Building codes play a dominant role in defining the level of fire protection that is expected by society. Typically, fire protection is implemented in design through code compliance. As a consequence, a working knowledge of building codes is an important prerequisite for contemporary design.
In the past, keeping abreast of building codes was difficult, even for the largest design offices, since most major cities and a number of states maintained locally developed codes.
Today, this impediment is less relevant. In the United States and Canada, with a few notable exceptions, the vast majority of cities, states, and provinces now enforce one of the following model codes:
• National Building Code, Building Officials and Code Administrators International, Homewood, Ill.
• Standard Building Code, Southern Building Code Congress International, Birmingham, Ala.
• Uniform Building Code, International Conference of Building Officials, Whittier, Calif.
• National Building Code of Canada, National Research Council of Canada, Ottawa, Ontario Two fire-related characteristics of materials influence selection and design of structural systems:
combustibility and fire resistance.
Combustible and Noncombustible Materials
Most fires are either accidental or caused by carelessness. Fires are usually small when they start and require fuel to grow in intensity and magnitude. In fact, many fires either selfextinguish due to a lack of readily available fuel or are extinguished by building occupants.
Furthermore, even though most fires involve building contents, a combustible building itself may be the greatest potential source of fuel.
By definition, noncombustible materials such as stone, concrete, brick, and steel do not burn and therefore do not serve as sources of fuel. Although the physical properties of noncombustible materials may be adversely affected by elevated temperature exposures, these materials do not contribute to either the intensity or duration of fires. Wood, paper, and plastics are examples of combustible materials.
Tests conducted by the National Institute for Standards and Technology (formerly the National Bureau of Standards) indicate that an approximate relationship exists between the amount of available combustible material (fire loading, pounds of wood equivalent per square foot of floor area) and fire severity, hours of equivalent fire exposure (Fig. 6.7).
Subsequent field surveys measured the fire loads typically found in buildings with different occupancies (Table 6.35).
A reasonable estimate of the structural fire loading for conventional wood-frame construction is 71⁄2 to 10 psf. For heavy-timber construction, the corresponding structural fire load may be on the order of 121⁄2 to 171⁄2 psf. As a consequence, building codes generally limit the permitted size allowable height and area) of combustible buildings to a much greater degree than for noncombustible buildings.
In addition to regulating building construction based on the combustibility or noncombustibility of structures, building codes also specify fire-resistance requirements as a function of building occupancy and size, i.e., height and area. In general, fire resistance is defined as the relative ability of construction assemblies, such as, floors, walls, partitions, beams, girders, and columns, to prevent spread of fire to adjacent spaces or perform structurally when exposed to fire. Fire-resistance requirements are based on tests conducted in accordance with ‘‘Standard Methods of Fire Tests of Building Construction and Materials’’ (ASTM E 119).
The ASTM E119 test method specifies a ‘‘standard’’ fire exposure that is used to evaluate the fire resistance of construction assemblies (Fig. 6.8). Fire-resistance requirements are specified in terms of the time during which an assembly continues to prevent the spread of fire or perform structurally when exposed to the ‘‘standard fire.’’ Thus fire-resistance requirements are expressed in terms of hours or fractions thereof. The design of fire-resistant buildings is typically accomplished in a prescriptive fashion by selecting tested ‘‘designs’’ that meet specific building code requirements. Listings of fire-resistance ratings for construction assemblies are available from a number of sources:
Fire-Resistance Directory, Underwriters Laboratories, Northbrook, Ill.
• Fire-Resistance Ratings, American Insurance Services Group, New York, N.Y.
• Fire-Resistance Design Manual, Gypsum Association, Washington, DC
In the past, the term fireproof was frequently used to describe fire-resistant buildings. The use of this and terms such as fireproofing is unjustified and should be avoided. Experience has clearly demonstrated that large-loss fires (in terms of both property losses and loss of life) can and do occur in fire-resistant buildings. No building is truly fireproof. (Fire Protection Handbook, National Fire Protection Association, Quincy, Mass.).
