Installing piles alongside existing buildings is difficult to accomplish. Another limitation of piling is cost: building codes often require load testing for piles with capacities in excess of 40 tons, often incurring significant costs to make other solutions worthwhile.

FIGURE 12.22 Shear load resistance of foundation piles: (a) vertical piles for moderate shear loads; (b) Mass piles for large loads.

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 360

CHAPTER TWELVE

12.5.7 Caissons Caissons, or caissons as they are commonly known today, are a type of cross between spread foundations and piles. Some caissons are formed in open excavations, but most are drilled with a special probe. Similar to piles, caissons are best suited for heavily loaded buildings on poor soils. A typical pier has a vertical round shaft with or without a flared bottom and functions through end supports, like a massive pillar of plain concrete with a deep foundation. Flared-bottomed coffins support the floor; The coffins supporting the rim are straight and embedded in the rock. Coffins can also rely on skin friction, like friction pins. These caskets are limited in their carrying capacity and are rarely used; A combination of final storage and some skin chafing is most common. Perforated piers are generally preferred over piles, not only because of their low cost, but also because they do not have disadvantages such as noise, vibration and ground elevation - important considerations in urban design. Another advantage of caissons over piles is the fact that the base course can be inspected before concreting. In contrast, ramming is done blind. In one hilarious case, while driving, the stakes warped into a semi-circular shape and penetrated the ground below, damaging nearby parked cars! Meanwhile, the piling workers had the impression that everything was in order. The caskets' considerable size—usually 2 to 6 feet in diameter plus bell—and heavy weight make them resilient to lifting. Some additional resistance to lifting is provided by the floor over the bell and skin friction. When a lifting force exceeds the pulling capacity of a simple concrete manhole, a full length reinforcement cage can help. Also, a half-length node cage can be specified to improve upper shaft flexing ability. Caissons resist lateral loads in the same way as piles—by flexing the shaft and transferring the load to the leveling beams, which attack passive soil resistance (Fig. 12.23). The depth and thickness of planar beams are determined by analysis.

FIGURE 12.23 Bell shaped caisson and beam foundation.

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

361

12.5.8 Set-back slabs The cheapest foundation imaginable—beyond simply supporting columns in the ground—consists of inclined slabs pointing downwards at the perimeter (Fig. 12.24). This design requires no foundation walls and is commonly used by homebuilders in warm, frost-free climates. The side beam depth used for homes and light temporary structures can be as little as 12 to 18 inches. This recessed slab version, whether in its plain or "heavy" (slightly flared) version, is generally unsuitable for engineering building forefoundations as it offers little dead weight to counteract wind uplift and does not contain enough soil to realize the passive benefits Print. This design assumes some contribution from cloud cover to help withstand elevation and lateral loads - with little justification. As close analysis would show, the thin unreinforced plate or the plate lightly reinforced with welded wire mesh is overloaded at the point of thickness change. According to Ref. 3: "At the point where the 'leveling beam' begins, a crack will almost certainly occur in the floor slab". Once the slab cracks, the thick part becomes a separate—and probably inadequate—foundation for the building's column. This type of foundation becomes more effective as the size of the recessed slab is increased. For example, the design commonly used in the southern states for metal building supports has a rim beam 24 to 30 inches deep and about 24 inches wide at the bottom. In some cases, the foundation is further widened at the column sites. For moderate stress levels, this size may be sufficient, but a better solution in terms of avoiding panel cracking is to design the edge beam independently of the panel, as described below. If a mixed foundation is desired, a rug can provide the solution.

FIGURE 12.24 Overturned plate.

12.5.9 Mass Foundations There is another alternative to foundations consisting of columns and column foundations that are placed separately - and more substantial than the recessed slab. A mass foundation, also known as an amorphous foundation, does not rely on a slab for stability. With this type of construction, both the foundation and the wall are laid together in an excavated construction pit, which means that there is no need for formwork and reinforcement (Fig. 12.25). Mass foundations are commonly used in residential and light commercial buildings in areas where cohesive soils can safely support vertical cuts in the ground. The safe word is key as many accidents have happened while digging open trenches. Some sources suggest so

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 362

CHAPTER TWELVE

FIGURE 12.25 Mass foundation (form foundation).

Finish off at least the top 4 inches of the concrete for a more refined appearance and to reduce rainwater infiltration into the excavation. From a static point of view, bulk foundations can be made as large and deep as necessary to withstand vertical and lateral loads.

12.6 ANCHOR BOLTS AND BASE PLATES Anchor bolts, or anchor rods as the AISC prefers to call them, transmit lifting and shear forces from the building columns to the foundations and act as a critical link between the metal building fabricator and the recorder engineer domains. Some pitfalls and misunderstandings in the specification of anchor bolts for metal construction systems are discussed in Chap. 10; here our focus is primarily on the technical aspects.

12.6.1 Anchor bolts: function and types Anchor bolts have two basic functions: 1. Position the column in the right place and keep it stable during assembly. The minimum number of anchor bolts for steel structure supports required by OSHA steel structure codes is four. The only exception made to OSHA's safety and health standards for the construction industry, 29 CFR 1926 Part R, Safety Standards for Steelwork, is for studs. OSHA

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

363

defines a post as a "substantially vertical" structural member that "(1) weighs 300 pounds or less and is axially loaded, or (2) is not axially loaded but is laterally restrained by the overlying member." In the past the minimum number of column anchor bolts was two, and some details from metalwork manufacturers still reflect the old practice. 2. Transfer of lateral and vertical loads from the column to the foundation. Anchor bolts have limited ability to transfer shear and for very large lateral reactions it may be better to use shear lugs. Most construction specifications require the use of a template to define the bolts. Anchor bolt steel must conform to a relatively new specification ASTM F1554, Standard Specification for Anchor Bolts, Steel, 36, 55 and 105 ksi Yield Strength, which has replaced the earlier ASTM A307. As the title suggests, the new specification includes three grades of steel, but the most common still uses the traditional 36 ksi grade. This grade is relatively inexpensive and weldable, which is important when field corrections and bolt lengthening are required. The highest grade, 55, can only be safely considered weldable if special requirements are placed on weldability and carbon equivalent. Some engineers crave high strength steel and specify anchor bolts to ASTM A 325, which is incorrect and should be avoided. ASTM A325 is typically used for high strength bolts that connect steel to steel, not steel to concrete, with maximum bolt lengths typically limited to 8 inches—too short for anchor rods. As discussed below, the strength of the bolt material is often less critical than the strength of the concrete supporting the bolts. Anchor bolts derive their support holding power from their bent or flared ends against the concrete. Except in the case of subsequently mortared-in adhesives and dowels, the adhesion contribution between the screw shank and the concrete is neglected. Depending on the configuration of their set back ends, anchors are called L-bolts, J-bolts, head bolts and anchor plate bolts. Of these, L and J screws used to be the most popular until it was found that L screws were less effective at preventing slipping than cap screws. has witnessed. Even screws with a plate at the end, once a favorite of structural engineers, fell out of favor when it was realized that the larger the plate, the greater the plane of weakness it introduces into the concrete. Shear slings - short flat bars welded to the underside of the column base plates - are used to transfer shear forces to the concrete without relying on anchor bolts. To be effective, shear shoulders must be sufficiently contained13 and require special formwork inserts. Shear braces can be used most effectively to withstand very large lateral loads and are most commonly found in West Coast prefabricated homes. As evidence accumulated that properly embedded limited head anchor bolts were capable of fully developing the tensile strength of regular and high strength bolts, 14 head bolts gradually became the anchors of choice.

12.6.2 Head Anchor: Design Principles A head anchor remote from a concrete edge and subjected to pull-out develops a 'concrete failure cone' (Fig. 12.26a). When multi-pitched anchor bolts are used, as is the case in most practical designs, their fracture cones partially overlap (Fig. 12.26b). Head anchors located close to a concrete wall edge or a pillar only develop a partial concrete failure cone (Fig. 12.26c). Theory and experimental data for all these models can be found in references 13, 14 and 15. The tensile strength of an overhead anchor depends on two factors: the tensile strength of the steel bar and that of the concrete failure cone. It's easy to scale up the former simply by using as many large screws as the project requires, but it's quite difficult to scale up the latter. To increase concrete failure cone size, longer anchor bolts are required or the bolts must be spaced farther apart. To ensure a degree of ductility and avoid brittle concrete failure, this is common practice

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 364

CHAPTER TWELVE

FIGURE 12.26 Concrete failure cone for head anchor: (a) single full concrete failure cone; (b) overlapping fault cones; (c) Partial failure of bolts embedded in pillars and walls.

Make the concrete breaking cone more tensile strength than the anchor bolt steel. With large loads, the length of the anchor bolt can reach several meters. A common complication is anchors that are too close to the concrete edge. In order to avoid splitting failure of the concrete, a sufficient edge distance must be provided. References 15 and 16 recommend a minimum edge spacing of five bolt diameters or 4 inches, whichever is greater. Model building codes (IBC, BOCA) now contain design methods for anchor bolts that act in tension and/or shear. The 2002 edition of ACI 3189 includes for the first time Annex D which lists very detailed requirements for anchoring in concrete. According to the provisions of the Code, the design possibilities of anchors are very sensitive to edge distances. When anchor bolts are placed too close to the concrete edge it can be very difficult to develop the forces required and the author's practice is to specify minimum edge distances on contract drawings rather than relying on manufacturers' standard edge distances. For example, see Fig. 12.20, which requires a minimum edge clearance of 7 inches.

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

365

As can be seen, it is much easier to determine the size, number and spacing of anchor bolts - a task normally performed by a fabricator - than to design them in concrete, a task left to a civil engineer.

12.6.3 Reliance on column reinforcement A practical approach to the design of anchor bolts relies on adequate vertical column reinforcement to transfer tension loads to the foundation. In this model, the closely spaced strap loops transfer shear and protect against concrete cracking. The bolts should be reasonably close to the pier reinforcement and long enough to allow adequate development length of the pier reinforcement. The available development length is measured from an intersection of the reinforcing bars with the concrete failure cone (Fig. 12.27). The required minimum embedment length of the screw is equal to the developed length of the bar plus the horizontal distance between bar and screw plus the concrete cover over the end of the bar.

12.6.4 Bolt pretension Should anchor bolts be tightened? Some say yes, claiming that a clamping force resulting from tightening the screw helps prevent slippage at the base of the spine. However, most prefabricated buildings can tolerate some ground slip without adverse effects; such slip can even be beneficial for the following reason. Column base plates typically come with oversized holes to accommodate tolerances in anchor bolt placement, and only one bolt is likely to actually rest against an edge of the plate hole. As Fisher17 points out, if the base is allowed to slip, another anchor bolt may come into contact, thus aiding in load transfer; he suggests not relying on more than two screws in a cluster to transmit base shear. As previously mentioned, fabricators and contractors of prefabricated buildings often omit grouted leveling boards under columns. Regardless of what one thinks of this practice, it makes the baseplate easier to slide. The shear resistance of a group of anchor bolts

FIGURE 12.27 Transfer of traction and lateral loads to the columnar pier when engaging the vertical reinforcement.

(Video) Suspension System Components

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 366

CHAPTER TWELVE

it is often not sufficient if only two of the screws are effective in terms of shear strength. To engage all four bolts (the typical number), some engineers dictate welding on heavy duty washers that are tightened around the bolts at the top of the base plate after they are installed. Others argue that this practice introduces flexing into the anchors, since the force is now being transmitted through the baseplate, and that the shims must be thick enough to prevent bearing failure. Another view simply assumes that the anchor bolts are slightly bent out of alignment by the loaded column baseplate until all bolts are engaged. In prefabricated buildings, anchor bolts are typically only preloaded to a “fixed” state, resulting in a low fastening force that is often overlooked in construction. Significant clamping effort is required only for fixed-based columns that rely on clamping forces for moment transfer, or for buildings where lateral deviation is tightly controlled.

12.6.5 Common Anchor Bolt Positions In mid-span and span buildings with studded columns, the most common number of anchor bolts used to be four for side wall (frame) columns and two for end wall columns (and sometimes interior times) (Fig. 12.28). As just mentioned, OSHA regulations for steel mounting now require a minimum of four bolts on all posts except "posts" (see Section 12.6.1). As a result, depending on the rod weight, some previously standardized details have to be revised.

FIGURE 12.28 Typical anchor bolt arrangement. (Note that recently enacted OSHA standards for steel assembly require a minimum of four anchor bolts for all columns except "posts" as defined in Section 12.6.1.) (Star Building Systems.)

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

367

Each manufacturer has a standard set of dimensions for placing anchor bolts. The distance to the concrete edge depends primarily on the type of belt insertion (bypass, flush or semi-flush) and, in the case of bypass belts, on the size of the belt. Figure 12.29 shows a manufacturer's standard dimensions; Clearances may be larger for other manufacturers as shown in the images below. The screw spacing is also standard for each manufacturer. The most difficult to estimate are the dimensions of the cornerstones. For non-extendable end walls, a representative detail is shown in Fig. 12.30. Figure 12.31 shows a detail of an expandable sidewall structure. Solid base columns require a large number of anchor bolts, usually eight, to develop the final anchorage (Fig. 12.32). Complete fixation cannot realistically be achieved with the spaced anchors shown in Fig. 12.33. As mentioned, these “standard” dimensions can easily be changed if needed. They should not be considered sacrosanct, as manufacturers acknowledge - see, for example, a related note in Fig. 29.12. The author's practice is to indicate minimum screw edge spacings on contract drawings and to warn manufacturers that their “standard” details will not be accepted. 12.6.6 Minimum Column Sizes Don't skimp on column sizes. The pier must be large enough to accommodate not only the column base plate, which may be of unknown size, but also sufficient space for placing concrete around anchor bolts, tie rods, vertical bars and formwork. Column clogging can lead to improper concrete placement and structural failure.3 When foundations are designed prior to structural steelwork, the contract drawings must specify the largest acceptable column base plate sizes. Such limitations do not indicate paranoia: on one project, the manufacturer submitted drawings showing a 6-foot-wide column intended to support a 2-foot-wide pier, ostensibly to emphasize the stringent lateral drift criteria specified in the contract documents.

FIGURE 12.29 Example of anchor bolt locations on column sidewalls: (a) with chords aligned horizontally; (b) with pulleys. (A&S Building Systems.)

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 368

CHAPTER TWELVE

An often overlooked detail is enlarged pillars on wall brackets (Fig. 12.34). As in chap. 3, These clips are used to avoid directly anchoring the wall mount to the thin structure web. Clips can be subjected to significant buoyancy and lateral forces without permanent load from the balance beams. To withstand these loads, each bracket must be attached to an expanded foundation pier that supports the column (Fig. 12.35). How big should these pillars be? At least large enough to accommodate the column base and clips and provide adequate edge clearance for your anchor bolts. Unfortunately, the manufacturer's information for this condition is not uniform; Coward. 12.36 shows a solution. A very similar situation occurs when portal frames are used. Figure 12.37 shows a manufacturer layout for this condition. Similar considerations apply to end wall columns that rest directly on foundation walls. An unfortunate situation of Fig. 12.38 could have been avoided with proper coordination between designer and manufacturer.

FIGURE 12.30 Detail of base plate and anchor bolts on corner wall studs. (Note that more than two anchor bolts may be required.) (Nucor Building Systems.)

12.7 GRADE PLATE DESIGN

A discussion of foundations for prefabricated buildings would not be complete without at least a brief mention of some of the problems in planar slab construction. Much of the information needed to design a stepped slab can be found in the ACI Concrete Floor and Floor Design Guide18 and the PCA Floor Concrete Floors.19 We have focused on just a few critical points. A typical floor slab consists of compacted subsoil; subgrade of gravel or crushed stone, usually 6 to 12 inches thick; vapor barrier (if necessary) covered with a layer of sand; and the slab itself with joints, reinforcement and finishing (Fig. 12.39). Proper subgrade preparation is critical to the performance of a slab, as even a carefully constructed slab will eventually fail if placed on poorly compacted or inadequate soil. It is quite common to find several feet of poor soil or even loose debris near the surface supported by better material. In such situations, geotechnical advice is essential: soil surveys would indicate whether the materials could be compacted on site or removed and replaced with engineered fills, in which case deep foundations combined with a structural slab could provide a cost effective alternative. A subgrade helps the slab extend over poorly compacted areas, distribute concentrated loads over large areas, and provide drainage under the slab. The thicker the substrate, the more effective it is. There is no need for a substructure if the subsoil consists of easily compactable, well-draining granules. FIGURE 12.31 Base plate and anchor bolts for vapor barriers The advantages and disadvantages of vapor barriers have been debated for years. Corner column made of rigid side wall frame. (Numbers of proponents cite the need to stop capillary action of soil and anchor bolts as dictated by design, but most keep flooring dry. Opponents point out the usual 6,000 is probably four.) (Nucor Building Systems. ) Polyethylene vapor barriers can degrade within a few years and please note that true vapor barriers are seldom effective in preventing moisture problems. However, polyethylene unintentionally impedes water drainage during curing: while surface water evaporates from the top of the sheet, water is retained on the underside,

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

369

FIGURE 12.32 Anchor bolt arrangement for a fixed base column. (Metal construction systems.)

FIGURE 12.33 Closely spaced anchor bolts are acceptable for stand feet, but not for solid-bottomed supports.

resulting in uneven healing rates and frequent plaque scaling. A layer of sand on top of the vapor barrier is said to mitigate the curling problem. Still, many engineers believe that unless a panel is covered with a moisture-sensitive finish, a panel that sits on a suitable substrate does not need a vapor barrier. The structural design of sloping ceilings is relatively easy, even with concentrated loads. Choice of slab thickness, amount of shrinkage reinforcement and design approach

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 370

CHAPTER TWELVE

FIGURE 12.34 Anchor bolts in wall reinforcement clip. (Star building systems.)

FIGURE 12.35 Enlarged columnar pier on foundation support clips.

Concentrated loads are discussed in ACI 30218 and Art. 20. However, some points are left to the discretion of the designer: ●

Support or insulate the panel on the outer walls? While a continuous insulation joint, commonly found around the perimeter, sounds like a good idea for shrinkage control, it's important to remember that proper compaction is difficult to achieve near walls. Brace the slab against the wall and provide wall-to-slab dowels (Fig. 12.21) to help the slab overcome weak points. the cones

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

371

FIGURE 12.36 Position of brace clamp relative to gantry column. (Corner bulkhead support shown.) (Nucor Building Systems.)

●

it has to be buckled in the field, since compaction near the wall is not possible if they simply protrude horizontally from the wall. Strengthen plate or not? If yes, with what? There are no specific code requirements for slab reinforcement. The well-known welded wire mesh (WWF) is designed to minimize the width of shrinkage cracks, not eliminate them. The same result can be achieved by close spacing of control and construction joints, by deformed reinforcement or by specifying shrinkage compensation cement (type K). If used, wire mesh or rebar should be properly supported by special braces or, better yet, closely spaced concrete blocks of the same concrete type as the slab. The question of whether or not to stop the welded wire mesh at the panel control joints was discussed in Section 12.5.2. A common control connection detail is shown in Fig. 12.40. How close to room control and construction joints? The smaller the distance, the less expected shrinkage. Too small a clearance, however, increases the cost of the seals and their future maintenance. As a rule of thumb, the joints are spaced 15 to 25 feet apart in each direction, which probably matches the column layout. Depending on the spacing between joints, the required amount of heat shrink steel can be determined using the resistance formula in ACI 302.18. While formerly slabs were commonly laid in a grid pattern, the long strip method is currently used, in which the slab is laid in alternating strips 20 to 25 feet wide and further divided into squares by control joints. A lesson the author learned in this regard is not to place construction and control joints parallel to one another: construction joints tend to accommodate all of the movement of the slab, leaving control joints uncracked and therefore ineffective. What type of construction joints should be specified? Of the two basic types of construction joints - wedged and dowelled - dowelled joints (Fig. 12.41) appear to result in better load transfer between them

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 372

CHAPTER TWELVE

FIGURE 12.37 Detail of portal frame base plate. (Nucor Building Systems.)

adjoining plate sections. Careful installation and alignment of the pins is critical, as misaligned pins can impede platen movement and cause cracking. Key connections tend to flake off. Diamond-shaped insulating gaskets placed around the posts can reduce cracking in this area. What surface finish and tolerances should be specified? Plate tolerances are a common source of confusion. Previously, a board was considered acceptable if the clearance under the 10-foot board was no more than 18 inches. Today's requirements are much more complex and involve two so-called F-numbers: FF, which measures the flatness or waviness of the floor, and FL, which controls the flatness of the panel. These F numbers are used by ASTM E 115521 and have been adopted by ACI 117.22 A good introduction to the topic is Tipping.23

●

Quality ceiling construction is not cheap. Ruddy1 notes that the slab accounts for between 5% and 18% of the total construction cost. The lower end of the spectrum relates to office-type uses, the upper end to production facilities with special floor coverings. The cost of some heavy-duty metal or emery roofing can far exceed the cost of the panel itself.