Effect of Temperature on Steel
The properties of virtually all building materials are adversely affected by the temperatures developed during standard fire tests. Structural steel is no exception. The effect of elevated temperatures on the yield and tensile strengths of steel is described in Art. 1.12. In general, yield strength decreases with large increases in temperature, but structural steels retain about 60% of their ambient-temperature yield strength at 1000F.
During many building fires, temperatures in excess of 1000F develop for relatively brief periods of time, but failures do not occur in structural steel members inasmuch as they are rarely loaded enough to develop full design stresses. As a consequence, in many instances, bare structural steel has sufficient load-carrying capacity to withstand the effects of fire. This is not recognized in the standard fire tests, however, inasmuch as, in the tests, the temperatures are continuously increased and structural members are loaded to design capacities.
Based on these tests, when building codes specify fire-resistant construction, they require fireprotection materials to ‘‘insulate’’ structural steel elements.
A variety of different materials or systems are used to protect structural steel. The performance of these are directly determined during standard fire tests. In addition to the insulation characteristics evaluated in the tests, the physical integrity of fire-protection materials is extremely important and should be preserved during installation. Required fire-protection
assemblies should be carefully inspected during and after construction to ensure that they are installed according to the manufacturers’ recommendations and the appropriate fireresistant designs.
Gypsum. This material, in several forms, is widely used for fire protection (Fig. 6.9). As a plaster, it is applied over metal lath or gypsum lath. In the form of wallboard, gypsum is typically installed over cold-formed steel framing or furring.
The effectiveness of gypsum-based fire protection can be increased significantly by addition of lightweight mineral aggregates, such as vermiculite and perlite, to gypsum plaster.
It is important that the mix be properly proportioned and applied in the required thickness and that the lath be correctly installed.
Three general types of gypsum wallboard are readily available: regular, type X, and proprietary. Type X wallboards have specially formulated cores that provide greater fire resistance than conventional wallboard of the same thickness. Proprietary wallboards also are available with even greater fire-resistant characteristics. It is therefore important to verify that the wallboard used is that specified for the desired fire-resistant design. In addition, the type and spacing of fasteners and, when appropriate, the type and support of furring channels should be in accordance with specifications.
(‘‘Design Data—Gypsum Products,’’ Gypsum Association, Washington, D.C.)
Spray-Applied Materials. The most widely used fire-protection materials for structural steel are mineral fiber and cementitious materials that are spray applied directly to the contours of beams, girders, columns, and floor and roof decks (Fig. 6.10). The spray-applied materials are based on proprietary formulations. Hence it is imperative that the manufacturer’s recommendations for mixing and application be followed closely. Fire-resistant designs are published by Underwriters Laboratories.
Adhesion is an important characteristic of spray-applied materials. To ensure that it is attained, the structural steel should be free of dirt, oil, and loose scale; generally, the presence of light rust will not adversely affect adhesion. When the steel has been painted, however, field experience and testing have demonstrated that adhesion problems can arise. (Paint and primers are not generally required for corrosion protection when structural steel will be enclosed within a building or otherwise protected from the elements.) If paint is specified for structural steel that will subsequently be protected with spray-applied materials, the specifier should contact the paint and fire-protection material suppliers in advance to ensure that the two materials are compatible. Otherwise, bonding agents or other costly field modifications may be required to ensure adequate adhesion.
Suspended Ceiling Systems. A wide variety of proprietary suspended ceiling systems are also available for protecting floors and beams and girders (Fig. 6.11). Fire-resistance ratings for such systems are published by Underwriters Laboratories. These systems are specifically designed for fire protection and require careful integration of ceiling tile, grid, and suspension
systems. Also, openings for light fixtures, air diffusers, and similar accessories must be adequately protected. As a consequence, manufacturer’s installation instructions should be closely followed.
In the case of load-transfer trusses or girders that support loads from more than one floor, building codes may require individual protection. As a consequence, suspended ceiling systems may not be permitted for this application.