REFERÊNCIAS 1. John L. Ruddy, „Evaluation of Structural Concepts for Buildings: Low-Rise Buildings“, BSCE/ASCE Structural Group Lecture Series Nr. MIT, Cambridge, MA, 1985. 2. BOCA National Building Code, 14ª ed., Building Officials and Code Administrators International, Inc., Country Club Hills, IL, 1999. 3. Metal Building Systems, 2ª ed., Building Systems Institute, Inc., Cleveland, OH, 1990.

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

FIGURE 12.38 A column designed without considering the size of the foundation.

FIGURE 12.39 Ending plate components.

373

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS 374

CHAPTER TWELVE

FIGURE 12.40 Typical slab control joint.

FIGURE 12.41 Typical ceiling construction joint.

4. NAVFAC DM-7.2, Erdfundamente und -strukturen, Department of the Navy, Naval Facilities Engineering Command, Alexandria, VA. 5. A. Kleinlogel, Rigid Frame Formulas, Frederick Ungar Publishing Co., New York, 1964. 6. A. Newman, „Engineering Pre-engineered Buildings“, Civil Engineering, September 1992, p. 58. 7. CRSI Design Handbook, Concrete Reinforcing Steel Institute, Schaumburg, IL, 2002. 8. Roy E. Hunt, Geotechnical Engineering Techniques and Practices, p. 548, McGraw-Hill, New York, 1986. 9. ACI-Ausschuss 318, Anforderungen der Bauvorschriften für Konstruktionsbeton (318-02) und Kommentar (318R-02), American Concrete Institute, Farmington Hills, MI, 2002. 10. ACI Committee 436, Suggested Design Procedures for Matching Shoes and Mats, ACI 336.2R-66, American Concrete Institute, Detroit, MI, 1966. 11. Edwin H. Gaylord, Jr. und Charles N. Gaylord (Hrsg.), Structural Engineering Manual, 2. Aufl., p. 5-66, McGraw-Hill, New York, 1979. 12. D. W. Lee und J. E. Breen, „Factors Affecting Anchor Bolt Development“, Research Report 88-1F, Project 3-5-65-88, Cooperative Highway Research Program with Texas Department of Highways und USA Bureau of Public Road, Center for Highway Research, University of Texas, Austin, August 1966. 13. R. W. Cannon, D. A. Godfrey und F. L. Moreadith, „Guide to the Design of Anchor Bolts and Other Steel Embedments“, Concrete International, Juli 1981.

FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS FOUNDING PROJECT FOR METAL CONSTRUCTION SYSTEMS

375

14. ACI Committee 349, Code Requirements for Concrete Structures Related to Nuclear Safety (ACI 349) Anhang B, American Concrete Institute, Farmington Hills, MI, 1997. 15. John G. Shipp und Edward R. Haninger, „Design of Headed Anchor Bolts“, AISC Engineering Journal, zweites Quartal, 1983. 16. Mario N. Scacco, „Design Aid: Anchor Bolt Interaction of Shear and Tension Loads“, AISC Engineering Journal, viertes Quartal, 1992. 17. James M. Fisher, „ Structural Details in Industrial Buildings“, AISC Engineering Journal, Third Quarter, 1981. 18. ACI Committee 302, Guide for Concrete Floor and Slab Construction, ACI 302.1R-96, American Concrete Institute, Detroit, MI, 1996. 19. Concrete Floors on Ground, Portland Cement Association, Skokie, IL, 2001. 20. Robert A. Packard, Slab Thickness Design for Industrial Grade Concrete Floors, Portland Cement Association, Skokie, IL, 1976. 21. Test Method Standard for Determining Floor Ebenheit und Ebenheit Verwenden des Systems Nummer F, ASTM E 1155-87, AST M, Philadelphia, 1987. 22. ACI Committee 117, Standard Specifications and Commentary for Tolerances for Concrete Construction and Materials, ACI 117-90, ACI, Detroit, MI, 1990. 23. Eldon Tipping, „Bidding and Building to F-number Floor Specifications“, Concrete Construction, Januar 1992.

REVIEW QUESTIONS 1 Compare the advantages and disadvantages of cap screws and L-shaped screws 2 Identify and explain design issues that should be addressed by screw designers. 3 Why do many engineers question the long-term viability of hairpin bars? 4 Using the tables in Appendix D, determine the approximate column reactions for a 60 foot wide rigid structure with a span, eaves height of 18 feet, and spaced 20 feet. The structure will be subjected to a live load of 20 psf on the roof and a wind speed of 80 mph, Exposure C, calculated according to ASCE 7-95. 5 Who is responsible for the design of the anchor bolt including the definition of minimum edge distances? 6 When determining column abutment sizes, what factors should be considered before selecting the manufacturer? 7 Which load combinations are most likely to determine the design of the foundation and anchor bolts? 8 List two types of lateral load-bearing devices attached to sidewalls that may require flared columnar piers. 9 List at least two challenges faced by designers of recessed slabs for metal building foundations.

FOUNDING PROJECT FOR METALLIC BUILDING SYSTEMS

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 13

SOME CURRENT DESIGN TRENDS

Metal construction systems are one of the youngest and most dynamic areas of the construction industry. New materials and metal applications for building design are constantly emerging, expanding the architect's choices. This chapter examines some of the latest trends in metal construction specification, some compelling design solutions and factors that continue to increase the competitiveness of prefabricated buildings.

13.1 FAÇADE SYSTEMS: MANSARDS AND ROOFING As metal building systems expand their acceptance in commercial, institutional and community settings, the formerly dull, utilitarian look of metal-clad gabled buildings is giving way to more interesting and diversified design solutions. Optical interest can not only be shown by the information given in Chap. 7, but also for various facade treatments ranging from simple roofing to sophisticated facade panels.

13.1.1 Roofs A functional and aesthetically pleasing roof is perhaps the most common facade treatment. The simplest way to construct a canopy is to provide a cantilevered extension of the primary structure at eave level and continue the roof construction into the canopy (Fig. 13.1). The eaves canopy works best for up to 10 feet of continuous, wide coverage that extends the full length of the building. For this solution to work, the building envelope must be visually compatible with the exposed cantilever rafter, which remains visible even with the soffit panels covering the underside of the canopy. A more sophisticated option is a flush-lined canopy, in which the entire framing is hidden (Fig. 13.2a). On the side walls, parallel to the purlins, this canopy is supported by a special cantilever at each column (Fig. 13.2b), while on the end walls the purlins can easily be extended beyond the wall line (Fig. 13.2c). A flush frame canopy, while elegant in design, is limited to a maximum width of 3 to 5 feet, or about one-half to one-third of the canopy from the eaves line. For an even more sophisticated treatment, a bullnose canopy can be specified (Fig. 13.3). A bullnose canopy looks and works best when it is some distance below roof level and attached to a contrasting wall material. Rather than being supported by an extension of the roof, it is supported by closely spaced frames attached to the building wall, with steel members shaped like a hat or duct between them. The supporting wall must of course be strong enough and is best made of masonry or concrete,

377 Conveyed from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

SOME CURRENT DESIGN TRENDS 378

CHAPTER THIRTEEN

FIGURE 13.1 Canopy in eaves with exposed rafters. (Constructive appendix systems.)

although it is possible to reinforce a structure of metal plates for this purpose. Bullnose panels should be curved as described in the next section. The spacing of the support structures is controlled by the bending capacity of the hat and channel sections. The structural properties of some representative top hat and channel sections manufactured by MBCI are shown in Fig. 13.4. The caliber of MBCI cuts is quite thin; similar sections of thicker metal are available from other manufacturers.

13.1.2 Parapets and mansards Facades and mansards appear so natural in prefabricated buildings that one might forget to specify and detail them separately. A vertical fascia and parapet panel, the most common type, is usually supported by a moment-resisting structure of its own that is rigidly attached to the main structure of the building. The primary structure must be designed for additional fascial loading. Some general details of this solution are shown in Fig. 13.5. Mansard-style bezels require few modifications to the details of the vertical panel (Fig. 13.6), but a completely different type of framing is required for the so-called double-curved brow panel (Fig. 13.7). A curved fascia combined with contrasting wall panels helped transform what could have been a basic prefabricated building into a modern-looking office (Fig. 13.8). For an even bolder design, a three-stage curved fascia (Fig. 13.9) can add spice to almost any building. A note of caution: Mansards and parapets can look great on metal buildings, but they must be specified with a full understanding of the potential hazards involved. Unlike gable roofs with free drainage, internal gutters can become clogged with ice or dirt. It is imperative that such systems be supplemented with overflow cuppers or culverts to remove standing water that could otherwise stress the roof structure. In cold weather, accumulated snow can pile up on railings and overwhelm purlins in outer bays. As in chap. 10, the failure of a single span can spread throughout the building and result in a total loss. The author examined a metal building with a parapet that actually depicted this scene, while none of the surrounding buildings without a parapet had collapsed. In general, it's best to avoid using railings on metal structures in snowy regions.

SOME CURRENT DESIGN TRENDS SOME CURRENT DESIGN TRENDS

FIGURE 13.2 Flush canopy with soffit board: (a) general appearance; (b) side panel; (c) end wall section. (Metal construction systems.)

379

SOME CURRENT DESIGN TRENDS 380

CHAPTER THIRTEEN

FIGURE 13.3 Bullnose Canopy with Soffit. (Central.)

13.2 CURVED PANELS As shown in the illustrations, well-proportioned curved panels can create an excellent visual impact. These panels have enjoyed great popularity since 1985, when Curveline, Inc. of Ontario, California brought to the United States the first patented curve crimping process for panels developed in the Netherlands. Crimp bending involves the incremental pushing and pulling of metal panels into rounded shapes in a computer-controlled process.1 Today, curved panels are used not only as fascias, mansards, and porches, but also as sidewalk roofs, decorative column covers, equipment shields, eaves, and even curved shapes for concrete. Curveline, Inc. remains the industry leader, offering the broadest line of curve products. The company can form slabs from 2 to 30 feet in length and a maximum width of 5 feet. Plate depth can range from 3Ⲑ4 to 6 inches, thickness from 0.016 to 0.052 inches (29 to 18 gauge). Curve configuration, Curveline can create complex and impressive multiple and S-curves (Fig. 13.10). Visible fastener plates are best suited for bending, although some products with hidden fasteners can also be bent. The rippled curvature process approximates the true curvature through many short chords, an appearance some dislike. Where a smoother line is desired, the "tangle-free" appearance of curved panel ribs can be avoided by rotating the panels so that their flats and not the ribs face outward. Each bent panel manufacturer has their own standards for minimum bend radii. Typically, the deeper the plate, the larger the radius. Plates made from thin materials, particularly high-strength steels, typically require a larger bend radius.

SOME CURRENT DESIGN TRENDS SOME CURRENT DESIGN TRENDS

FIGURE 13.4 Section properties of hat and channel sections. (MBK.)

381

SOME UPDATED DESIGN TRENDS 382

CHAPTER THIRTEEN

FIGURE 13.5 Vertical stripes. (Metal construction systems.)

SOME CURRENT DESIGN TRENDS SOME CURRENT DESIGN TRENDS

FIGURE 13.6 Mansard panels. (Metal construction systems.)

383

SOME CURRENT DESIGN TRENDS 384

CHAPTER THIRTEEN

Fig. 13.7 Doubly curved brow fascia. (Central.)

The most bendable steel composition, according to Curveline, is ASTM A 446 Grade D G-90 carbon steel with a tensile strength of 50,000 lb/in2.3 Galvanized steel, aluminum and stainless steel panels can be bent. Curved panels are structurally more efficient than straight panels and can often be made from thinner metal, resulting in material savings. For continuous support, curved chords and purlins can be made in the same source to conform to the contour of the slab. Wherever curves follow straight panels, e.g. B. on building corners, depending on the provider, a separate round piece may or may not be required. According to Curveline, Inc., a separate curved connector is generally not required and the curve can be built into one end of the straight panel. An additional seal is also dispensed with for aesthetic and functional reasons. A notable exception is the beveled corner (Fig. 13.11), which is best made separately. When factory bending is not practical, on-site bending is possible. Some companies, such as B. Berridge Manufacturing of Houston, Texas offers both roll forming and sheet metal bending on site. Alternatively, a rounded corner can be obtained without bending if the panel is bent parallel to the ribs, a relatively simple operation. While curved panels are visually appealing, bending can seriously degrade the surface finish of the panel. Some manufacturers who entered the curve business in the 1980s might do so

SOME CURRENT DESIGN TRENDS SOME CURRENT DESIGN TRENDS

385

FIGURE 13.8 A curved facade adds interest to an office building. (Curveline, Inc.)

not overcome technical difficulties and survive. To date, some of the major manufacturers such as Butler not only do not offer curved panels, they also advise against having their products bent by others. For the same reason, many architects avoid specifying curved panels in corrosive climates. Before specifying bent crimp plates, designers should consult with some of the local manufacturers involved in this business for information on available plate profiles, finishes, bend radii and product warranties. It is instructive to look at some of your previous projects, preferably at least several years old, to look for signs of corrosion. Inspect for incomplete bends and crimps in the panels, proper crimping of all trim pieces, and acceptance of tolerances. Besides the above companies, some other companies involved in the manufacture of curved panels are ATAS Aluminum Corp. from Allentown, Pennsylvania; Floline Architectural Systems of St. Louis, Missouri; Petersen Aluminum Corp. from Elk Grove Village, Illinois; Centria of Moon Township, Pennsylvania; and BHP Steel Building Products USA, Inc. of West Sacramento, California.

13.3 HOUSES IN STEEL CONSTRUCTION Always looking for new opportunities, the metal construction industry began at a spectacular pace to supply prefabricated frames for housing. According to AISI, 13,000 steel-frame houses were built in this country in 1993, compared to just 500 built in the previous two years. In 1994, 40,000 steel houses were scheduled to be built in North America.4 Historically, steel was prohibitively expensive for housing, but when lumber prices rose sharply in the early 1990s, steel suddenly became cost-competitive. Next

SOME UPDATED DESIGN TRENDS 386

CHAPTER THIRTEEN

Fig. 13.9 Fascia curved in three stages. (Central.)

Despite the sometimes temporary price development, steel has some real advantages over wood: It is non-flammable, dimensionally stable, does not warp or rot and is not attacked by termites. The main disadvantage of steel is its poor thermal properties. Surely steel houses have been tested before. Peter Naylor's "portable iron houses", described in chap. 1, were offered to Californian gold rush fortune hunters as early as the mid-19th century. A century later, after World War II, the United States government granted the Lustron Corp. from Columbus, Ohio to build steel houses. Lustron Homes were constructed from steel frames and clad with porcelain-coated steel exterior panels. Also the inner walls and

SOME CURRENT DESIGN TRENDS SOME CURRENT DESIGN TRENDS

387

Figure 13.10 Complex curves adorn the US space warehouse in Huntsville, Alabama. (Curveline, Inc.)

The ceilings were made of steel. According to a website dedicated to Lustron Homes, these homes were made in 1949 and 1950; They sold for approximately $7,000. There are three methods to build the steel house. The first is to simply replace the steel posts and beams with wood, essentially following traditional stud and beam construction with either 16 or 24 inch spacing. Everything else - roof, panelling, doors, windows - remains the same as in a timber frame house. This method allows for easy frame-to-steel conversion in both standard and custom homes; It is undoubtedly used in most steel frame houses. The second framing method is panel construction: the framing is constructed from pre-assembled steel beam wall panels and roof trusses. Both the studs and the trusses are spaced 32 to 68 inches apart on center, with under girdles and under purlins in the hat area, similar to Fig. 13.4 inclusive between them. Despite claims of efficiency, this system is structurally quite complex and may require more bracing and anchoring than others.5 It is unfamiliar to both traditional homebuilders and prefab builders, and can create a great deal of confusion in the workplace.

SOME UPDATED DESIGN TRENDS 388

CHAPTER THIRTEEN

FIGURE 13.11 Beveled corner. (Central.)

The third method is to construct an actual, albeit small, prefabricated building, complete with the usual main structures, braces, purlins, bar braces, and metal roofing. The shaft spacing of such prefabricated structures varies between 6 and 10 feet. A rigid frame gabled building may be appropriate for a large modern home with an open floor plan and high ceilings. However, the system is unfamiliar to most home designers and builders. It is also so different from traditional construction that it is difficult to present it as a framing replacement for already complex projects. Additionally, commonly available components of commercial-style metal building systems, such as doors, windows, siding and roofing, may not meet homeowner's expectations. Since few homeowners dream of a house clad in metal siding, a brick or wood facade is desirable. These traditional surfaces typically must be supported by 3Ⲑ4 inch plywood panels that can span about 4 feet between metal supports. The thermal bridging problem can be solved by installing rigid insulation on the outside of the steel studs. Polyisocyanurate insulation offers excellent insulation value (see Chapter 8) and is available as an insulating coating. Rigid insulation can also be installed in ETICS external walls, as described in Chap. 7. A typical high quality exterior wall may consist of 6" steel studs covered with 3Ⲑ4" plywood and 1" insulation board coated with an EIFS finish. The stands can be filled with 6-inch fiberglass insulation covered with heavy-wall vapor-retardant plastic and 5Ⲑ8-inch drywall. 2 seconds. 2.10.

13.4 INDUSTRY COMPUTING In short, what contributed to the transfer of old "ready-made" designs to modern metal construction systems? computers! Heavy reliance on these machines made it possible to manufacture metal structures.

SOME CURRENT DESIGN TRENDS SOME CURRENT DESIGN TRENDS

389

Manufacturers to ditch the old menu of some predefined build settings in favor of unlimited theme options. In fact, almost all metal buildings built today are specifically designed for a specific project. While architects celebrate the new freedom of design, homeowners are happy about quick price offers. Advanced software allows quotes to be created in just 5 minutes, a task that used to take days. Builders, on the other hand, are happy about the fast delivery times: it is not uncommon for the delivery time to be reduced to 5 weeks, a task that recently took at least 3 months. Leading manufacturers are racing to develop the most comprehensive and easy-to-use software systems that, based on input data, generate a price quote, design calculations, manufacturing drawings, and even presentation materials. Investing in such premium systems gives the biggest players in the industry a clear advantage over smaller stores. Not surprisingly, the aptly named Butler Advantage System won first place among hundreds of entrants in the Manufacturing and Distribution category at the Windows World Competition in April 1995, as well as other awards. (Windows World is sponsored by Microsoft Corp. and Fortune and Computer World magazines.) Over 1000 Butler creators are reported to be using this software. VP Edifícios has developed its own Command Information System, which has excellent graphic capabilities and allows orders to be placed 24/7. Some examples of program outputs are given in Chap. 9. Smaller manufacturers who cannot afford large investments in software development can purchase one of the many commercially available computer programs such as B. that offered by Loseke Technologies, Inc., The Colony, Texas, or Metal Building Software, Inc. , from Fargo, N.Dak. Computerization allows manufacturers to centralize labor, accounting, and inventory control systems — and create more accurate quotes. It also enables savvy manufacturers who are willing to make the investment to compete in a world with many building codes, languages, and units of measurement to respond more quickly to market changes. Technological advances are also likely to impact the construction side of the industry. The Standard Commodity Accounting and Tracking System (SCATS) allows barcode tracking of every piece of steel in the project. Some Louisiana structural steel fabricators are already using SCATS to track materials on accelerated projects. With their legendary build speed, metalwork system assemblers can't be far behind.

13.5 METAL MULTI STORY BUILDING SYSTEMS Although the vast majority of metal building systems are single storey, the market for multi storey buildings offers tremendous growth opportunity. With land costs rising and space constrained in built-up areas, multi-storey prefabs are a logical answer for homeowners who need more than one floor of floor space but still want to reap the benefits of the construction industry. metallic . According to some estimates, the prefabricated structure can save about 15% of construction costs in a four-story commercial building. Metal building systems using truss beams work well with traditional metal decking and concrete infill floor frames. Systems based on lightweight C and Z sections may face some acceptance issues related to their fire rating, deflection and vibration characteristics. Designers wishing to specify multi-storey prefabricated buildings should first inquire whether local dealers have experience with this type of construction. Some manufacturers offer multi-story steelwork systems that include open-web steel beams supported by moment-rigid frames running in two directions (Figs. 13.12 and 13.13). These systems closely resemble their conventional cousins in composition and appearance. In fact, the office building of Figs. 13.12 and 13.13 is faced with an elegant veneer brick, as shown in Fig. 1.5 of Chap. 1.

SOME CURRENT DESIGN TRENDS 390

CHAPTER THIRTEEN

FIGURE 13.12 Multi-story steelwork system with rigid moment-bearing frames and open web girders. (HCI Building Systems, Inc.)

REFERENZEN 1. Marc S. Harriman, „Full Metal Jacket“, Architecture, März 1992. 2. Steve Moses, „Curved Panels Expand Design Possibilities for Metal Products“, Metal Architecture, Januar 1994. 3. Custom-Curving of Profiled Metallplatten, Curveline, Inc., Ontario, CA. 4. Richard Haws, „Steel Framing: The Right Choice in Residential Construction“, Architectural Specificationr, September 1994. 5. Juan Tondelli, „Steel Homes Slowly but Steadily Penetrating Residential Market“, Metal Architecture, Dezember 1993. 6 Sam Milnark, „Steel macht ein Statement auf Catawba Island“, Metal Home Digest, Ausgabe von 1994.

REVIEW QUESTIONS 1 What types of metal plates are best for bending? 2 What are the main advantages of multi-story metal building systems? 3 How is the canopy made? At which edge of the building can it make sense to build?

SOME CURRENT DESIGN TRENDS SOME CURRENT DESIGN TRENDS

391

FIGURE 13.13 Details of the outer rigid structure shown in Fig. 13.13. 13.12. (HCI Building Systems, Inc.)

4 List some of the concerns in defining parapets for steel buildings in snowy regions. 5 What simple steps can you take to prevent water from pooling around clogged indoor gutters?

SOME CURRENT DESIGN TRENDS

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 14

RECONSTRUCTION AND RENOVATION OF METALLIC BUILDING SYSTEMS

14.1 INTRODUCTION Building reuse and refurbishment is becoming increasingly popular and the volume of refurbishment work is now rivaling that of new construction. The principles of metal construction, such as cost efficiency and responsibility from a single source, apply not only to new construction, but also to building renovation. As the first generations of prefabricated buildings near the end of their useful lives, they may be refurbished or partially replaced with new metal structures. Some components of metallic building systems have their place in the renovation of conventionally constructed buildings. A common problem faced by low-rise building owners is leaking roofs. When we talk about renovating a building, the first thing that comes to mind is roof renovation or roofing renovation. Therefore, this chapter will first focus on conventional and metal roof renovations with metal roofs. External skin modifications are discussed next, followed by primary and secondary frame reinforcement. The chapter ends with a discussion of some of the difficult problems involved in adapting existing prefabricated buildings to new conditions of use.

14.2 RETROFITTING ROOFS WITH METAL BUILDING SYSTEMS 14.2.1 The Roofing Problems No building problem seems to cause more trouble than a leaking roof. Dealing with occupants who report a leak—a problem that's easy to spot but difficult to fix—is one of a building owner's most frustrating tasks. After a few rounds of thankless repairs, a complete roof overhaul seems like the only solution left. Not surprisingly, roof renovation work accounts for two-thirds of all roof projects in this country. Why do roofs fail so quickly? The roof is the hardest part of the building, protecting it from the scorching sun in summer, from snow in winter and from rain all year round. Every storm attacks the roof, first literally trying to lift it off the building and then throwing it back down. Ultraviolet radiation shortens the life of unprotected monolayer membranes and slowly robs them of their elasticity and strength; It also causes damage to other types of roofs. Most conventional low-rise buildings have flat or nearly flat membranes or recessed roofs, a solution that was popular until a few years ago. A minimum slope of 1Ⲑ8:12 was considered sufficient for drainage; probably was - in theory. In real life, however, the foundations of buildings lay a 393 Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

MODIFICATION AND REFURBISHMENT OF METAL CONSTRUCTION SYSTEMS 394

CHAPTER FOURTEEN

somewhat uneven, roof beam heights vary slightly due to manufacturing and installation tolerances, and the beams and deck flex under load. Additionally, point loads from HVAC equipment, overhead wires, lights, and the like cause some structural roof members to deflect more than others. All of these factors can result in the actual roof profile being far from what is assumed - with some areas of the roof having no slope at all - and lead to water accumulation. Roofs that are not designed to remain submerged for long periods of time, like some asphalt-based products, can slowly disintegrate and eventually leak. Other factors that lead to roof failure include localized damage from careless foot traffic or equipment maintenance, clogged roof drains - which in turn lead to ponds - and poorly secured roof penetrations. Deterioration usually begins at spots with ridges, expansion joints, and poorly attached gravel screens. Regardless of the source, roof leakage can result in saturation and destruction of fiberglass insulation, staining of surfaces, and corrosion of the roof deck. If immediate remedial action is not taken, the damage can progress to the point where the roof is beyond repair, so disassembly and replacement is the only solution.

14.2.2 Roof Renovation Options A popular choice for covering conventionally constructed roofs is single ply membranes, particularly the fully bonded or mechanically bonded lightweight varieties. This material is not without its drawbacks. Covering an old tar and gravel roof with a single-ply membrane usually requires removing all of the gravel. This tricky operation, if handled improperly, can result in a poorly excavated roof that will need to be covered with protective sheets or even torn down entirely. If the roof pitch was previously unsuitable, it can only be changed with expensive gradient insulation. And, as mentioned earlier, in sunny locations, exposure to sunlight causes unprotected membranes to quickly collapse and warp this system. Another increasingly popular option is metal roofing. This solution offers numerous advantages. As in chap. 6, Metal roofing is available in a variety of finishes including polyvinylidene based coatings which are extremely durable and UV resistant. Vertical seam covers with sliding clips can handle thermal expansion and contraction better than membranes. Even with pitches as low as 1Ⲑ4:12 for structural panels and 3:12 for architectural roofs, water can drain faster than almost flat roofs. On steeper slopes, as described in Chap. 6 should do the coverage even better. The necessary gradient can be achieved by erecting a light construction on the old roof. The combined weight of the metal roof and new structure generally does not exceed 2 to 4 lb/ft2, making this one of the lightest systems on the market. This small additional load can often be safely carried by the existing roof structure, while heavier systems such as built-in roofs would overwhelm it. If the existing roof structure does not have excess capacity, a rafter or truss system can be erected between the new upper supports on top of the existing building supports. Some experienced architects believe that properly designed and constructed metal roofing will last 40 years.1 Although metal roofing can initially cost more than competing systems, their exceptional durability combined with ease of maintenance often makes metal a winner in life cycle cost comparisons. Because they are recyclable, metal roofs also deserve an environmental rating. Metal roofs are very useful when it comes to replacing an existing slate or tile roof that is supported by an aging and undersized roof structure. These roofs can benefit from a Bermuda metal roof with the profile of slate, trepidation or tile, or a PVDF coated metal roof tile product designed to resemble traditional materials.

14.2.3 Tear or Cover? The decision to keep the existing roof or remove it often depends on the moisture content of the existing roof system. It is clear that some water has penetrated the roof insulation as a result of the previous leaks; The only question is how much water? Roof renovation over existing roofs and wet insulation possible

RECONSTRUCTION AND RENEWAL OF METALLIC CONSTRUCTION SYSTEMS RECONSTRUCTION AND RENEWAL OF METALLIC SYSTEMS

395

invite various problems. In addition to the problems of reduced insulation performance and corrosion of existing decking mentioned above, trapped moisture can cause unpleasant odors and mold growth, leading to serious indoor air quality problems. Additionally, adaptive fasteners that enter wet areas can eventually corrode and undermine the integrity of a newly designed roof system. The degree of water saturation can be determined by a moisture measurement carried out by the planner or by professional advice. The latter can provide more reliable results as specialized companies are likely to use advanced testing methods such as infrared thermography, capacitance and nuclear backscatter.2 The survey produces a rough sketch of areas of wet insulation and determines the level of saturation, which can be confirmed by removing a few cores of insulation. Only then can the extent of the problem be reasonably assessed. A common solution to the problem of trapped moisture is to install multiple vents and wait for the moisture to escape before closing the last vent. However, this approach only works at very low humidity levels. If the existing roof and insulation are completely saturated with water from frequent leaks, ventilation through some holes may not be adequate, especially if a structural ceiling or vapor barrier restricts the downward movement of moisture. Studies show that under such circumstances it would take 30 to 100 years for the insulation to dry out, even with vents installed.2 A better course of action would be to remove the roof and wet insulation. Tobiasson3 states, “In most cases, wet insulation should be considered a cancer that must be removed prior to roof renovation.” He points out that each inch of saturated insulation can contribute up to 5 lb/ft2 of permanent stress – a significant one Quantity. A moisture survey that indicates multiple areas of wet insulation should be taken seriously: complete removal may be the only reasonable option left. A structural roofing survey is also extremely helpful. Persistent leaks may have resulted in widespread deck corrosion beyond repair. Likewise, the presence of some potentially corrosive roof components could have damaged the deck. An example is phenolic foam roof insulation manufactured in the late 1980s through 1992. It has been reported that this type of insulation, when wet or damaged, can contribute to the corrosion of metal decks, sometimes to the point where walking becomes unsafe. Replacing the roof deck is serious business as it opens up the interior of the building to the elements and impacts operations. There are arguments against removing the roof entirely, primarily the high cost. The bill for disposal of removed materials containing potentially hazardous waste such as asbestos screens can also be significant. The pros and cons of the two approaches require careful consideration. Interestingly, the 2 billion square feet of roof repair work completed in this country annually is split evenly between removal and coverage.2 Of course, if your local building code prohibits adding another layer of roofing, the decision of whether to remove or re-cover the roof is an easy one target.

14.2.4 The Design Responsibility Problem Who determines whether an existing roof is structurally capable of withstanding the additional, albeit modest, load of a roof renovation? As we have already discussed, manufacturers of metal building systems are not willing to go beyond the design of metal components; They usually disclaim any responsibility for evaluating the existing roof structure and its ability to support additional loads. Hire a local structural engineer to analyze the existing roof structure, they suggest. The problem is that the engineer can readily prove only the average uniform load-bearing capacity of the roof, a calculation that is only applicable if the new roof is simply laid on top of the existing one. However, any change in slope requires a new ("retrofit") structure supported by a few discrete supports that transfer concentrated, uneven loads to the existing structure. At the evaluation stage, the engineer often has no way of knowing which manufacturer will be selected for the work and what the spacing between the columns will be. If the manufacturer is already on board, all the better. Otherwise, engineers can choose one of the popular systems, such as B. that described below, base their analysis on this system and require the contractor to follow it. Or they can assume the worst case scenario and use a

RECONSTRUCTION AND RENOVATION OF METALLIC BUILDING SYSTEMS 396

CHAPTER FOURTEEN

costly approach in which the new truss only has to support the existing supports and the existing roof structure is completely ignored. Alternatively, they may require the new supports to be spaced to achieve approximately even loading - also not the most cost-effective solution. To complicate matters further, any significant change in roof pitch increases the vertically projected area of the roof and results in a higher wind load on the building. Now the entire lateral resistance system of the building may need to be re-evaluated, involving the engineer even more deeply in the design. In order to be prepared for such complications and to be able to make informed design decisions during the preparation of construction documents, designers need to educate themselves about the types of metal roof support structures available.

14.2.5 Supporting structure for pitch changes The proposed roof structure shall have the same level of strength and rigidity as any other metal roof structure. In practice, this means that the spacing of the new purlins ('retrofit') is likely to be limited to 5 feet or more (Fig. 14.1). In the areas of the "raised corners" in the area of the eaves, the roof pitch and the ridge (with certain roof pitches) a smaller distance, about half the width, is required to withstand the increased wind load there. Likewise, the purlin distance is reduced in areas with potential snowdrift accumulation. Wherever the existing roof construction remains, the new construction not only withstands the wind and snow loads on the new roof, but also the wind loads on the new gable walls. Therefore, two types of bracing are required for stability: vertically between the structural mullions and horizontally in the plane of the new roof to act as a membrane. Membrane action can be provided by rod or angle bracing, by steel decks, or by certain types of solid covering (but not generally vertical seam covering, as discussed in Chapter 5). Another important point is the lateral reinforcement of the new purlins. When spaced and sectioned, the verticals of the structure provide the necessary reinforcement, but this may not be the case if the supports are widely spaced. As in chap. 5, manufacturers' design practices for lateral support of purlins vary widely, ranging from the conservative to the ignorant. To ensure consistency of design assumptions among bidders, the owner's requirements for acceptable membrane construction and purlin reinforcement methods must be specified in the contract documentation. Proper anchoring to the existing structure is vital to prevent the new metal roof from exploding, which acts like a giant sail erected on top of the existing building. Anchorage details should be designed by the metal system manufacturer and carefully considered by the approval engineer. Details should be tailored to the existing roof structure, rather than showing how the new fasteners end up in a mass of concrete, a simple but pointless "solution" all too often touted. If the existing roof structure - not just the roof - needs to be removed for easier attachment to the existing structure or due to excessive corrosion, consider replacing it with a new panel or horizontal brace to provide lateral support for the roof. existing roof beams and to restore the existing roof skin. Roof details for new work, such as B. Clip design, top attachment, placement of stepped expansion joints and the like remain the same for new roofs (see Chapter 6). The new 'attic' needs to be carefully evaluated for fire safety, ventilation and egress requirements.

14.2.6

Determining New Support Points The positions of new roof supports are determined by the type and spacing of the existing load-bearing roof elements. New supports are best placed directly over existing roof purlins, whether the purlins are steel, concrete or wood. This approach avoids overlays on the existing roof deck, which could corrode or be damaged by water leakage, even when theoretically appropriate. The desired roof pitch is achieved by varying the height of the new supports. All variable height columns must be pre-cut to exact size unless licensed to use adjustable brackets first

RECONSTRUCTION AND RENEWAL OF METALLIC CONSTRUCTION SYSTEMS RECONSTRUCTION AND RENEWAL OF METALLIC SYSTEMS

397

FIGURE 14.1 Option 1: Retrofitted structure with columns placed over existing roof thirds and spaced no more than 5 feet apart at center. (MBK.)

received from the patentee. Height-adjustable columns—no matter how accomplished—are protected by several patents held by the Re-Roof America Company of Tulsa, Oklahoma, which is enforcing their rights. A grid of purlins and columns no more than 5 feet apart in either direction, and perhaps closer together in some areas, is ideal. This is only possible if the existing purlins are also 5 feet apart or less (Fig. 14.1). The actual spacing of new support columns along the length of the purlin may depend more on the type and span of the existing purlins than the restrictions imposed by the new framing. For example, if the existing roof is framed with batten rafters and steel decking, the new studs should be placed directly over the points of the rafter panels and spaced a quarter of their span or even closer at points to allow them to match and carry uniformly. The reason: Due to their proprietary design, the only available structural information about beams is typically their uniform load capacity; different

RECONSTRUCTION AND REMEDIATION OF METALLIC BUILDING SYSTEMS 398

CHAPTER FOURTEEN

hot-rolled girders, bar girders cannot easily be tested for point loads. Keep this nuance in mind when a bidder begins to argue that such close support spacing is not necessary - the bidder is not responsible for assessing the strength of the existing roof. If you are supporting a support column directly onto the existing corrugated steel deck, remember to check the buckling strength of the deck ribs. A support plate or channel is often required to distribute the support load over several shafts, as shown in Fig. 14.1. In a more common - and difficult - situation, existing purlins are not 5 feet apart. In this case, a different design approach is required, providing a grid of base support members, typically C and Z purlins, spaced 5 ft 0 on center perpendicular to the existing roof purlins. The new retrofit supports support these bars and support the retrofit purlins (Fig. 14.2). Note that backplates are provided under the base support members and that the retrofit posts are arranged on a checkerboard. Lateral stability of the base support members can be achieved by properly attaching them to the existing roadway and by properly bracing or locking between the attachment points.

FIGURE 14.2 Option 2: Attaching the structure over existing purlins - a general case. (MBK.)

RECONSTRUCTION AND RENEWAL OF METALLIC CONSTRUCTION SYSTEMS RECONSTRUCTION AND RENEWAL OF METALLIC SYSTEMS

399

14.2.7 Lateral bracing requirements As mentioned earlier, the bracing requirements for light C and Z purlins are surrounded by some controversy and misunderstanding. We recommend a well-spaced support for these limbs, perhaps 5 to 6 feet apart. It can be assumed that the retrofitted supports will interlock both the upper and lower purlins, but reinforcement of the purlin above and between the supports is still required. Some manufacturers appear to agree and provide such reinforcement on both flanges of retrofit purlins to a maximum of 5 feet on center when the supports are spaced more than 5 feet apart on center and there is a permanent splice roof fig. 14.3. In accordance with our in Chap. 5, we prefer to see slight angles or channels able to act under pressure rather than strapping only under tension. In addition to the purlin reinforcement, transverse reinforcement is also required for the transverse load resistance of the entire arrangement. The main purpose of the bracing is to provide lateral stability to the new structure and to transfer wind loads to the existing building structure. The exact configuration and spacing should be left to the discretion of the manufacturer, but at least some bracing should be provided at regular intervals. Typical bracing requirements are shown in Fig. 14.4; some typical details are shown in Fig. 14.5.

14.2.8 Reroofing over an existing metal roof An old metal roof that is damaged beyond repair can be considered for reroofing. If there is sufficient gradient, no new construction is necessary; The new cover can be mounted directly over the old one. A common situation is vertical seam metal panels plus fiberglass insulation installed on an old roof. Here, the only new supporting structure consists of light top-hat rails - sometimes even two by four made of pressure-treated wood - which are arranged directly above the existing purlins (Fig. 14.6). To enhance the insulation, thermal spacer blocks can be installed over the hat channels where the insulation is compressed. A weak point of this project is the existing metal roof: if it corrodes or is structurally inadequate for the new loading - two very common scenarios - it will hardly be able to provide adequate support for the structure. A direct attachment to the purlins using a filling material that matches the roof profile is the only solution. One product that allows for this attachment is the Roof Hugger*, which is basically a low Z purlin with a notch on the bottom flange and a core that fits into the existing corrugations on the roof (Fig 14.7). According to the company, this product has proven itself in numerous applications and continues to grow in popularity.

14.2.9 Construction Details Each manufacturer involved in roof rehabilitation applications has developed their own details for different conditions, as above and in Fig. 14.8. (Data from the MBCI4 catalog are used for consistency.) The data given here only covers the situation when the existing purlins are perpendicular to the new roof pitch; Slightly different details are used when the existing structure runs parallel to the new slope. However, the general concepts discussed in this chapter should be applicable to any manufacturer and roof condition. A design detail that should not be overlooked is additional insulation treatment that may be required to meet code required U-values. Insulation made of hard foam or glass fiber fleece is often placed on the existing roof between the new supports. The problem designers face is how to provide a vapor retardant in such insulation. If the existing roof system has

*Roof Hugger is a registered trademark of Roof Hugger, Inc.

RECONSTRUCTION AND RENOVATION OF METALLIC BUILDING SYSTEMS 400

CHAPTER FOURTEEN

FIGURE 14.3 Contraventamento of the mothers of the retrofit. (MBK.)

RECONSTRUCTION AND RENEWAL OF METALLIC CONSTRUCTION SYSTEMS RECONSTRUCTION AND RENEWAL OF METALLIC SYSTEMS

FIGURE 14.4 Typical longitudinal bracing for embankment formation. (Transverse reinforcement is not shown for clarity.) (Butler Manufacturing Co.)

FIGURE 14.5 Typical bracing details. (MBK.)

401

RECONSTRUCTION AND RENOVATION OF METAL BUILDING SYSTEMS 402

CHAPTER FOURTEEN

FIGURE 14.6 Reconstruction of an existing metal roof with the correct pitch. (Courtesy of the NAIMA Steel Construction Committee.)