Concrete and Masonry. Concrete, once widely used for fire protecting structural steel, is not particularly efficient for this application because of its weight and relatively high thermal conductivity. As a result, concrete is rarely used when the purpose is fire protection only.
Concrete floor slabs are acceptable as fire protection for the tops of flexural members.
Concrete or masonry is also sometimes used to encase steel columns for architectural or structural purposes or when substantial resistance to physical damage is required (Fig. 6.12).
Design information on the fire resistance of steel columns encased in concrete or protected with precast-concrete column covers is available from the American Iron and Steel Institute, Washington, D.C. Information on the use of concrete masonry and brick to protect steel columns may be obtained from the National Concrete Masonry Association, Herndon, Va., and the Brick Institute of America, Reston, Va., respectively.
Architecturally Exposed Steel
This concept involves the architectural expression of structural systems on building exteriors in contrast to the general practice of concealing them behind decorative facades. Design of architecturally exposed steel is strongly influenced by building code requirements for fireresistant construction.
One approach for meeting code requirements for structural fire protection when the appearance of architecturally exposed steel is desired is illustrated in Fig. 6.13. As shown, flanges of a steel column are fire protected with a spray-applied material, insulation is placed against the web between the flanges, and the assembly is enclosed in a metal cover with the shape of the column.
Another approach is to use tubular columns filled with water (Fig. 6.14). Originally patented in 1884, this system was neglected until the late 1960s, when it was adopted for the 64-story U.S. Steel Building in Pittsburgh, Pa. Since then, several other buildings have been designed using this concept. In a fire, the entrapped water in a tubular column is expected to limit the temperature rise in the steel. Generally, corrosion inhibitors should be added to the water, and in cold climates, antifreeze solution should be used for exterior columns. (Fire Protection through Modern Building Codes, American Iron and Steel Institute, Washington, D.C.)
In still another approach, sheet-steel covers are applied on the outside of a building to insulated flanges of steel spandrel girders to act as flame shields, as illustrated in Fig. 6.15. These sheet-steel covers not only serve to deflect flames away from the exposed, exterior web of a girder but also provide weather protection for the insulated flanges. As shown, a flame-shielded spandrel girder is protected in the interior of the building in a conventional manner.
As illustrated by full-scale fire tests on flame-shielded spandrel girders, the standard fire test is not representative of the exposure that will be experienced by exterior columns and girders. Research on fire exposure conditions for exterior structural elements has led to development of a comprehensive design method for fire-safe exterior structural steel which has been adopted by some building codes.
(Design Guide for Fire-Safe Structural Steel, American Iron and Steel Institute, Washington, D.C.)
Restrained and Unrestrained Construction
One of the major sources of confusion with respect to design of fire-resistant buildings is the concept of restrained and unrestrained ratings. Fire-resistant design is based on the use of tested assemblies and is predicated on the assumption that test assemblies are ‘‘representative’’ of actual construction. In reality, this assumption is extremely difficult to implement in laboratory-scale fire tests. The primary difficulty arises from the size of available test furnaces, which typically can only accommodate floor specimens in the range of 15 by 18 ft in area. As a result, a typical test assembly actually represents a relatively small portion of a floor or roof structure. Thus, even though the standard fire test is frequently described as ‘‘large scale,’’ it clearly is not ‘‘full scale.’’
In the attempt to model real floor systems in a representative manner, several problems arise. For example, since most floor slabs and roof decks are physically, if not structurally, continuous over beams and girders, real beams and girders are usually much larger than can be accommodated in available furnaces. Also, beams frame into columns and girders in a number of different ways. In some cases, connections are designed to resist only shear forces.
In other cases, full- or partial-moment connections are provided. In short, given the cost of testing, the complexity of modern structural systems, and the size of available test facilities, it is unrealistic to assume that test assemblies can accurately model real construction systems.