Insulation with a proprietary vapor retarder, or if the roof remains essentially intact, the new vapor retarder probably isn't needed. Conversely, if the existing cover was cut for ventilation purposes, this may be the case. A similar situation occurs when new rigid insulation is mechanically attached to an existing roof. A profusion of resulting holes could compromise the vapor retardant properties of the old membrane, although we're not sure if this will actually happen.5 In any case, perhaps the worst situation of all arises when there are multiple vapor retardants at the top and bottom of the membrane existing roof structure. The insulation present between the relatively impermeable barriers tends to absorb the moisture that seeps in through the inevitable imperfections of vapor retarders, but does not lose it through evaporation. Unfortunately, it is quite difficult to design a roofing system completely free of this problem unless the existing roof is removed. Vapor migration issues deserve careful consideration by the design professionals involved. Another common issue is the treatment of existing rooftop HVAC components. With the new slope retention structure, all existing ventilation ducts on the roof can be extended. When replacing a metal roof, it can be installed over existing vent pipes and other penetrations, with new panels folded around them. Special grooved aftermarket boots are available for this condition (Fig. 14.9). The exhaust air fans can also be moved to the new roof area and supported on suitable curbs. But what about large devices such as chillers, air conditioners or cooling towers? The new lightweight aftermarket structure is clearly not strong enough to support it. It is of course possible to extend the structural framework to a higher height, but the cost can be prohibitive. It is also possible to treat the space between the new and existing coverings as a kind of mechanical covering and fill the new gable walls with shutters or large openings. In this case, the HVAC system can stand on top of the existing roof and be fully or partially enclosed by the retrofit roof.

RECONSTRUCTION AND RENEWAL OF METALLIC CONSTRUCTION SYSTEMS RECONSTRUCTION AND RENEWAL OF METALLIC SYSTEMS

403

Abb. 14.7 Telhado-Orthese. (Roof Hugger Inc.)

It is advisable to stipulate that the existing roof penetrated by columns or post-connections be patched to provide temporary leakage protection.

14.2.10 What the manufacturer needs to know for metal roof renovation Contract documents for metal roof renovation must contain at least the following information: ● ● ● ● ● ●

●

The current building code and its issue; Desired UL and insurance ratings Design of live, snow, wind and seismic loads and load combinations Existing dimensions and construction details (some original drawings can be attached) Plan, pitch and configuration of the proposed roof Desired roof type, including profile and finishing Structural requirements for the new support structure, such as support spacing, bracing and roof panel construction.

It is advisable to request the submission of detailed construction calculations and shop drawings, accompanied by certification from a professional engineer that the new roof meets contract requirements.

RECONSTRUCTION AND RENOVATION OF METAL BUILDING SYSTEMS 404

CHAPTER FOURTEEN

FIGURE 14.8 Typical eaves and bank details for bank formation. (MBK.)

14.2.11 Refurbishment of Existing Metal Roofs Like any other type of roof, metal roofs have a limited lifespan that can be shortened by exposure to the elements or improper installation. Metal roof damage can manifest as seams pulling apart, fasteners springing back, or rust spreading from the panel edges to the rest of the roof. All of this leads to leaks. Sometimes the paint fails first and the roof just looks bad.

RECONSTRUCTION AND RENEWAL OF METALLIC CONSTRUCTION SYSTEMS RECONSTRUCTION AND RENEWAL OF METALLIC SYSTEMS

405

FIGURE 14.9 Fitting a new covering over existing roof penetrations, overlapping and sealing the panels and using a retrofit split hose. (Central.)

When the builder identifies a suspected leak on the roof, the first impulse is usually to cover the suspected area with tar or asphalt. These simple on-the-spot repairs don't last long as they are often poorly done and can trap water and cause corrosion. Instead, a leaking or deteriorating roof should be systematically evaluated to get an overall picture of its condition and pinpoint other potential problem areas. The first step in solving the problem is a roof inspection. If the existing roof pitch, insulation, and other parameters are satisfactory, and the problem is only the deteriorated finish, a full roof rehabilitation job may not be necessary. A simple cladding can be enough to drastically improve the image of the building. The most problematic parts of the roof – flashings, gutters, fascia flashing and roof penetrations – need to be inspected prior to cladding and replaced if necessary. Obviously, rusting metal cannot be helped by a new coat of paint. The roof support structure and fasteners are also vulnerable and should be inspected for signs of corrosion. The most critical areas in this regard are the edges of the purlin holes made by the roof fasteners where the protective coating on the purlins is broken and bare metal is exposed. The only protection given to this area is the compressible mounting gasket, if fitted, but gasket performance is highly dependent on the skill of the installer. Widespread corrosion around the fasteners can dramatically weaken the roof's anchorage and require the installation of new fasteners, challenging the cost-effectiveness of the entire cladding project. The actual refurbishment process, much like refinishing, consists of removing all or part of the existing paintwork (or sometimes just pressure washing it), priming and repainting with a new substrate-compatible material. The field is rife with coating products specifically designed for aging metal roofs. Many of these products are based on elastomer membranes that seamlessly cover the roof; Some use seamless foam systems consisting of sprayed polyurethane foam insulation covered with a sprayed silicone membrane. Prudent planners thoroughly examine the actual performance of these products before selecting them. A detailed explanation of this process, complete with a flow chart, is provided by Newman.6

RECONSTRUCTION AND RENOVATION OF METAL BUILDING SYSTEMS 406

CHAPTER FOURTEEN

14.3 EXTERIOR WALL REPLACEMENT AND REPAIR Metal-clad walls age in the same way as metal roofing: surfaces deteriorate, seams crack, fasteners loosen and corrode. Sooner or later, the siding will scream for attention. Luckily, walls tend to be easier to update than roofs. The simplest type of wall renovation is of course repainting, a relatively simple and well-known process. A more complicated situation arises when walls are damaged beyond repair and need to be replaced with a product that works with the existing belt system. The structural evaluation of existing belts should be the first task. Are the straps the right size? Is there excessive corrosion? And the side reinforcement? Links? These questions must be answered satisfactorily before a liner change. If the straps are too far apart or missing in some way, it may be more cost effective to add them and use the standard fairing rather than specifying extra strong replacement panels that can span large distances. Most metal wall replacement projects involve new windows, doors, flashing and maybe even molding around openings to provide a coordinated exterior wall system. In fact, the corrosion of door and window frames, together with rusted side rails, is often one of the first signs of aging in metal construction. Simply repainting the wall rarely solves the problem, as rust tends to seep through the new paint (Fig. 14.10). In this situation, it is better to replace the wall. A special, albeit unusual, situation arises when existing metal panel walls have to be converted into 'hard' walls. In this case, the existing building is likely not rigid enough to laterally support such walls and may need to be replaced. In order to keep operations running smoothly in a medium-sized building, it may be possible to construct the new "hard" outer walls and the new roof construction outside the existing building envelope and later to remove the old building piece by piece.

14.4 STRENGTHENING THE STRUCTURE FOR CHANGES IN LOAD CONDITIONS Changes in ownership, occupancy, manufacturing process, or mechanical systems often result in some changes in structural loads. Assessing such changes in conventionally designed buildings is a relatively easy task for structural engineers. However, prefabricated buildings need to be "re-engineered" or at least evaluated for the new conditions, which is far from easy given the proprietary nature of the structure. Some common examples: ●

●

Change of crane loading or layout. As explained in the next chapter, cranes apply concentrated loads to the metal structure of the building. Cranes are usually supported from the pillars of the building, but monorails are usually suspended from the structure's girders. Any increase in monorail capacity or relocation of the crane runway will result in new stresses on the prefabricated structures that the structures are not designed to support. A new swinging door is required in the compartment with an existing cross brace. Can the bracing be moved to another bay? Next to "our" building, a new taller building or an extension will be erected. As a result, accumulated snow is likely to build up on the existing roof and possibly weigh it down. Can the roof of the building be reinforced? As a result of the mechanical system upgrade, a new high-capacity roof-mounted HVAC equipment is proposed. Is the prefabricated roof structure strong enough to support it? New state-of-the-art process equipment can only fit into the existing building if one of the beams of the main structure "loses" a few inches in depth. It is possible?

In all these cases an intelligent answer can only be given if the sizes of the metallic structures are known and the building can be analyzed well. The first step in this endeavor is to locate the original design drawings and calculations of the structure made by the fabricator. these could

RECONSTRUCTION AND RENEWAL OF METALLIC CONSTRUCTION SYSTEMS RECONSTRUCTION AND RENEWAL OF METALLIC SYSTEMS

407

FIGURE 14.10 Rusted wall fixtures and window frames bleed through paint.

in the client's file or the city's building authority or, if an architect is involved, in the architect's file. Unfortunately, if something is found, it probably consists only of the assembly plans without information on the component sizes. Remember that manufacturing drawings and calculations are not normally provided by the manufacturer unless required by contract. Nevertheless, plans can at least contain the name and order number of the building manufacturer. Armed with this information, the owner's engineers can proceed to step two – contacting the building's original manufacturer. If they're still in the market, manufacturers may have the coveted design files that include element sizes. In any case, they can help identify the design assumptions, steel grade, and similar information needed for the analysis. If the document search is unsuccessful, it is usually possible to measure the frame sizes for a new analysis on site. Such an analysis may seem daunting to some civil engineers unfamiliar with prefabricated building designs. In this case, the third step is to seek help from a fellow metallurgy fabricator who may be in a better position to reanalyze the building for the new loading conditions. Although rare, the need for such support is good reason to keep working.

RECONSTRUCTION AND RENOVATION OF METAL BUILDING SYSTEMS 408

CHAPTER FOURTEEN

Relationships with some manufacturers who could be contacted for help. Another option is to hire a specialist engineer who is well versed in the design and renovation of metal buildings. There are several techniques to strengthen primary and secondary framing. Perhaps the simplest method is to add additional primary frames in between the existing ones. In this way, the load on existing structures is reduced by half and the purlin span is halved, meaning purlin capacities are quadrupled. Of course, primary cadres will need their own foundations. To fit under the existing purlins, the new frames can easily be placed under the existing frames and the purlins secured to them with bearing clips. The clips not only compensate for minor differences in frame height, but also protect against weakening of the purlin network on the consoles. Another method of strengthening the frame is to add additional columns. This method is less desirable, not only because it introduces new obstacles into the building, but also because it requires careful consideration of the resulting stress redistribution. It is also possible to reinforce the primary and secondary frames by welding continuous plates or angle plates to their flanges. These and other reinforcement methods are discussed in detail in another book by the author, Structural Building Renovation: Design Methods, Details, and Examples.6 The book also includes a comprehensive case study of renovating an existing prefabricated building.

14.5 EXPANSION OF EXISTING METAL BUILDINGS One of the frequently cited advantages of metal building systems is their ease of expansion. It is true that it is relatively easy to extend a prefabricated building by adding several more matching framing panels to your expandable front wall and cutting a door into the wall. Complications begin when anything more than that is attempted. A case in point involves the removal of the old bulkhead structure to unify the new and existing spaces. As in chap. 3, this task is easy only if the building is intended for expansion and a moment-fixed frame is installed in each expandable side wall. Otherwise, removing an old bulkhead requires some technical gymnastics, such as B. temporarily bracing the purlins and struts supporting the end wall structure, removing the latter and erecting a new free field moment structure in its place - no easy task as illustrated in the case study in Sect. 14.6. The extension next to the existing building is even more treacherous. In northern regions, two gabled buildings sharing a common wall form a valley that is likely to be filled with accumulated snow (Fig. 14.11). The resulting snow load on the roof will likely exceed the design load for the original building. While the cultivation could certainly be designed for higher loads, the stock could be seriously endangered. In fact, the downtime losses due to this very condition run into the hundreds of millions of dollars.3 Unless an expensive structural upgrade is considered, it is best to avoid such an expansion of the building. A similar problem always arises when the extension has a higher roof height than the original building. Again, accumulated snow on the existing lower roof can lead to overstressing and failure. In some cases, the only option is to expand upwards. Second floor annexes are not very common in metal buildings, but they can be built in situations where site constraints leave no other option. Rather than supporting second story columns on top of existing ones, it is generally best to place new columns outside the building. The offset between new and existing supports is largely determined by the distance required to allow the new foundations to cross the existing ones (Fig. 14.12). A similar procedure can be followed if the existing operation has to be maintained during the construction of the new building. In fact, some metal building manufacturers make replacement systems designed to span the existing building (or multiple small buildings). An example is the one already mentioned in Chap. 1 mentioned Coronis Building Systems, Inc. 4, which sells its retroframes* for this purpose (Fig. 14.13). According to the manufacturer, retroframes have been used successfully in various applications. *Retroframe is a registered trademark of Coronis Building Systems, Inc.

RECONSTRUCTION AND RENEWAL OF METALLIC CONSTRUCTION SYSTEMS RECONSTRUCTION AND RENEWAL OF METALLIC SYSTEMS

409

FIGURE 14.11 Adding constructions adjacent to an existing structure can cause overloading.

14.6 CASE STUDY: ADDITION TO A PORCH BUILDING* 14.6.1 Project Background A public agency wanted to add an extension to their 10 year old early 1990's warehouse. The existing prefabricated building was approximately 100 feet tall, 60 feet wide and had an overhang of 20 feet . It was a simple gabled building with a longitudinal ridge and rigid single span structures spaced 25 feet apart at the center. Existing structural and architectural contract drawings were available. According to the drawing, a two-field horizontal roof panel made of steel rods, which reached from eaves to eaves, ensured the lateral stability of the building. Transverse bracing was provided on all four walls. The metal cover with vertical joints, with a pitch of 1:12, was held in continuous Z purlins. The roof was insulated with a fiberglass blanket laid over the purlins. The field isolated metal case was promoted by bypass design Z-tapes. The end walls were non-extendable, with end wall columns spaced 20 feet from the center. In the center of each end wall were 12 foot wide drop down doors. The building had a 6 inch thick panel on the level reinforced with welded wire mesh. The foundations consisted of 12 inch thick foundation walls on 20 inch wide wall foundations. The walls were partially exposed 3 to 4 feet and continued below ground level to the ice line for at least 3.5 feet. The drawings did not show vertical wall reinforcement or wall foundation reinforcement, but did show two #5 horizontal bars at the top and bottom of the walls. The walls were attached to the slab with #5 anchors 2 feet above the floor. The pillars of the structure were supported by integral columns and columnar foundations. A tie rod (1.5 inch diameter threaded rod) in a thick plate was provided on each pair of portal columns to resist their lateral reactions. The proposed extension measured 34 ⫻ 60 feet. It was to be fully integrated into the existing building, so one of the existing side walls had to be removed.

14.6.2 The design criteria The design criteria are listed below. Current Building Code: BOCA 1996 Roof Snow Load: 30 psf *This case study is based on a design by the Maguire Group, Inc. who provided the accompanying illustrations (Figs. 14.14-14.19).

RECONSTRUCTION AND RENOVATION OF METAL BUILDING SYSTEMS 410

CHAPTER FOURTEEN

FIGURE 14.12 Second floor extension over existing metal building.

Collateral Load: 5 psf Wind Load: Per BOCA 1996, Exposure C, Ground Wind Speed 85 mph Allowable Floor Bearing: 2 tons/ft2 Concrete Strength (new and existing): fc′ ⫽ 3000 psi Part of the extension was planned to be finished with drywall. The design intent was to separate the finished portion from the rest of the building by 2 inches so as not to penalize the entire building by limiting its deviation to H/500 (see discussion in Chapter 11). With a 20 foot eaves height, a 2 inch spacing would theoretically limit the allowable deviation to H/120. However, there was no guarantee that real-world design details would ensure full separation, so tighter lateral drift limitation was deemed necessary. A limit of H/200 was chosen as a good compromise, although a spacing of 2 inches was expected.

RECONSTRUCTION AND RENEWAL OF METALLIC CONSTRUCTION SYSTEMS RECONSTRUCTION AND RENEWAL OF METALLIC SYSTEMS

411

FIGURE 14.13 Proprietary structure surrounds an old building. (Coronis building systems.)

14.6.3 The Challenges of the Structure As the existing side walls were not extensible, one of them had to be removed and replaced with a rigid structure similar in span and roof pitch to the existing structures. The existing chords and purlins originally supported by the bulkhead had to be temporarily supported and eventually supported by the new frame. The new frame supports could not be placed on the existing foundation wall corners and still develop the required anchor bolt forces, since the drilled anchor bolts would have too small an edge distance. Instead, new piers and foundations had to be provided at the new positions of the frame columns. It was less clear how to encompass the 34-foot length of the annex. A review of Z-purlin load tables, similar to those in Appendix B, indicated that the proposed span and load were outside the economic range of cold-formed purlins. Therefore, two other framing system options were considered: 1. Two 17-foot purlin spans, which would require additional rigid framing, column footings, and ties to resist lateral column reactions. 2. A single 34-foot span framed with open mesh steel girders with custom seating details as discussed below. Both options were equally acceptable. The second system was chosen for this project. The roof plan for the combined building is in Fig. 14.14 and the foundation plan for the extension is in Fig. 14.15.

FIGURE 14.14 Roof plan for the case study.

RECONSTRUCTION AND RENOVATION OF METALLIC BUILDING SYSTEMS

412 Conveyed from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

RECONSTRUCTION AND RENEWAL OF METALLIC CONSTRUCTION SYSTEMS RECONSTRUCTION AND RENEWAL OF METALLIC SYSTEMS

413

FIGURE 14.15 Basic plan to supplement in the case study.

14.6.4 Roof Design Open web steel beams were spaced 5 ft o.c. - the typical third-octave spacing. The overall design load of the unit (nearby, safety and snow) was 30⫹ 5⫹ 5⫽ 40 psf, assuming a sustained load of 5 psf. The total load per linear foot of beam was 40 psf ⫻ 5 ft ⫽ 200 lb/ft. Load per linear foot of beam was 30 psf ⫻ 5 ft ⫽ 150 lb/ft

RECONSTRUCTION AND RENOVATION OF METAL BUILDING SYSTEMS 414

CHAPTER FOURTEEN

The most economical joist size selected from the load charts published by the Steel Joist Institute was 22K5. The secondary wall frame was limited to cold-formed Z-beams and a pair of intermediate wind columns were provided for beam support. These wind columns were placed on wall pillars and the columns were supported at the top by a special roof support. As in chap. 5, Open web sloped steel girders supporting vertical seam roofs often require spaced girder bracing and cross bridges. In addition, beam seats and brackets must be designed to withstand tipping caused by membrane forces. A corresponding note has been added to the general notes on adding prefabricated buildings (Fig. 14.16). The new intermediate structure was designed to support the existing Z purlins on one side and the new open web steel beams on the other. The standard seat depth for open web steel beams (2.5 inches) would not match the depth of existing purlins and in this case a custom beam seat was required (Fig. 14.17).

14.6.5 Foundation design The vertical loading of the new inner columns of the Portikus was calculated directly, as the product of the secondary area multiplied by the design load. The width of the tributary structure was 34 ⫹ 24.5 ᎏᎏ ⫽ 29.25 (ft) 2

HINTS ON ADDING PREDESIGNED BUILDINGS 1.

SEE GENERAL STRUCTURAL NOTES ON DRAWING S001 FOR CONSTRUCTION AND LOAD PATTERNS.

INTO THE PROJECT A MINIMUM SAFETY LOAD OF 5 P.S.F.

PROVIDE ADDITIONAL STRUCTURE FOR MECHANICAL AND OTHER EQUIPMENT WITH ROOF MOUNTING. PROVIDE ADDITIONAL SUPPORT (AND FRAMEWORK, IF REQUIRED) IN EXISTING SPA FOR SUCH FEE.

THE MAXIMUM ALLOWABLE VALUE OF LATERAL DRIFT UNDER ANY CONSTRUCTION LOAD COMBINATION SHALL NOT EXCEED H/200. SEPARATE DRYWALL PARTITIONS FROM BUILDING STRUCTURE BY AT LEAST 2 INCHES.

TYPE OF MAIN STRUCTURE: SINGLE-CLAMP FIXED CONSTRUCTION WITH FIXED COLUMN BASE (WITHOUT FLEXING MOMENTS AT BASE).

TYPE OF SECONDARY CONSTRUCTION: STEEL BENCHES OF THE SPECIFIED SIZE AND SPACING. PROVIDE BEAM STRENGTH WITH CONFINED SPACE AND TRANSMITTER TO COMPENSATE FOR BEAM SLOPE (APPROXIMATELY 1:12) AND LACK OF STRENGTH THROUGH THE ROOF. CONSTRUCTION OF BEAM SEATS AND ACCESSORIES TO ROLL OVER BY MEMBRANE FORCES.

COLUMN BASE PLATE SIZES MUST NOT EXCEED 12"x12". THE DISTANCE FROM THE EDGE TO THE ANCHOR SCREWS MUST BE A MINIMUM OF 7" FOR SIDE PILLARS, 6" FOR PULL-WALL PILLARS AND 4" FOR DOOR JAMBS.

FIGURE 14.16 Notes on adding a prefabricated building to the case study.

RECONSTRUCTION AND RENEWAL OF METALLIC CONSTRUCTION SYSTEMS RECONSTRUCTION AND RENEWAL OF METALLIC SYSTEMS

415

FIGURE 14.17 Section through a new external structure in the case study.

The inflow area for a new internal column was 29.25 ⫻ 30 ⫽ 877.5 ft2 The total vertical load on the column for a unit load of 40 psf was 877.5 ⫻ 40 ᎏᎏ ⫽ 35.1 kip 1000 The shoe size needed was (35, 1 kip) / 4 ksf ⫽ 8.8 ft2, so a 3 ft2 shoe can be used. The horizontal frame reactions were estimated from tables similar to those in Appendix D for a span of 60 feet, an eaves height of 20 feet, a tributary width of 29.25 feet and the design loads listed above. Table values for bays 25 feet wide have been increased by a factor of 29.25/25 ⫽ 1.17. The design value of the horizontal frame reactions was determined to be 17 kip. To withstand the reactions of the horizontal structure and to enclose the existing concrete walls to provide sufficient 'ballast' against wind uplift, the new piers were attached to the existing foundation walls (the old end walls) so that the existing walls acted as tie rods. The dowels were drilled into the average wall thickness of 12 inches. The design process for this task is described in a separate section below. The downward load on the two new center end wall posts was 0.04 kSF ⫻ 20 ft ⫻ 34 ft ᎏᎏᎏ ⫽ 13.6 kip 2 Likewise, the corner posts would be loaded at 13.6/2 ⫽ 6.8 kip. The shoes of the center and corner columns were nominal; Its size (3 ft ⫻ 3 ft) was determined simply by a 6-inch overhang around the pillars.

RECONSTRUCTION AND RENOVATION OF METAL BUILDING SYSTEMS 416

CHAPTER FOURTEEN

The pillar sizes (2 ft ⫻ 2 ft) were chosen to provide a generous tolerance for pillar sizes, ease in anchor bolt development, and provide 2 inch clearance for the plate seat on the pier. Despite the large size, a note was added to the drawings limiting the size of the column base to 12 in ⫻ 12 in, which is also well suited for the span and loading of the building. To maximize anchor bolt design options, the same note also required a minimum edge spacing of 7 inches for side jambs, 6 inches for end wall jambs, and 4 inches for door jambs (see Fig. 14.16). Details of the foundation piers in the corner and central columns are shown in Fig. 14.18. Once the outer foundation walls were partially exposed to match the existing ones, they acted as basement walls retaining the floor and had to be fixed to the slab at level with #4 dowels spaced 12 feet above the floor. Soil compaction near the walls with protruding pegs is very difficult and the pegs should be bent in the field (see Fig. 14.19). The final step in the foundation design involved a height check of the structure's new piers. The design lifting capacity was first estimated. As of BOCA 1996, the gross lifting capacity on the roof was determined by the following formula: P ⫽ Pv I [Kh Gh Cp ⫺ Kh (GCpi) ] where Pv ⫽ 18.5 psf for 85 mph wind I ⫽ 1.08 (importance factor) from Die corresponding tables in the code: kh ⫽ 0.87 gh ⫽ 1.29 gcpi ⫽ 0.25 cp ⫽ ⫺0.7 p ⫽ 18.5 ⫻ 1.08 [0.87 ⫻ 1.29 ⫻ (⫺0.7) ⫺ 0.87 ⫻ 0.25] ⫽ 1.29 ⫻ (⫺0.7) ⫺ 0.87 ⫻ 0.25) ] ⫽ 0 puplift is obtained by subtracting two thirds of the permanent load of the unit building: Pnet ⫽ 20.0 ⫺ 2Ⲑ3 ⫻ 5 ⫽ 16.67 (psf) The total lifting force at the foundation is then 16.67 60 ᎏᎏ ᎏᎏ 29.25 ⫽ 14.63 kip 1000 2 This lifting force was supported by two thirds of the weight of the foundation pillar, foundation, soil supported on the foundation and the new and existing foundation walls attached to the pier. The height of the walls was assumed to be 4.75 feet to the top of the base of the wall. As can easily be demonstrated, the available "ballast" was more than sufficient.

14.6.6 Development of Horizontal Frame Reactions The total available horizontal reinforcement in the existing walls consisted of four #5 bars with a total area of 1.23 in2. Assuming the bars were properly spliced and the steel's allowable tensile stress of 24 ksi was used, the bars could safely develop a tensile strength of 24 ⫻ 1.23 ⫽ 29.5 kip ⬎ 17 kip

(OK)

FIGURE 14.18 Details of the foundation piers in the new end wall in the case study: (a) corner; (b) middle.

RECONSTRUCTION AND RENOVATION OF METALLIC BUILDING SYSTEMS

417 Conveyed from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

RECONSTRUCTION AND RENOVATION OF METALLIC BUILDING SYSTEMS 418

CHAPTER FOURTEEN

FIGURE 14.19 Section through the new foundation in the case study.

The calculation of the size and number of the drilled dowels placed at the edge of the front wall was carried out using two different approaches, since there was no generally applicable method for determining the tensile strength of glued dowels at that time. In one approach, a reputable anchor manufacturer's catalog (Ref. 7) was consulted to determine the final steel and bond strength for rebar in concrete. This number was then divided by a safety factor of 4 and multiplied by the 6 inch margin adjustment factor. For No. 5 bars with a 10-in.

RECONSTRUCTION AND RENEWAL OF METALLIC CONSTRUCTION SYSTEMS RECONSTRUCTION AND RENEWAL OF METALLIC SYSTEMS

419

The final bond strength was reported as 33,820 lb, the edge spacing adjustment factor was 0.88, and the pin design capacity was calculated as 33,820 1 ᎏᎏ ᎏᎏ 0.88 ⫽ 7.44 (kip) 1000 4 . This approach would require at least three pins total. In the second approach, the maximum pull-out strength of a concrete pin was determined using the method of BOCA 1996 Sec. 1913.1.2.1 for the tensile strength of head anchors in concrete. The final concrete strength Pc for a 12 inch bar bed was found to be: ␾Pc ⫽ ␾␭ f c′ (2.8 As) where As ⫽ is the projected area of the concrete cone with a radius considered conservatively equal to the edge distance , or 6 in: As ⫽ ␲ 62 ⫽ 113 in2 ␭ ⫽ 1.0 for normal concrete ␾ ⫽ 0.65 (strength reduction factor) 2.8 ⫻ 113 ␾Pc ⫽ 0.65 ⫻ 1 .0 ⫽ 3000 ᎏᎏ ⫽ k ip0. per do0

The maximum strength of steel Ps was determined as follows: Ps ⫽ 0.9 Ab fs′ where Ab ⫽ bar area (0.31 in2 for #5) fs′ ⫽ yield strength of steel (60 ksi) Ps ⫽ 0.9 ⫻ 0 .31 ⫻ 60 ⫽ 16.74 kip Concrete control capacity. The factored design tensile strength was calculated by multiplying the working load of 17 kip by a combined load factor of 1.6 and an additional factor of 1.3 to account for a special check design: Tu ⫽ 17 ⫻ 1.6 ⫻ 1.3 ⫽ 35.4 kip The minimum required number of dowels was 35.4 ᎏᎏ ⫽ 3.14, say 4 11.26 An additional dowel was drilled into the shoe, for a total of five. A section through the new pier is shown in Fig. 14.20. Drilled studs alone were not sufficient to convert the existing wall into a tie beam: proper load transfer from the studs to the bars in the existing wall was required. Because of this, the existing transoms could not simply be cut where the wall was cut to make room for a new foundation pier. Instead, the existing bars were to be salvaged, cleaned and relocated to the new pier. Technically, even with these measurements, the existing wall could not be considered a true tie rod because overlapping splices in tie rods were not allowed by regulations. In this case, however, it was decided that retaining a 6-foot-deep wall with rods spliced and tied to the existing slab was preferable to removing the wall and replacing it with a 12-inch-thick slab, as was done on the original building. . The same dowel connection was made with the existing side wall foundation.

RECONSTRUCTION AND RENOVATION OF METAL BUILDING SYSTEMS 420

CHAPTER FOURTEEN

FIGURE 14.20 Section through new rigid frame foundation piers in a case study.

RECONSTRUCTION AND RENEWAL OF METALLIC CONSTRUCTION SYSTEMS RECONSTRUCTION AND RENEWAL OF METALLIC SYSTEMS

421

LITERATUR 1. James E. Grimes, „Metal Reroof Systems Deserve a Closer Look by Specificationrs“, Metal Architecture, November 1993. 2. Maureen Eaton, „Re-covering Roofs Requires Careful Evaluation“, Building Design and Construction, Dezember 1994 3 Wayne Tobiasson, „Some Thoughts on Snowloads“, Structure, Winter 1995. 4. NuRoof Design/Installation Information, MBCI Publication, 7. Oktober 1993. 5. Rene Dupuis, „How to Prepare Comprehensive Re-roofing Specifications“, 1989 Handbook of Commercial Roofing Systems, Edgell Communications, Inc., Cleveland, OH, 1989. 6. Alexander Newman, Structural Building Renovation: Methods, Details, and Design Examples, McGrawHill, New York, 2001. 7. Technical Product Guide (HVA Adhesive System), Hilti Corp., Tulsa, OK, 2000.

REVIEW QUESTIONS 1. Explain some pitfalls of attempting to add tiles to an existing prefabricated building. 2 A contractor proposes to erect a new sloping roof and place the supports of the structure on a regular 7 ⫻ 7 foot grid. The spacing of the existing secondary roof panels is unknown as they are covered by a sheet of drywall and the homeowner does not wish to remove it. What advice would you give to builders and owners? 3 What would you recommend to a homeowner looking to replace rusted metal siding with brick veneers? 4 Name at least two methods to increase the bending resistance of the primary structure. 5 Can rusty metal roofs be reliably coated? 6 What to do about rusty screws in a 30 year old metal roof? 7 It is proposed to renovate the existing metal roof with a new metal roof, fiberglass insulation and a good vapor retarder. The existing roof also has vapor retardant coated fiberglass insulation. Are there problems with this plan? 8 Homeowner asks for your "gut feeling" about the feasibility of adding a 20 ton HVAC unit on top of an existing metal building. What's your "gut feeling" in the absence of any design data?

RECONSTRUCTION AND RENOVATION OF METALLIC BUILDING SYSTEMS

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 15

CRANE BUILDING SPECIFICATION

15.1 INTRODUCTION Two out of five metal construction systems are built for manufacturing facilities that often require cranes for material handling. A construction crane is a complex construction system consisting of the actual crane with trolley and crane, crane rails with their attachments, crane runway girders, construction supports, stops and bumpers. A powered crane would also contain electrical and mechanical components that are not discussed here. Our discussion is still limited to hall construction cranes. The main focus of this chapter is on the correct integration of the crane and steel construction in a coordinated and networked system. Any attempt to add a heavy crane to the prefabricated building that has already been designed and built is likely to be fraught with frustration, high costs and inefficiencies. However, if the necessary planning is done in advance, a cost-effective solution is much more likely. We are also interested in a relationship between the three main parties with design responsibilities - architect-engineer, steel structure manufacturer and crane supplier. Disputes occasionally arise when contractual documents do not clearly delineate their respective roles in the project.

15.2 CONSTRUCTION CRANES: TYPES AND APPLICATION CLASSIFICATIONS Various types of cranes are suitable for industrial metal construction systems, the most common being overhead cranes (either overhead or overhead), monorail and jib cranes. Occasionally stacker cranes and gantry cranes may be required for special warehousing and manufacturing needs. Jib, monorail and bridge cranes are examined here in this order - in ascending order of the static requirements for a prefabricated structure. Space limitations prevent us from discussing gantry cranes and stacker cranes, as well as conveyors and similar material handling systems. Within each type, cranes are classified according to frequency and severity of use. Each crane must meet one of six service ratings established by the Crane Manufacturers Association of America (CMAA). The six classes are: A (standby or irregular duty), B (light), C (moderate), D (heavy), E (heavy) and F (continuous heavy). Guidance for assigning a service rating is provided in CMAA standards 701 and 742 and the MBMA manual.3 The Manual Design Practices apply only to cranes with service ratings A through D. Information for cranes with service ratings E or F, including design loads and influencing factors , is included in the AISE 13.4 Technical Report. Another way to classify cranes is by type of movement - manual or electric. Push trolley cranes are physically pulled along the rail by the operator and are cheaper but slower.

423 Conveyed from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

SPECIFICATION OF CRANE BUILDINGS 424

CHAPTER FIFTEEN

as electric cranes. Manual transmission cranes operate with less structural impact than their faster-working electric cousins. The operator controlling an electrically powered crane can either stand on the ground using an overhead drop button station or sit in a cab located on the movable bridge.

15.3 JIB CRANES Jib cranes require relatively little planning effort on the part of the prefabricated house designer. Floor-mounted slewing jib cranes (also known as pillar cranes) are not dependent on the superstructure of the building for support and stand on their own foundations (Fig. 15.1). Pillar slewing cranes, on the other hand, are supported or supported by the building's metal columns and therefore impose certain strength and rigidity requirements on the structure. For example Ref. 5 recommends that structural columns supporting jib cranes be rigid enough that the relative vertical deflection at the end of the jib is limited to the jib length divided by 225. Floor mounted jib cranes can rotate a full 360°, while column cranes are generally limited to 200° jib rotation. A crane takes the load onto a trolley, which runs on the bottom flange of the jib and loads a chain hoist. The hoist can be electric or manual. As the load is lifted, the boom pivots around the crane's stationary column and lowers the object to the desired location. These two operations - trolley travel and boom rotation - are often performed manually. The crane boom length ranges from 8 to 20 feet. Lifting capacities range from 1Ⲑ4 to 5 tons,3 with 1Ⲑ2 and 1 ton cranes being the most popular.

FIGURE 15.1 Floor mounted jib crane. (American Crane and Equipment Corp.)

CRANE BUILDING SPECIFICATION CRANE BUILDING SPECIFICATION

425

Common jib crane applications include machine maintenance, assembly lines, steam hammers and loading docks. Sometimes a pair of cranes and a monorail combined with forklifts are sufficient to move cargo from a loading dock to the area served by overhead cranes or stacker cranes. Inexpensive jib cranes can relieve main cranes of many smaller jobs that would tie them up for a long time. Floor mounted jib crane manufacturers normally provide recommended foundation sizes for their equipment, but foundation design is still the responsibility of the designer. Whenever ground mounted crane foundations are added to an existing metal building, care must be taken not to damage any tethers or hairpins that may be in the slab. Otherwise, the building's lateral load-carrying system may be damaged. Obviously, an additional crane on pillars will need to be approached with even more caution, as the building's existing pillars will likely need to be reinforced to withstand the new loads imposed. Basic design concepts for jib cranes are discussed in Ref. 6.

15.4 MONWAYS 15.4.1 The Monorail System The monorail crane is a familiar sight in many industrial plants, workshops and storage facilities. Monorails are cost effective for applications that require material transfer along specified routes without lateral deviations; its traversing range can be extended with the help of switches and turntables. The monorail crane is essentially a crane supported by a trolley whose wheels move on the bottom chord of a single-track girder. Monorails can be used to move loads from 1 to 10 tons and can be manual or electric. Traditionally, monorail beams are made from standard wide flange sections that accept straight profile wheels, or from I-beams that support tapered profile wheels. (The straight wheel is essentially a short cylinder; the tapered wheel is a short truncated cone.) Today, these standard rail supports are increasingly being replaced by proprietary rail support products with non-uniform flange configurations. Figure 15.2 illustrates such a hard alloy steel inverted tee offered by crane manufacturers; also shows the loads applied to the track by the hoist. Some advantages of proprietary belts over laminated backings include better wear resistance, easier rolling, longer life, and weight savings. Chains have been specifically designed to overcome problems common to standard shapes such as: B. Excessive local flange deflection due to wheel load. The advantages of proprietary products make their use worthwhile for most monorail applications.

15.4.2 Loads acting on beams of a monorail

FIGURE 15.2 Loads acting on the monorail beam (proprietary rail shown).

The weight shown in Fig. 15.2 includes the weight lifted and the weight of the hoist and trolley. It also includes the effects that the AISC7 specification prescribes with 10% of the maximum wheel load for cranes with suspension and 25% of the wheel load for cab-operated cranes. Transverse thrust S is given as 20% of the load lifted, including the crane trolley; this lateral force is evenly distributed to all crane wheels. According to AISC, the longitudinal force L caused by the trolley deceleration is to be taken into account with 10% of the maximum wheel loads. Some building codes contain design codes other than these, and the AISC specification's load percentages apply only where otherwise specified in other referenced standards. In addition, some codes contain more detailed provisions for monorail cranes. The International Building Code8

SPECIFICATION OF CRANE STRUCTURES 426

CHAPTER FIFTEEN

does not require a surge load increase for overhead traveling cranes and monorail cranes with overhead traveling cranes, trolley cranes and hand gear cranes, but for motor overhead traveling cranes the IBC specifies a surge load increase of 25 percent. In the case of bridge cranes, various building authorities place higher minimum requirements on impacts and lateral loads on the track supports. For example, ANSI MH 27.19 specifies shock tolerance for electric hoists as 1Ⲑ2% of rated load for each foot per minute of lifting speed, with a minimum tolerance of 15% and a maximum tolerance of 50%. For bucket and magnet applications, the shock tolerance must be 50% of the rated load. Struts and side braces have to withstand each of these forces.

15.4.3 Suspension and Bracing Systems Monorails running perpendicular to the main portal are normally designed to run between the portals without intermediate supports. When monorails run parallel to structures, additional support beams are required. There are two basic suspension systems for attaching monorail beams to a structure or beam: rigid and flexible. In a rigid or fixed system, the beam is connected to the structure by relatively rigid steel support members. Depending on the available vertical clearance of the structure, the support member can consist of a simple strut welded to the underside of the frame beam or support beam (Fig. 15.3a) or a longer steel section (Fig. 15.3b). A short support can usually withstand vertical and lateral shear reactions, but longer sections may need to be supplemented with diagonal angles. The suspended load exerts a concentrated force on the supporting structure, and additional reinforcement and welding is usually required to reinforce the beams at these locations. Some manufacturers supply a single reinforcement (Fig. 15.3), others two double reinforcements (Fig. 15.4). Irrespective of the specific details, it must be clear from the contract documents who is responsible for the various components of the suspension system. Figure 15.4 illustrates a manufacturer's approach, where the frame reinforcements and a shop-welded short strut are supplied by the manufacturer, while the other components and reinforcements are supplied by the crane supplier. A flexible suspension uses suspended rods instead of supports (Fig. 15.5). The bars are usually attached to hooks placed over the top flange of the frame and the length of the bar can be easily adjusted. According to published data3,10 flexible suspension tends to result in lower crane loads

FIGURE 15.3 Rigid monorail mounts: (a) minimum length clamp; (b) long support. (Metal construction systems.)

CRANE BUILDING SPECIFICATION CRANE BUILDING SPECIFICATION

427

FIGURE 15.4 Details of fixed suspension and reinforcement welding. The manufacturer usually excludes dashed points from its scope of services. (Nucor Building Systems.)

FIGURE 15.5 Weld detail of mesh reinforcement for flexible suspension. The manufacturer usually excludes dashed points from its scope of services. (Nucor Building Systems.)