In recognition of the practical difficulties associated with testing, ASTM E119 includes two test conditions, restrained and unrestrained. The restraint that is contemplated in fire testing is restraint against thermal expansion, not structural restraint in the traditional sense.
When an assembly is supported or surrounded by construction that is capable of resisting expansion, to some degree, thermal stresses will be induced in the assembly in addition to those due to dead and live loads. Originally, it was thought that thermal stresses would reduce the fire resistance of many assemblies. However, extensive research indicated that restraint actually improved the fire resistance of many common types of floor systems. The two test conditions in E119 recognize the complexity of this issue.
The restrained condition applies when the assembly is supported or surrounded by construction that is capable of resisting substantial thermal expansion throughout the range of anticipated elevated temperatures. Otherwise, the assembly should be considered free to rotate and expand at the supports and should be considered unrestrained. Thus a floor system that is simply supported from a structural standpoint may often be restrained from a fireresistance standpoint. To provide guidance in the use of restrained and unrestrained ratings, ASTM E119 includes examples in an explanatory appendix (Table 6.36) which indicate that most common types of steel framing systems can be considered to be restrained from a fireresistance standpoint.
Temperatures of Fire-Exposed Structural Steel Elements
Basic heat-transfer principles indicate that the rate of temperature change of a beam or column varies inversely with mass and directly with the surface area through which heat is transferred to the member. Thus the weight-to-heated perimeter ratio W/D of a structural steel member significantly influences the temperature that the member will experience when exposed to fire. W is the weight per unit length of the member (lb / ft), and D is the inside perimeter of the fire protection material (in). Expressions for calculating D are illustrated in Fig. 6.16 for columns and beams with either contour or box protection. In short, the weightto- heated-perimeter ratio defines the thermal size of a structural member.
Since the temperature of a structural steel member is strongly influenced by W/D, it therefore follows that the required thickness of fire-protection material is also strongly influenced by W/D. This interrelationship is clearly illustrated in Fig. 6.17, which gives the fire resistance of steel columns protected with different thicknesses of gypsum wallboard as a function of W/D. The curves show that in determination of fire resistance, W/D is significant, as is the thickness of the fire-protection material.
In recognition of this basic principle, several semiempirical design equations have been developed for determining the thicknesses of fire protection for structural steel elements as a function of W/D for specific fire-resistance ratings. These equations have been incorporated into the Underwriters Laboratories Fire-Resistance Directory and are described in the following publications available from the American Iron and Steel Institute: Designing Fire Protection for Steel Columns, Designing Fire Protection for Steel Beams, and Designing Fire Protection for Steel Trusses. These calculation methods also have been recognized by model building codes and are widely used in design of cost-effective, fire-resistant steel buildings.
Rational Fire Design
Building code requirements for structural fire protection generally are prescriptive and are based on standard fire tests. This approach suffers from the following significant deficiencies:
• The fire exposure is arbitrary and does not necessarily represent real building fires. In many cases, real fires result in high temperatures of short duration.
• In effect, the standard fire test presumes that structural members will be fully loaded at the time of a fire. In reality, fires occur randomly, and design requirements should be probability-based. Rarely will design structural loads occur simultaneously with fire.
• Given the scale of available laboratory facilities, structural interaction cannot be directly evaluated. In effect, ASTM E119 unrestrained ratings presume virtually no structural interaction, i.e., simple supports without continuity. To some degree, structural interaction is indirectly considered in establishing restrained ratings. The boundary conditions are arbitrary, however, and the extension to real buildings is largely based on judgment.
As a consequence of these shortcomings, a rational engineering design standard for structural fire protection is desirable. Standards of this type have been developed and are now routinely used in Japan, Australia, and throughout much of Europe (International Fire Engineering Design for Steel Structures: State-of-the-Art, International Iron and Steel Institute, Brussels, Belgium) and are being developed in the United States by the American Society of Civil Engineers in cooperation with the Society of Fire Protection Engineers.
(The SFPE Handbook of Fire Protection Engineering, National Fire Protection Association, Quincy, Mass.)