SPECIFICATION OF CRANE STRUCTURES 428

CHAPTER FIFTEEN

and reduced wear. Monorail supports supported by suspended bars always require lateral anti-pivoting bracing for stability. Any suspension system must be vertically adjustable to bring the monorail supports to a truly horizontal position before the side braces are installed. Of course, the flexible suspension makes adjustments easy. Lateral thrust S was traditionally counteracted by a shop-welded channel laid flat on an I-beam (Fig. 15.3). The proprietary rail with its wide top flange eliminates the need for the channel. Lateral shear can also be resisted by intermittent lateral reinforcement of the upper beam flange; these braces shall be designed not to affect the vertical deflection of the monorail beam. In the case of prefabricated houses with cold-formed purlins, such a transverse bracing can be impractical: purlins have little lateral stability of their own and cannot absorb any bracing loads. Some additional structural members must then be introduced to withstand the lateral bracing forces, or appropriate lateral bracing between the purlins must be provided to distribute these forces in the roof membrane. The longitudinal L force of the track can be counteracted by diagonal angle braces placed at approximately 100 foot intervals and at all curves of the track in the plane of the monorail girder (headers placed between purlins) are necessary to withstand the bracing forces. The locations and conceptual details of all vertical and lateral supports are to be specified in the contract documents. Vertical supports should ideally be on each main frame or at 20 to 25 foot intervals if the monorail runs along the frame. A special case occurs when two parallel monorail beams must merge into a single vertical beam. This problem can be solved with monorail switches.

15.4.4 Rail Rail Design Considerations ANSI MH 27.1 requires that the allowable stress in the lower flange (tension) of monorail rail rails be limited to 20% of the maximum steel strength. It also specifies the deflection criterion for monorail beams with a span of 46 feet or less as the length between vertical supports divided by 450. The L/450 deflection criterion is also found in other sources, including the MBMA Handbook. The monorail girders must be carefully spliced to allow smooth movement of the wheels between each girder. The best seam detail is the full penetration welding of the lower flange in combination with thrust plates bolted to the web. Some fabricators of prefabs prefer to use anchor plates welded in place and to locate the seams under each column (Fig. 15.6). To ensure that the splice will not interfere with the trolley run, it is advisable to request a test run of the trolley along the entire length of the monorail before accepting an order. Who supplies the roadway beam and its supports? The metal construction manufacturer already supplies the suspensions and can also supply the track girder if expressly requested by the contract (crane work is usually excluded from the manufacturer's scope of services). The track support supplied by a building fabricator will likely be of standard structural form. Wherever proprietary track supports or switches are specified, they must be supplied and installed by the monorail supplier.

15.4.5 Special requirements for support frame beams

FIGURE 15.6 Splicing of deck beams at suspension points. (Metal construction systems.)

Whatever suspension system is chosen, the weight of the loaded monorail must be transferred to the beam of the supporting structure. A hanging load attempting to pull the bottom beam chord from its web must be resisted by welds between the chord and the web; Web tearing is counteracted by reinforcements (Figs. 15.3, 15.4 and 15.5).

CRANE BUILDING SPECIFICATION CRANE BUILDING SPECIFICATION

429

Fabricators of prefabricated buildings generally provide single-sided web-to-flange welds in their primary structures. Such welds may be unsuitable to withstand high localized levitation loads. Single-side welded sleepers subjected to cyclic levitation loads can experience fatigue problems caused by a notch sometimes created by single-side welding.11 For this reason, the areas around the sleepers may need to be reinforced with double welds on both sides, to work in combination with fabric reinforcements to withstand suspended loads. Occasionally, a large part of the structure needs to be reinforced with double welds. This nuance is just one more reason to coordinate the design of steel construction systems and cranes. Loads acting on the metallic structural system of monorails or other cranes supported by the building structure should be included in the values given in Chap. 3.

15.5 LOW OVERHEAD CRANES 15.5.1 System description An overhead crane, as the name suggests, has a lifting carriage that moves along the crane and “hangs” on the track girders. The crane usually consists of a single and occasionally a double girder supported by two wheeled end girders which run on the bottom chords of the track girders (Fig. 15.7). The runway supports, in turn, are suspended from the girders or girders of the building structure. (In the latter case, the supports are located at the truss plate points.) ANSI MH 27.1 requires a minimum clearance of 2 inches between the overhead traveling crane and any obstructions on the side or overhead. Overhead cranes have relatively modest lifting capacities - from 1 to 10 tons - and are generally limited to spans of 20 to 50 feet. Barrel cranes should generally be of the overhead type. Manual and electric transmission overhead cranes are available. Electric cranes are usually controlled from a pendant station, although cab-operated and automatically controlled bridge cranes also exist. The main advantage of the suspended construction is that the crane span does not have to extend completely between the columns of the building. Overhead cranes are therefore particularly suitable when only part of the building aisle needs maintenance and the building has a large free span. The overhead design allows the trolley to move beyond the centerlines of the track girders and allows load transfer between adjacent crane aisles or between multiple parallel overhead traveling cranes in one aisle. 15.5.2 Track Girders The design and construction of track girders for overhead cranes is similar to that of monorails. Traditionally, both types of cranes were based on rails made of I-beams (today called S-profiles).

FIGURE 15.7 Hanging single girder crane. (FKI Industries, Inc.)

INDICATION OF CRANE BUILDINGS 430

CHAPTER FIFTEEN

thick cores for good resistance to lateral loads and tapered flanges useful for self-aligning wagon wheels. Recently, the proprietary tracks described above are gaining more and more popularity. The girder suspension and shoring details for overhead cranes are the same as for monorails, except that only one of the two runways needs to be braced laterally to allow for deviations in alignment and crane deflection.3 While monorails and overhead cranes can patent the same rail construction, they can be controlled by different locking devices and switches are connected to each other. The resulting combined crane cover can be individualized for specific processes and is still very cost-effective.

15.5.3 Design data An example of required minimum clearances for cranes suspended from a girder is given in Figure 15.8, reproduced with permission from Ref. 12. Figure 15.8 also includes maximum crane wheel loads, crane weights and lifting data. ANSI MH 27.1 provides additional information on some design issues related to overhead cranes. As in Sect. 15.4.5, the girders of the main frames supporting overhead cranes shall have two-sided welds connecting the flanges to the web. Wherever proprietary track sections, stops or switches are specified, it is best to have these all provided by the crane supplier for product compatibility reasons. Standard section rail supports can be supplied by the steel fabricator if required by contract.

15.6 OPERATION OF GOVERNING CRANES 15.6.1 System Description A typical crane is supported on structural columns and covers most of the aisle – the space between the columns. Crane movement is performed by two trucks mounted on rails supported by track girders. Travel speed is generally higher than that of bridge cranes. The crane can consist of a single girder supporting a trolley crane traveling on its lower chord in monorail construction (Fig. 15.9a) or a double girder supporting an upper trolley (Fig. 15.9b). Double girder cranes can lift heavier loads and accommodate greater lifting heights than single girder cranes; Even larger loads can be lifted with box girder cranes. Overhead cranes are primarily electrically operated, with the exception of some single-girder models, and are usually controlled by hanging knobs. Some heavy box girder cranes have operator controlled cabs. Top jib cranes are limited to lifting capacities from 1 to 10 tons and spans from 20 to 60 feet. They are a good choice for budget conscious homeowners whose modest material handling needs can be met with the low duty crane. An example of dimensional and load data for single-girder cranes is shown in Fig. 15.10. Cranes with double girder suspension can lift 5 to 25 tons, can be scaled for higher operating capacities and have spans from 30 to 100 feet. Lifting requirements in excess of 25 tons typically require a box girder bridge. Some cab powered box girder models can lift 250 tons and bend up to 100 feet. Double-girder overhead traveling cranes with trolleys offer the longest hoisting distances of any crane type reviewed here, and low-profile models are available for cases where a few inches more headroom is worth the inevitable sacrifice in wasted lifting capacity and capacity. An example of dimensional and load requirements for double-girder cranes is shown in Fig. 15.11. (As with all tables reproduced in this chapter, other manufacturers' cranes will not necessarily agree with this data.) For all top operating cranes, a minimum of 3 in. pol. from CMAA 70 and CMAA 74.

CRANE BUILDING SPECIFICATION CRANE BUILDING SPECIFICATION

FIGURE 15.8 Dimensional and load data for single-girder cranes. (FKI Industries, Inc.)

431

SPECIFICATION OF CRANE STRUCTURES 432

CHAPTER FIFTEEN

FIGURE 15.9 Mole taxiway bridges: (a) single girder; (b) Bars. (FKI Industries, Inc.)

15.6.2 Forces acting on overhead crane runways Runway girders supporting overhead cranes are generally made of standard structural sections; A combination of wide flange beam and gutter is typical. Because the wheels on these cranes ride over the rails and not the actual girders, hardened proprietary rails are not required. The crane wheels exert the same three types of reactions on the deck girder as described in Section 15.4.2, but applied to the top chord (Fig. 15.12). The AISC specification and IBC assign equal factors for vertical, lateral, and longitudinal impact forces, regardless of capacity. Other sources recommend a sliding approach, designing heavier cranes for proportionately larger loads. For example, Weaver13 suggests adjusting the load values according to the crane's CMAA service rating. CMAA Class A cranes would be rated for 10 percent vertical impact, 10 percent side load and 5 percent longitudinal load of the wheel loads, while F Class cranes would be rated for 25 to 50 percent, 15 to 20 percent and 20 to 30 percent, For most CMAA service classifications, this approach results in a higher design burden than the AISC requires. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

CRANE BUILDING SPECIFICATION CRANE BUILDING SPECIFICATION

FIGURE 15.10 Dimensional and load data for overhead working single girder cranes. (FKI Industries, Inc.)

433

SPECIFICATION OF CRANE STRUCTURES 434

CHAPTER FIFTEEN

FIGURE 15.11 Dimensional and loading data for top-operated double-girder cranes. (FKI Industries, Inc.)

CRANE BUILDING SPECIFICATION CRANE BUILDING SPECIFICATION

435

CMAA 70 and CMAA 74 require the shock factor (crane load factor) to be accounted for at 0.5 percent of the hoisting speed in feet per minute, but no less than 15 percent and no more than 50 percent. For bucket and magnet cranes, the shock factor should be assumed to be 50% of the lifting capacity. In addition to this raised load impact factor, CMAA 70 and 74 also require that the impact factor be attributed to the standing load of the crane, trolley and associated equipment. The continuous load factor is specified as 1.1 for cranes with travel speeds of up to 200 ft/min and 1.2 for faster cranes. CMAA 70 and 74 also include other crane loads and load combinations for which crane supports must be designed. Alternatively, some engineers assume a shock factor of 0.25 for the preliminary design of most cranes, as in Ref 14 for example. FIGURE 15.12 Forces acting on the rail The support and bracing systems are able to withstand the large loads applied to the upper jib. Overhead cranes are much more complex than those used for monorail support or overhead cranes. These support systems are discussed in separate sections below. For the sake of simplicity, our discussion is limited to buildings housing a single crane. Additional considerations for buildings with multiple cranes are covered in the MBMA manual.3

15.6.3 Craneway Girder Structural Design Structural design of track girders for combined loads is well covered in many design manuals as well as in the AISC Design Guide 7.5, so only a general procedure is described here. The first design step is to determine if fatigue drives the design. Fatigue cracking is responsible for perhaps nine out of ten crane girder problems.15 Given the expected number of cycles by the owner and a building lifespan - 50 years can be taken as a standard value - it follows the procedure in the AISC specification Annex K to determine the allowable stress range. CMAA Class D, E and F crane girders are often fatigue controlled, meaning the allowable bending stress in these components is reduced from 24 to perhaps 16 kip/in2. 16. When fatigue is not critical, the combined allowable stress of the beam can be determined in a conventional manner, given the properties of the beam and its free length. Without additional lateral bracing, the unbraced length of a simply supported deck girder corresponds to the support spacing (span). For shaft sizes found in most prefabricated buildings - 20 to 30 feet - the allowable combined stress is often 0.6 Fy. The second step involves calculating the required stiffness (moment of inertia) of the track beam based on the allowable deflection criteria. Readers who read our discussion in Chap. 11 should keep in mind that there is no consensus among engineers on deflection criteria in general; Crane runway deflections are no exception. One source of information is the AISC Design Guide 3.17, which recommends the following design criteria: ●

● ●

For CMAA classes A to C, the vertical deflection of the track beam under wheel load is limited to L/600 (for class D, L/800). For CMAA Classes E and F, the maximum vertical deflection is limited to L/1000. The maximum lateral deflection of the track girders is limited to L/400 for all crane classes.

Other sources suggest slightly more restrictive criteria for vertical deflections, such as B. L/1000 for CMAA classes A, B and C and L/1200 for CMAA classes D, E and F. The lateral deflection criterion of L/400 seems to be generally accepted. Impact factors used for stress analysis do not need to be included in deflection calculations. The third step is to determine the maximum bending moments of horizontal and vertical movement loads. Horizontal loads can be assumed to be carried only by the runway girder top chord and gutter, if any.

SPECIFICATION OF CRANE STRUCTURES 436

CHAPTER FIFTEEN

Next, a test section is selected that keeps the deflection criteria and combined stresses within the previously established limits. Beam Tables from AISC Handbook, 18 AISC Design Guide 7 or Ref. 19 may be helpful for this task. The last step is usually to check the selected section for lateral web dents per section 1.K1.5 of the AISC specification. The AISC Design Guide 7 points out another step that is often overlooked: a calculation of the local longitudinal bending stress in the top chord of the track girder caused by the moving crane wheels. This additional contribution to the total bending stress in the top flange can add 1 to 4 kips/in2. In most cases, the design process outlined above will result in a selection of the combination of Wide Flange Beam Sections. For the heaviest cranes or for long track spans, steel girders with built-in cover troughs or even top-flanged horizontal trusses may be required. For light cranes and relatively small enclosures, it may be possible to select a single heavy wide-flanged girder with no top channel. The increased beam weight can be more than offset by the labor savings required to weld the duct. As a rule of thumb, a wide flange rail combination must be at least 20 lb/ft lighter than a single wide flange beam to be economical. If a single beam is used, its flange must be wide enough to accommodate rail attachment. One problem with gutters is tolerances: since neither the gutter nor the wide flange beam are perfectly straight, there are likely to be small gaps between the two. As the crane's wheels pass over the gaps, some wear may occur on the connecting welds or on the duct itself. For this reason, end rails or end plates should be avoided on girders used for crane classes E and F.15 Crane runway girders can be single-span or continuous. Continuous beams flex less under load and require lighter and therefore less expensive profiles. However, continuous elements are susceptible to damage due to uneven seating of supports and thermal stress build-up. Not only are single span tracks virtually unaffected by such problems, they are also easier to design, construct, and replace if necessary. We recommend designing all track beams as single-span elements. 15.6.4

Rail Girder Supports The simplest, and perhaps most common, method of supporting rail girders for overhead cranes is to use supports that are shop-welded to rigid frame columns (Fig. 15.13). The brackets of the supports are

FIGURE 15.13 Track girder supported by crane support with upper bearing. (Metal construction systems.)

CRANE BUILDING SPECIFICATION CRANE BUILDING SPECIFICATION

437

more suitable for relatively light cranes with a capacity of up to 20 tons; with heavier cranes, eccentrically loaded columns become uneconomical. In addition, any slight impact on the street will be transmitted to the metal structure of the building, possibly causing vibration and disturbing the occupants. In this system, building columns are typically reinforced with fabric reinforcements at the column attachment points. In addition, to improve the fatigue behavior of the frame, continuous welding of the frame web to the flanges on both sides is recommended - or by means of milling. The bottom flange of the deck beam is attached to the bracket with heavy duty bolts. The upper chord of the runway girder is braced laterally in the structure. As discussed further in the following section, this fixed connection must allow in-plane movement of the track beam ends in two directions - horizontal and vertical - to ensure load transfer perpendicular to the plane of the beam. Figure 15.14 shows a schematic of how the beam ends move and bend under load. A second track support method uses staggered pillars (Fig. 15.15), a solution common in older factories. Staggered supports are suitable for heavy cranes and for buildings with large eaves, which can benefit from the significant rigidity of such supports. With this construction, as with the previous one, the vibrations from the crane are likely to be transmitted to the rest of the building and felt by the occupants. Crane tracks with a capacity of more than 20 tons can be economically supported by a third method - separate crane columns. The separate supports are positioned directly under the runway girders and receive only vertical loads, while the building structure only withstands lateral loads from the crane. A separate set of small struts can be economical even for crane capacities less than 20 tons but spans in excess of 50 feet.3 Lateral reactions are transmitted to the building structure by bracing between the two sets of struts, which also serve as lateral support for the top chord of the runway beam (Fig. 15.16). Runway support columns are normally oriented with their webs perpendicular to those of the main frames. Some engineers prefer to construct these columns with a fixed base to minimize column displacement, although, as discussed in Chap. 12, the column fixation in real construction may be difficult to achieve.

15.6.5 Bracing against lateral and longitudinal forces of the roadway Of the three forces acting on a roadway beam according to Fig. 15.12, the vertical reaction V is taken up by the column or support column. The lateral impulse S is resisted by a cover channel or the top of the beam

FIGURE 15.14 Jib movement and deflection under load.

SPECIFICATION OF CRANE STRUCTURES 438

CHAPTER FIFTEEN

FIGURE 15.15 Detail of the stepped structural column supporting the crane. (Nucor Building Systems.)

Flange and transferred to the building column via the top connection described above. For heavy cranes, a horizontal truss or structural member may be required instead of the cover channel. In order to design a bracing system to withstand lateral shear forces, an assumption must be made as to whether or not these forces will be shared between the lanes on opposite sides. The answer depends on the shape of the crane wheels. Most medium-duty cranes have double rim wheels that "grip" well on the rail and ensure good lateral support. When these wheels are used, the lateral thrust can be shared equally between the two track beams. Some large cranes have double rim wheels on one side only and rimless rollers on the other side. This design is to prevent the wheels from binding due to changes in bridge length due to temperature fluctuations. If such wheels are used, only one track girder has to withstand the entire lateral thrust.14 The longitudinal force L poses some difficulties. Longitudinal forces act on the building, mainly caused by the braking or acceleration of the crane and by its impact with stops on the runway

CRANE BUILDING SPECIFICATION CRANE BUILDING SPECIFICATION

439

FIGURE 15.16 Crane runway girder supported by separate columns. (Constructive appendix systems.)

Columns for a racking action. Unfortunately, most of the crane buildings are poorly reinforced towards the runway. In fact, as Mueller20 notes, "the number one complaint of steel mill buildings in the past has been too lateral, too much vibration, too much movement." This opinion is shared by many crane manufacturers supplying their products for metal construction systems. Longitudinal movement of the structure is best minimized by providing cross braces of structural steel sections under the deck girders to transfer longitudinal forces from the deck girders to the foundations. For multi-bay rails, Mueller (as well as the MBMA manual) recommends that support curves be in the middle of the bay and not at the ends of the rails. He strongly advises against using knee pads to stiffen track girders. Proper splicing of track supports is extremely important. As Mueller demonstrates, a simple butt joint with web plates can cause the beam web to fail because the web plate restricts rotation of the simply supported beam ends. Instead, he recommends bolting the bottom flange of the girder to a crane column deck plate with bolts designed to transfer longitudinal forces from the runway to the cross brace below. As already mentioned in the previous section, the lateral connection of the drawstring is quite difficult to manufacture. This connection must accommodate in-plane movement of the track girder ends in horizontal and vertical directions (Fig. 15.14) but allow load transfer perpendicular to the plane of the girder. None of the many previously popular details completely solve the problem. The rigid vertical membrane, the horizontally curved plate, the oblong holes in Fig. 15.13, even the thin rods of Fig. 15.16 - none of these connections allow unrestricted crimping of the ends. The design solution using a rigid vertical plate diaphragm may be the least effective at providing the desired movement and has resulted in many failures that lead to diaphragm rupture

SPECIFICATION OF CRANE BUILDINGS 440

CHAPTER FIFTEEN

Membrane itself or the web of the ray (Ref. 20–23). Whichever fitting is used, it should preferably be connected directly to the top flange or cover rail of the crane girder rather than to its web. A prime method of laterally bracing track supports involves a proprietary back-tie, such as that shown in Fig. 15.17. The hinge allows rotation and movement of the beam ends while providing the desired resistance to lateral thrust. The details described above apply to light and medium capacity cranes commonly found in prefabricated buildings. Much more extensive detail is required for heavy cranes, some of which are presented in AISC Design Guide 7. Heavy cranes may also require additional reinforcement, which is not discussed here. For example, AISE 13 requires that bottom chords on crane girders longer than 36 feet be laterally braced.

15.6.6 Specification of Crane Rails and Rail Accessories Crane rails are normally selected and supplied by the crane supplier, but methods of attachment, connection and arrangement may be at the designer's discretion. Crane rails, if installed incorrectly, can cause a number of crane performance problems such as: B. premature wear of various crane components, high noise and vibration levels and even failure of the rail support. According to AISE 13.4, the rails must be centered on the beam; the maximum allowable eccentricity shall not exceed three quarters of the thickness of the web of the beam. The rails must be arranged so that the joints on both track supports and the joints in the supports and roof ducts are offset from each other by at least 1 foot. The crane supplier should ensure that the wheelbase of the crane does not match the stagger quantity. All crane rails must be marked "for crane operation". Rail joints can be made by bolting or welding. Successful welds require special final preparation and very strict control of the operation; both can be difficult to achieve in the workplace. Screw splices offer a more practical alternative. The best splice detail is a tight connection between rail sections and not the commonly used connection with a small gap (1Ⲑ16 to 1Ⲑ8 inches).18 This detail

FIGURE 15.17 Proprietary rail clips and tie rod linkage. (Gantrex Corporation.)

CRANE BUILDING SPECIFICATION CRANE BUILDING SPECIFICATION

441

requires milling of the rail ends and a specific rail drilling pattern. Adjacent rail sections are secured with heavy-duty bolts through the rails and two connecting rods that match the rail profile. Perhaps the biggest challenge for designers of crane rails is attaching the rails to the girders. The ideal joint should be able to withstand lateral thrust from the crane wheels without restricting the rail's ability to expand and contract with temperature changes independently of the support. (Hence, rails should never be welded to the track.) Rails can be attached to crane girders in three different ways. The first is by hook bolts, which were once popular but are now not recommended except for the lightest cranes. Hook bolts are not very good at resisting lateral shear forces because the bolts tend to stretch and the nuts tend to loosen under load. For this reason, hook bolts must be checked and adjusted frequently to maintain track alignment. 24 hook bolts are dimensionally suitable for capless crane girders with narrow flanges that do not leave room for other methods of attachment. Hook Bolts are supplied in pairs and are typically spaced 2 feet apart (Fig. 15.18). Rail clips work slightly better than hook bolts and can be used on almost any crane. The clamp assembly consists of a clamp plate connected to the carrier by two heavy duty bolts through a filler rod. Rail clips are spaced no more than 3 feet apart (Fig. 15.19). Clamps can be fixed or floating. Floating clamps are said to allow for thermal expansion of the rail, but unfortunately they can also allow for excessive transverse rail movement, making it difficult to maintain rail alignment. Tight clamps, on the other hand, tend to impede any rail movement and can lead to thermal stress build-up in the rail and beam. The third rail attachment method involves proprietary adjustable rail clamps available from various crane manufacturers. The patented clips are designed to allow the rails to expand longitudinally and limit their lateral movement. Some crane manufacturers also advocate the use of synthetic rubber pads under the rail at the attachment points, which provide smoother rail roll and less shock, noise and girder wear. An example of proprietary clips and pads is shown in Fig. 15.17. Currently, adjustable clips represent the best rail attachment method available. Regardless of the rail attachment method, regular inspection and maintenance is required to tighten the bolts and keep the rail aligned.

FIGURE 15.18 Fastening the crane rail to the track girder with hook bolts. (Nucor Building Systems.)

SPECIFICATION OF CRANE STRUCTURES 442

CHAPTER FIFTEEN

FIGURE 15.19 Attachment of crane rail to track girder by floating clamps. The manufacturer typically uses 1 inch ASTM A325 bolts with spring washers. (Nucor Building Systems.)

15.6.7 Rail Stops and Buffers Rail stops and buffers are often the last elements considered when designing a crane system. They shouldn't be any less important though, as a poorly designed crane stop can ruin both the crane and the building. Chain buffers work in a similar way to cars, absorbing the kinetic energy of the crane impact. Old blocks of wood and rubber are modern hydraulic and spring dampers like the one in Fig. 15.20. ANSI B30 and CMAA 70 standards require bridge fenders to be designed to withstand the force resulting from a crane impact at 40% of rated speed. If the building is designed in accordance with the AISE 13 regulations, it is assumed that the crane will be protected by bumpers up to the maximum travel speed.25 Track stops, as the name suggests, are intended to stop a moving crane. While proprietary bumpers are typically selected by their suppliers, crane bumper design belongs to the design professional. For monorails and bridge cranes, a small elbow attached to the web of the track beam may be appropriate. However, overhead cranes require a heavy prop bolted to the top of the rail girder. A bumper is attached to the bracket or intended to contact a bumper installed on the crane truck. Ultimately, of course, the force on the abutment must be absorbed by the building structure and its bracing. In the absence of crane and fender specific data, this force can be estimated as the greater of 10 percent of the crane's unloaded weight or twice the design pull force as suggested in AISC's internal Design Guide 7 for cranes.

15.7 HOW TO CHOOSE AND SPECIFY A CONSTRUCTION CRANE The layout of new industrial buildings, especially those with cranes, is often determined by the equipment inside. It therefore makes sense to first select a suitable crane cover for the process equipment and then determine the dimensions of the metal construction system and not the other.

CRANE BUILDING SPECIFICATION CRANE BUILDING SPECIFICATION

443

FIGURE 15.20 Hydraulic damper. (Gantrex Corporation.)

way around. The process outlined below closely follows the ACCO Crane Planning Guide for Steel Buildings.12 The first step involves determining the required hook coverage – width, length and height that the crane must meet. While plan dimensions are determined by equipment layout, vertical coverage (height) depends on the size of items being lifted, plus a tolerance for the height of any floor-mounted equipment to be removed and for spreaders or other equipment under the hook. The second step is to determine the type, lifting capacity and operational performance of the crane(s) using the information in this chapter or other guides. For example, a 25 ton CMAA Class B electric double girder crane combined with two 5 ton Class A jib cranes may be required in the short term but also for any likely future operational changes. When in doubt, it is best to invest in additional crane capacity early on. Crane operating power should not be chosen lightly. According to Dunville26, this is the most critical point affecting the cost of the crane. Dunville reports that in 1995 his company sold a 10-ton, 84-foot span, CMAA Class C (moderate duty) crane for $34,380, while another crane with the same capacity and span but the CMAA Class F (continuous heavy duty) sold for over $400,000! In addition, some manufacturers will not design monorail and overhead cranes for service performance above CMAA C.27 class (the same source suggests that the maximum monorail crane capacity does not exceed 5 tons and the overhead crane capacity does not exceed 10 tons). , finally with the dimensions of the building. The minimum free span of the building is found by adding the dimensions EG, B and C to those of the hook cover (Fig. 15.21). The minimum building height is calculated by adding dimensions A and R to the hook lift; is measured to the lowest point on the roof, either at the bottom of frame joists or an overhead sprinkler pipe. All of these dimensions are included in Figures 1 and 2. 15.8, 15.10 and 15.11. In the fourth step, the runway beams, their bearings and bracing methods are selected and dimensioned. Finally, the inside dimensions calculated in step 3 and the ones given in chap. 3. It is advisable to select a slightly larger building to allow for some design variability between manufacturers. The contract documents should specify who is supplying items that are likely to fall into the “grey area” of responsibility, such as: B. track carriers, rails and track stops. Unless provided by the crane supplier and specifically required by contract, these parts may be the responsibility of the metal structure manufacturer. In this case, all information about the crane, its wheel loads, outriggers, etc. must be included in the contract documents. Unless otherwise noted, fabricators of prefabricated buildings are likely to follow the impact tolerances, loading and deflection criteria given in the MBMA manual, still one of the best sources of crane-related information available. If stricter design standards are to be met by the builder and crane supplier, the relevant requirements must be included in the contract documentation.

SPECIFICATION OF CRANE STRUCTURES 444

CHAPTER FIFTEEN

FIGURE 15.21 Hook Cover. (FKI Industries, Inc.)

It is important to specify the allowable installation tolerances for steel rail supports because, according to AISC Design Guide 7, “The standard steel structure tolerances for buildings are not tight enough for buildings with cranes. Also, some of the required tolerances are not included in the standard specifications.” There have been instances where crane operation has been affected even when normal assembly tolerances are maintained.

REFERENCES 1. Specifications for Electric Multi-Girder Gantry Cranes, Specification CMAA 70, Crane Manufacturers Association of America, Inc., Charlotte, NC, 2000. 2. "Specifications for Electric Single Girder Cranes with Trolleys", CMAA Specification 74, Crane Manufacturers Association of America, Inc., Charlotte, NC, 2000. 3. Metal Building Systems Handbook, Metal Building Manufacturers Association, Cleveland, OH, 2002. 4. "Guide for the Design and Construction of Mill Buildings", AISE Technical Report 13, Association of Iron and Steel Engineers, Pittsburgh, PA, 1997. 5. James M. Fisher, "Industrial Buildings—Roofs to Column Anchorage", AISC Steel Design Guide 7, Chicago, 1993. 6. James M. Fisher and Steven J. Thomas, “Jib Crane Design Concepts,” AISC Engineering Journal, vol. 39, No. 2 (Second Quarter), 2002. 7. Specification for Structural Steel Buildings, Allowable Stress Design and Plastic Design, American Institute of Steel Construction, Chicago, 1989.

CRANE BUILDING SPECIFICATION CRANE BUILDING SPECIFICATION

445

8. International Building Code, International Code Council, Falls Church, VA, 2000. 9. Specifications for Overhead Cranes and Monorail Systems, ANSI MH 27.1-1981, Monorail Manufacturers Association, Pittsburgh, PA, 1981. 10. Overhead Cranes and Monorails, Monorail Manufacturing Association, Pittsburgh, PA, 1977. 11. Kunming Gwo, „Steel Interchange“, Modern Steel Construction, März 1995. 12. Steel Building Crane Planning Guide, ACCO Chain & Lifting Products Division, 76 ACCO Drive, Box 792, York , PA, 1986. 13. W. M. Weaver, Overhead Crane Handbook, Whiting Corporation, Harvey, IL, 1985. 14. Richard White und Charles Salmon (Hrsg.), Building Structural Design Handbook, Wiley, New York, 1987. 15. Julius P. Van De Pas und James M. Fisher, „Grane Girder Design: An Examination of Design and Fatigue Considerations“, Modern Steel Construction, März 1996. 16. James M. Fisher und Julius P. Van De Pas, „New Fatigue Provisions für die p Konstruktion von Kranbahnträgern“, AISC Engineering Journal al, vol. 39, Nr. 2 (Q2), 2002. 17. James M. Fisher und Michael A. West, „Maintenance Design Considerations for Low-Slender Buildings“, AISC Steel Design Guide 3, Chicago, 1990. 18. Steel Construction Handbook, Stress Permissible Design, 9. Ausgabe, American Institute of Steel Construction, Chicago, 1989. 19. Structural Beam Design Guide and Selection Chart for Overhead Crane Track System, ACCO Chain and Hoisting Products Division, York, PA, 1991. 20. John E. Mueller, „ Lehren aus Kran-Landebahnen“, AISC Engineering Journal, vol. 2, nein. 1, 1965. 21. John C. Roswell, „Crane Runway Systems“, MSc Thesis, Department of Civil Engineering, University of Toronto, 1987. 22. John C. Roswell und Jeffrey A. Packer, „Beam Tie Connections Crane Crane, Iron and Steel Engineer, Januar 1989. 23. P. H. Griggs und R. H. Innis, „Support Your Overhead Crane“ Proceedings, 1978 Annual Convention, Association of Iron & Steel Engineers, Chicago, 1978. 24. David T. Ricker, „Tips to Kranbahnprobleme vermeiden“, AISC Engineering Journal, vol. 19, Nr. 4, 1982. 25. Paul G. Kit, „Hydraulic Bumpers for the Protection of Buildings, Cranes and Operators from Impact Damage“, Iron and Steel Engineer, September 1997. 26. Larry Dunville, „Tips, Tricks and Traps: Overhead Cranes in Metal Buildings“, Metal Construction News, Mai 1998. 27. „Product and Engineering Manual“, Nucor Building Systems, Waterloo, IN, 2001.

REVIEW QUESTIONS 1 Which crane girder components resist horizontal thrust? 2 Which of the two crane systems is expected to be more expensive: (a) the 10 ton, 80 ft span, Class A ACM or (b) the 10 ton, 90 ft span, Class A E ACM? 3 The owner's valuation review focused on the horizontal anchoring connections between crane girders and building columns. The review suggested that the proprietary binding specified in the contract documents be replaced with a less expensive vertical plate welded to the beam and column. Should this proposal be accepted? Why or why not? 4 What are the three ways to support crane girders?

SPECIFICATION OF CRANE BUILDINGS 446

CHAPTER FIFTEEN

5 What modification to the frame rails should be made in the area of the monorail supports? 6 Is it acceptable to connect the crane rail and beam at the same place? Explain why or why not. 7 Which AISC document must be considered when dimensioning crane girders? 8 Which type of crane discussed in this chapter interferes the least with the metallic construction system?

Quelle: METALLIC CONSTRUCTION SYSTEMS

CHAPTER 16

AVOID CONSTRUCTION PROBLEMS

16.1 INTRODUCTION Any metallic building system, no matter how well constructed, can become a constant source of problems if improperly installed. Metalwork is a specialty where a builder's success depends on years of experience with a particular system or manufacturer. There is no perfect way to build a metal building, because different manufacturers propose slightly different assembly methods, and different teams' assembly techniques are different. It is not the purpose of this chapter to guide designers, owners and facility managers through every single task of a construction process. Aside from being inconvenient in this context, it's simply unnecessary for those readers who only visit the site regularly. Rather, our goal is to provide just a general idea of how metal fabrication should proceed, and to describe some common "red flags" that indicate problems. Knowing if the builder is following best practice - and what good practice is - is a valuable skill for anyone involved in the construction of metal building systems.

16.2 BEFORE YOU BEGIN THE STEEL INSTALLATION At this point we set the in chap. 9. Hopefully by this point all the required documents such as the design certification letter and design approval kit for the store have been reviewed, all colors selected and the site prepared. Of course, some construction—such as foundations—takes place before steel is erected. Any flat plate used can be placed either before or after erection of the metal structure. In conventional construction, the slab is usually built after the building is closed, but in some prefabricated buildings, the slab is placed first. These buildings include those with cantilevered walls that require a flat panel for wall casting and those that rely on panel-to-cast connections for lateral strength. Some topics of ceiling construction are discussed in Chap. 12. No matter how tight the project schedule, steel erection should not begin until the concrete foundations have sufficiently cured: "green" concrete will not hold the anchor bolts and may crack under structural loads. Ideally, concrete should set for 28 days, although in practice this time is reduced to a week. When time is critical, high strength concrete can be used, reaching the required strength in as little as 3 days. It may be of interest to the owner and approval engineer to observe the process of delivery, unloading and interim storage of the metalwork system; A general impression from this observation, whether favorable or not, is likely to be confirmed during construction. Manufacturers concerned about the quality of their systems don't ship packages of unmarked metal and let the builders handle everything. In fact, some building codes, such as BOCA, require1

447 Conveyed from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies. All rights reserved. All use is subject to the terms of use provided on the website.

AVOIDING CONSTRUCTION PROBLEMS 448

CHAPTER SIXTEEN

that each structural element, including side and top panels, is marked by the manufacturer - imprinted with the manufacturer's name or logo and the part number or name. A similar wording can be found in MBMA's Common Industry Practices manual.2 Unfortunately, complaints about deliveries without part markings are quite common. Some manufacturers even resort to using "universal" (one size fits all) purlins and slings to avoid confusion on the job site. For fitters, the way vans are packed can make a difference. The most sensible packing method is the reverse order of assembly so that the items required first are removed from the truck first. This type of packaging is reserved for resellers who require this in their contracts (order documents) with the manufacturers. Otherwise, manufacturers are free to load flatbed trailers as they wish. Upon delivery, the builder will check the shipment. If inspection reveals that packaged or nested metal components have become wet during transit, the manufacturer is expected to unpack and dry them to prevent rusting. Then the builder has the material stored, even if only for a short time. This is where caution, or lack thereof, comes into play. Careful builders follow proper procedures when lifting slender, cold-formed members that are prone to twisting and warping, bearing in mind that damage caused by rough handling of the "iron" is readily apparent. Components are to be stored according to the manufacturer's instructions. Roof and sidewalls are usually held in a slightly inclined position for drainage, while cold-formed beams and purlins can be stored flat to eliminate hard spots in the supports. Proper stowage keeps metal parts off the ground rather than letting them sink into the mud.3 It is best to keep casters with fiberglass insulation off the ground and cover them. Experienced builders store components in a logical way that aids, rather than hinders, future assembly. They also plan the assembly process in advance and know when each part of the build will be needed. The old adage – if you don't plan, you plan to fail – is very true when it comes to building metal structural systems.

16.3 ASSEMBLING THE MAIN STRUCTURES: THE BASICS 16.3.1 The Reinforced Span Now that the metalwork package is in the hands of the builder and the foundations have been properly constructed, cured and inspected, the steelwork can begin. The fitter can be the builder or a separate subcontractor. Typically, building fabricators are not involved in the actual construction - unless, of course, they erect the building directly - and they do not oversee or inspect the steel assembly process. However, for critical projects or where the expertise of the installer is in doubt, the project specification may require a competent manufacturer representative to be present on site throughout the construction process to ensure the building is assembled correctly. A careful decision must be made as to whether the additional expense and possibly the contractual uncertainty of such a representation is justified, since the client alone is responsible for the means, methods, techniques and construction processes. In any case, the manufacturer must provide assembly instructions, assembly drawings and printed instructions. Some manufacturers do not provide installation manuals due to differences in installation methods, local conditions and installer experience. The most common method of assembly begins with the construction of a bracing, which consists of two parallel frames connected by joists, purlins, and wall and roof bracing. The bracing field serves as a stable element against which adjacent superstructures and side walls can "lean" during assembly. It is usually located in the second compartment of an end wall. The erection process begins when two adjacent supports are brought into place using a crane or other equipment and temporarily stabilized by cross bracing. (For small rigid gantry cranes, the entire gantry can be floor mounted and installed as described below.) The columns are then joined together by one or two rows of wall studs, which provide some stability and allow the columns to be plumbed (Fig 16.1). .

AVOIDING CONSTRUCTION PROBLEMS AVOIDING CONSTRUCTION PROBLEMS

449

FIGURE 16.1 Assembling rigid multi-span frames. The left supports are stabilized by diagonal braces and wall struts. (Photo: Maguire Group Inc.)

After plumbing the columns, the wall bracing is tightened and the two columns on the opposite side wall are set up in a similar manner. Then the frame girders, pre-assembled on the ground, where the connections are much simpler than in the air, are lifted into place with a crane and bolted to the columns. The bolts are only tightened after the crane boom has been repositioned to create some slack in the cables, allowing the beams to stretch slightly under their own permanent load so the roof is secure. At the points where the transverse roof reinforcement is attached to the frames, some purlins, usually including the apex purlin, are installed to form a truss-like roof panel. Installation of column flange braces and rafter braces to the inner flanges, as shown in Figures 4.19 and 4.20, completes assembly of the braced shelf. Flange clamps protect against lateral buckling of constructions and should not be neglected at this point. In fact, some assembly manuals5,6 recommend installing column flange braces before placing the beams. Next, the final wall structure is erected. For spans under 60 feet, it can be pre-assembled in the ground, lifted into place as a unit, and supported by purlins and struts that extend into the enclosed bay. For spans greater than 60 feet, the end wall frame can be erected similarly to the inner frame4 or pre-assembled in staggered sections (due to common connections).7 Typical assembly details for a post and beam end wall are shown in Figure 16.2. The assembly then moves to adjacent frames which are laterally supported by struts and purlins attached to the braced span. Finally, the opposing front wall frame is set up, followed by a final check and cleaning. 16.3.2 Other methods of assembling frames A slightly different method is used for assembling fully assembled small single span rigid frames. With this method, the first structure is erected with a crane, probed and stabilized on both sides by clocking.

PREVENTING CONSTRUCTION PROBLEMS 450

CHAPTER SIXTEEN

FIGURE 16.2 Mullion and beam end wall details: (a) intermediate mullion; (b) Corner Post. (Butler Manufacturing Co.)

rare threads. After the second frame is erected, straps and purlins are installed to attach it to the first frame (Fig. 16.3). The permanent wall and roof supports are attached before the support cables are removed from the second structure. Yet another construction method is used for buildings with sloping concrete walls. Swivel walls are installed and fixed first so as not to interfere with the assembly of the structure. Wall bracing is best accomplished with temporary adjustable tubular supports anchored to the concrete slab, or concrete piers cast in the ground for this purpose. If purpose-built mounts are used, a qualified engineer

AVOIDING CONSTRUCTION PROBLEMS AVOIDING CONSTRUCTION PROBLEMS

451

FIGURE 16.3 Assembly of pre-assembled single span rigid frames with guy wires.

Neer needs to be maintained by the contractor to design the supports and their connections. Poorly planned panel bracing can result in panels snapping to the ground.8 Once erected and reinforced, pivot walls provide lateral stability for the sidewalls and columns of the sidewall structure. The first intermediate frame can now be connected to the bulkhead frame with purlins and beams (Figure 16.4). Proceed in a similar way for the other inner frames. The main steel assembly continues until all permanent roof braces are in place and all connections are made. Regardless of the installation method, the manufacturing and assembly tolerances contained in the MBMA manual general industry practices apply to most metal buildings. Crane designs require particularly tight tolerances because poor assembly can place excessive forces on the crane's bearing system and lead to a number of performance and durability issues.

16.3.3 The Critical Nature of Erection Reinforcement Whatever method of installation is preferred by installers, it must contain adequate erection reinforcement, which may be stronger than the permanent reinforcement of the building. Referring to common MBMA industry practices: "It cannot be assumed that the manufacturer's support for the metal structure system will be sufficient during assembly." thereby accommodating more wind load. The owner or recording engineer usually has no way of knowing what type of bracing is appropriate for the setup and cannot determine if the wrong type was used. However, some sort of temporary wall and roof support will certainly be needed. If none is present, or only one wall support is installed, the problem can loom large: prefabricated buildings with insufficient support have been known to collapse during construction. A case in point is documented by Sputo and Ellifritt,9 who describe the collapse of a rigid building in Florida with a free span of 206 feet. . They started by erecting the first rigid structure, only mounting cross braces on the side walls. Next, the second rigid structure was erected without reinforcement; Instead, the builders seemed to rely on purlins installed between the two structures and the front wall. No roof cross braces or frame flange braces were provided, although both were specified on the assembly drawings. Interestingly, a building inspector noticed the defect

AVOIDING CONSTRUCTION PROBLEMS 452

CHAPTER SIXTEEN

FIGURE 16.4 Steel assembly in a high-rise building. (Star building systems.)

Roof racks during a site inspection and mentioned this to the fitters as a courtesy. No corrective action has been taken. When the third rigid structure was placed in a similar manner the next day, the entire structure collapsed (Fig. 16.5). It later emerged that the builders had neither the assembly instructions nor the assembly drawings and had never worked on a building of this size. The authors conclude, "Insufficient support during erection likely contributes to more collapses of metal structural systems than all other factors combined."

16.4 INSTALLATION OF THERMAL AND GROUNDING As we have seen, some cross members and purlins required for bracing are installed along with the main frames, while the rest are placed right after. The girders and purlins can be bolted directly to the main frame steel or fastened to it with clamps as described in Chap. 5 and in Figs. 16.2 and 16.6. Installation details depend on whether the beams and purlins are intended to behave as single or continuous spans, the criticality of the crippling web stresses, and the design span and loading. It is often more economical to bring purlins to the roof in bundles than individually. The beams are placed near the eaves, from where fitters can manually push the individual purlins into the desired position.5 Purlin reinforcement, the meaning of which is given in Chap. 5, must be installed as soon as possible and definitely before the roof installation. Thin cold-formed C and Z profiles are easily deformed during construction. Sagging, twisted braces and purlins, a sad but common result of poor storage and installation practices, do not inspire confidence in the builder's work. Special care must be taken to ensure that these parts are not damaged by the assembly of equipment and by careless people using them to support ladders, tool boxes and similar equipment. Many manufacturers prevent purlins from rotating during installation by using temporary wooden blocks. These blocks work according to the same principle as those in Chap. 5. Installation of the secondary frame around the wall and ceiling openings completes the secondary steel assembly. Some problems in specifying this framework are discussed in Chap. 10

AVOIDING CONSTRUCTION PROBLEMS AVOIDING CONSTRUCTION PROBLEMS

453

FIGURE 16.5 A prefabricated building collapsed during construction. (Photo: Prof. Duane S. Ellifritt.)

16.5 FITTING THE INSULATION The insulation is fitted after the secondary steel is installed but before the cladding. In buildings with fiberglass insulation under fixed roofs, the roof is purlined through the insulation. Installing insulation typically begins with a 3 foot wide starter roll installed near a bulkhead. Subsequent full-width (4 to 6 foot) rolls are then attached to the purlins with self-tapping screws that must be the correct length. The 6 inch. and thicker bolts require longer bolts (11Ⲑ2 or 13Ⲑ4 in.) than those commonly used for roof fasteners to avoid tightening the insulation enough to set the panel back. Attaching insulation to the eaves and edges of framed openings requires some finesse. Most manufacturers recommend leaving the roof insulation over the edge of the frame, stripping about 6 inches of fiberglass from the face, and then folding the face over the insulation to prevent moisture pick-up (Fig 16.7). The wall insulation is hung from the eaves bracket, temporarily held with Vise Grip pliers or similar fasteners (Fig 16.8), pulled down to give a smooth, taut inner surface, and fastened to the eaves bracket and base. Attaching insulation to the roof deck is a dangerous job as no cover has yet been installed to support workers in the event of a fall. Butler Manufacturing Company uses their proprietary SkyWeb* Insulated Support and Fall Protection System, essentially a 1Ⲑ2-inch coated polyester mesh, to provide fall protection at the leading edge of the roof. Mesh has the added benefit of providing protection from falling objects to workers below and, in some cases, supporting the roof insulation. *Sky-Web is a registered trademark of Butler Manufacturing Company.

AVOIDING CONSTRUCTION PROBLEMS 454

CHAPTER SIXTEEN

FIGURE 16.6 Details of belt installation (bypass belts shown): (a) restraint belt bolted directly to the abutment; (b) Continuous beams simultaneously lapped and bolted to the column. (Butler Manufacturing Co.)

FIGURE 16.7 Detail of insulation at eaves and base. (VP building.)

AVOIDING CONSTRUCTION PROBLEMS AVOIDING CONSTRUCTION PROBLEMS

455

FIGURE 16.8 Temporary attachment of fiberglass wall insulation to eaves brace. (Butler Manufacturing Co.)

16.6 INSTALLATION OF ROOF AND WALL PANELS Whether roof or wall panels are installed first is at the contractor's discretion. We suggest raising the roof first, so that any interior work can begin and to improve the effect of the roof pane on the partially erected building (Fig. 16.9). Installation of roof panels usually begins at one of the selected end walls to allow placement of the panel away from prevailing winds; This order is intended to reduce the likelihood of windblown water entering the sides of the panel. The process begins with the roof panels being lifted in bundles by crane and placed directly over the main structures. Each purlin supporting the joists receives a wooden block that fits snugly between its top edge and the framing joists to protect the purlin from deformation. Roof panels are laid without fastening from the eaves to the ridge or from the lower eaves to the upper. On the ridge line, the panels are held at the distance specified in the manufacturer's specifications. After further adjustments to the recommended panel overlapping seams and small eaves overhang distances, sealants are applied if necessary and the roof secured. The process then moves to the next row of panels which are installed in a similar manner except the panel seams must now be formed by hand or by mechanical stitching. A common phenomenon when installing ribbed metal roofing is the "swelling" and "shrinking" of the panel width. Fitters can inadvertently increase the width of the panel by stepping on the ribs and partially flattening them. On the other hand, they can reduce the width by not applying enough pressure when bringing the cover into contact with the purlins. These dimensional changes for each panel are small, but the cumulative error can be large enough to be noticeable. In these circumstances, a special jig or spacer is helpful.6 Stepping on panel ribs is strongly discouraged by manufacturers, as is walking on partially fixed or loose panels. The safest way to walk on a fully paved roof is to step on walking boards placed in the flat areas of the panels and extending between the purlins. To avoid skidding, the walk

AVOIDING CONSTRUCTION PROBLEMS 456

CHAPTER SIXTEEN

FIGURE 16.9 The roof of this prefabricated building is installed in front of the walls. (Photo: Maguire Group Inc.)

The boards must be fixed to the roof. If stepping on the panels is unavoidable, try to walk as directly over the purlins as possible and stay away from the center of the flat part of the panels. Wall panels are mounted similarly to the roof. To minimize the visibility of vertical seams, panels are best lifted in a direction where their overlaps face away from the main viewpoint of the building. On-site cutting of panels, particularly those with state-of-the-art galvanized steel coatings, weakens the panels' resistance to corrosion by exposing bare metal, although the galvanizing protects the cut edges to some extent. For this reason alone, pre-cut factory panels are preferable to site-formed. If it is unavoidable, the panel cutting must be done carefully and precisely on the ground. The edges must be touched up with a special putty from the manufacturer; all metal dust and shavings must be removed promptly so that rust stains quickly develop in wet weather.

16.7 SOME COMMON PROBLEMS DURING CONSTRUCTION Failure to follow best practices will inevitably lead to problems with the appearance, function or durability of a newly constructed metal building. Some of these glitches, which occur with discouraging regularity, are detailed below.

AVOIDING CONSTRUCTION PROBLEMS AVOIDING CONSTRUCTION PROBLEMS

457

16.7.1 Problems with missing or misaligned slab and foundation anchor bolts. If anchor bolts aren't placed with a jig, they're likely to end up in the wrong position, causing a mini-crisis that entails frantic calls to the designer and manufacturer. A few simple fixes, like drilling new holes in the column baseplate and drilling new extensions or chemical anchors, can solve the problem. If not, a new, larger base plate or an extension of the existing one may be required; In really critical cases it may even be necessary to replace the foundation. All anchor bolts in oversized holes must be fitted with thick washers under the nuts. Washers are typically 5/16 to 1/2 thick; this thickness should be considered as 10 when screw protrusion is detailed. Anchor bolts that do not protrude high enough to allow proper nut engagement are no less of a problem. The best solution to this common problem is to lengthen the bolts by welding short lengths of threaded rod together; The welded end of each extension piece must be cut at 45° to allow full weld penetration. Alternatively, a special screw coupler can be used for splicing, which may require removing some metal from the baseplate and concrete. In either case, some washers are needed to lift the nut above the fastener material in the threaded area of the bushing. The contractor's usual suggestion of filling the cavity in the nut with solder does not make a strong connection and deserves to be rejected, particularly for bolts designed to withstand tensile loads. A classic example of the failure of an anchor bolt welded in this manner occurred at a school in Louisiana in high winds. When welds and threads of minimal length suddenly failed, a spring-like action occurred and columns with attached beams were thrown into the air.11 Plate cracks or waviness. Cracked floor tiles are the constant source of complaints from homeowners. Most of the many possible reasons for this crack have to do with poor build quality. Dry shrinkage cracks that occur within days of slab installation are usually caused by a lack of proper control and construction joints. For example, a popular control connection detail requires that all other strands of welded wire mesh be cut at the control connection points. The detail doesn't work if the wire cutting isn't done - a common oversight - and the board isn't weakened enough at the joints to cause cracks there. As a result, the plate will crack elsewhere. Large cracks in the slab, accompanied by settlements, indicate insufficient preparation of the subsoil; Most other cracks can be attributed to insufficient curing of the panel. Waviness of slab edges usually results from improper detailing and execution of construction joints. For example, it is known that keyed construction joints are more prone to corrugation than dowel joints. The use of vapor barriers under the slab without sand cushions has also been linked to slab curling. Individually. 12 gives some recommendations for building better slabs on slopes and avoiding cracks and bumps in slabs. What to do with a cracked or wavy record? Corrective action can range from doing nothing for small cosmetic cracks and pits to filling the cracks with epoxy or completely replacing the slabs for critical super-flat flooring. Problems in repairing level boards are discussed at length by Newman.12 Misplaced or missing wall anchors. Rebar bushings extending from foundation walls to a sloping slab may be required for a number of reasons. Dowels are usually required to support the panel to allow it to be passed over poorly compacted soil adjacent to the wall and to laterally support the top of the wall. These studs must be bent on site as studs protruding from the wall cannot compact the subsoil. Unfortunately, regardless of what is shown on the design drawings, dowels are usually supplied pre-bent and then re-bent on site - out of the way. Or there are no dowels at all and require subsequent drilling and grouting. Or the sockets are too short to use. Due to their large number, poorly placed anchors can become a significant source of friction between the foundation contractor and the homeowner.

AVOIDING CONSTRUCTION PROBLEMS 458

CHAPTER SIXTEEN

16.7.2 Problems with gaps in the metal superstructure under the column base plates. In conventional designs, column bases typically rest on cast leveling plates or are positioned over special leveling nuts. Very large base plates can be installed separately from the column in a mortar bed and then welded to it. In any case, the aim is to ensure complete support under the support plates and the plumb of the support. Leveling boards and joint mortar are often dispensed with in prefabricated construction. The MBMA Industry Common Practices specifically exclude "grouting or filling work of any kind under columns" from the work of the fabricator. The concrete supplier is probably long gone by the time mortar is needed. Who should do this? It is not surprising that precast columns, which must be within MBMA tolerances, are often placed directly on top of concrete piers or foundation walls that are not perfectly level. As a result, the column base plates can only support the concrete on one edge, with a slight gap under the rest of the plate (Fig. 16.10). If this gap is not filled with grout or washers, the concrete may crack or the column base may deform slightly and become loaded, creating a "gap" between the base plate and the anchor nut above. The base plates of building columns that support cranes and therefore require stability and accuracy of installation must always be grouted, possibly also by the owner's staff. Loose roof or wall bracing. Rod or cable mounts installed on roofs and walls of metal buildings are often observed to be loose, bent or even missing. Such bracing does not serve its purpose of stabilizing the building and ensuring that it can withstand external loads without excessive deformation. Cumulative movement of the structure, which can occur before the slack support has stretched sufficiently to be effective, can shatter skylights and windows, jam doors, and disrupt the operation of equipment carried by the structure—similar to excessive history drift . . . Too large a "gap" can also damage brittle wall coatings and result in noticeable noise and vibration. Luckily, this common erectile dysfunction is easy to spot and fix. Lateral bracing for missing or poorly secured primary and secondary beams. The importance of properly laid purlins and flanges is discussed in Chap. 5. Equally important is flange reinforcement for internal flanges of columns and beams when required for the design. However, both of these types of braces are regularly found missing, improperly connected, or incorrectly installed (e.g., a parallel purlin brace is installed where a cross brace was specified). This deficiency is also usually easy to remedy if it is noticed in good time before the interior fittings and insulation are installed.

16.7.3 Roof Leaks One of the most common complaints about metal building systems is metal roof leaks. At the design stage, the likelihood of leaks can be greatly reduced if the roof type and details are properly selected, as discussed in Chap. 6 and 14. However, the most common reason for leaks is poor design. An example: the watertightness of vertical joint covers with trapezoidal seams depends very much on the professional installation of the corrugated end strips on the eaves. That detail isn't perfect, even under the best of circumstances. However, if the terminations are simply omitted or poorly sealed, leakage is virtually guaranteed as there is nothing to prevent water from draining from an overflowing gutter or ice dam. Two other examples are unprotected roof penetrations - a notoriously fertile ground for leaks - and the omission of necessary sealants. As Star Manufacturing Company's Installation Guide5 states, 99% of leaks can be traced back to:

AVOIDING CONSTRUCTION PROBLEMS AVOIDING CONSTRUCTION PROBLEMS

459

FIGURE 16.10 A gap under the column base.

●

● ●

Omitting or misplacement of required sealing strips and caulking in the voids of the roof rails Failure to apply the additional sealing strip known as a pigtail to the four-way panel overlays and eaves joints Failure to apply the sealing tape under the screw heads No caulking between the eaves edge and underside of the roof panels

The presence of the required sealants at these critical locations can be checked with a special "feeler" tool made of material thin enough (0.005 inch) to fit into the seams of the roof. Wherever the seal is missing, the only safe repair method is to remove the offending panel and reapply the seal. These tests are time consuming and well beyond the experience of most homeowners and design professionals for whom few other installation quality guarantees are available. The most basic and simple precaution - checking the installer's success rate in making watertight roofs - should of course be taken before signing the contract.

PREVENTING CONSTRUCTION PROBLEMS 460

CHAPTER SIXTEEN

16.7.4 Construction Acceptance There may be many other construction residue defects, major and minor, structural and cosmetic, that will be identified by the owners and their design professionals towards the end of the work. The elements related to structure are the most serious. In addition to those described above, they can include poorly made frame connections such as loose, missing, or improperly tightened bolts and poor welds. Some elements may be poorly positioned or have insufficient load-bearing length. Exposed panel fasteners may not be properly aligned or properly tightened; Undertorque on screws with neoprene washers can leave penetration unprotected; If you apply too much, the panel may curl. Identifying these problems is best left to experienced building inspectors or owner-contracted engineers. Sloppy fit causes appearance problems, which can't lead to structural problems, but certainly emotional problems. It's hard to miss out of plumb door jambs, 1/2 inch wide caulked joints and sagging gutters. Field formed roof and wall panels seem to suffer more than their share from problems with rusting, buckling, oil exposure and improper seating. A convenient order fulfillment checklist used by Star Building Systems contractors, similar to that originally developed by MBMA, is shown with permission in Fig. 16.11. The checklist is primarily intended for steel installers and can be of use to both homeowners and design professionals in their quest to realize the full benefits of metalwork systems. A properly designed metal building system will provide an aesthetically pleasing, practical and virtually maintenance free environment for many years.

REFERÊNCIAS 1. O BOCA National Building Code, Building Officials and Code Administrators International, Inc., Country Club Hills, IL, 1999. 2. Metal Building Systems Manual, Metal Building Manufacturers Association, Cleveland, OH, 2002. 3. Metal Building Systems , 2ª ed., Building Systems Institute, Inc., Cleveland, OH, 1990. 4. Donna Milner, „Metal Building Basics“, The Journal of Light Construction, April 1989. 5. Guia de Montagem, Star Building Systems, Star Manufacturing Co., Oklahoma City, OK, 1989. 6. Manual de Montagem, Steelox Systems, Inc., Mason, OH, 1994. 7. Widespan Buildings, Erection Information, Butler Manufacturing Co., Kansas City, MO, 1978. 8 „Diretrizes para Bracing Tilt-up Walls“, Concrete Construction, Dezember 1995. 9. Thomas Sputo und Duane S. Ellifritt, „Collapse of Metal Building System during Erection“, Journal of Performance of Constructed Facilities, vol. 5, nein. 4, November 1991, Sociedade Americana de Engenheiros Civis. 10. David T. Ricker, „Some Practical Aspects of Column Base Selection“, AISC Engineering Journal, Third Quarter, 1989. 11. Pesquisa von der Texas Tech University, zitiert von Thomas C. Powell in „Steel Interchange“, Modern Steel Construction, Dezember 1992. 12. Alexander Newman, Renovação Estrutural de Edificios: Métodos, Detalhes e Exemplos de Projeto, McGraw-Hill, Nova York, 2001.

AVOIDING CONSTRUCTION PROBLEMS AVOIDING CONSTRUCTION PROBLEMS

FIGURE 16.11 Job Completion Checklist. (Star building systems.)

461

AVOIDING CONSTRUCTION PROBLEMS 462

CHAPTER SIXTEEN

FIGURE 16.11 (continued) Job Completion Checklist. (Star building systems.)

AVOIDING CONSTRUCTION PROBLEMS AVOIDING CONSTRUCTION PROBLEMS

FIGURE 16.11 (continued) Job Completion Checklist. (Star building systems.)

463

AVOIDING CONSTRUCTION PROBLEMS 464

CHAP

[PDF] CHAPTER 1 METALLIC CONSTRUCTION SYSTEMS: YESTERDAY AND TODAY - Free download in PDF (2023)