Thermal Restraint and Fire Protection of Structures

To view the figures and tables associated with this article, please refer to the flipbook above.

Beginning in the early 1970s, design of fire protection has been governed by a structural element’s classification as thermally “restrained” or “unrestrained.” While these designations are given to the building community as a basis of design, what constitutes thermal restraint is still a topic of contention. This article hopes to clarify the use of restrained versus unrestrained assemblies for the purposes of fire protection design.”

In the United States, structural design for fire protection is handled by numerous codes and standards: Chapters 7 and 9 of the International Building Code (IBC), Appendix 4 of the American Institute of Steel Construction’s AISC 360 Specification for Structural Steel Buildings, the American Concrete Institute’s ACI 216 Code Requirements for Determining Fire Resistance of Concrete and Masonry Construction Assemblies, Chapter 16 of the American Wood Council’s (AWC) National Design Specification for Wood Construction (NDS), and the AWC’s Fire Design Specification (FDS) for Wood Construction. These codes and standards provide guidance on how to protect various structural materials from fire.

When designing fire protection in accordance with the building code, particularly steel-framed assemblies, most design professionals are interested in the hourly resistance rating of structures as defined for specific types of building construction. The hourly rating depends on whether the assembly is considered thermally restrained or unrestrained.

The American Society for Testing and Materials’ ASTM E119 Standard Test Methods for Fire Tests of Building Construction and Materials and the Underwriters Laboratory’s UL263 Standard for Fire Tests of Building Construction and Materials define a thermally restrained condition as “one in which expansion and rotation at the ends and supports of a load carrying assembly resulting from the effect of a fire are resisted by forces external to the assembly exposed to fire.” Put another way, a thermally restrained structure is one that can resist or redistribute loads to elements which are not affected by fire.

Both restrained and unrestrained ratings are developed from the same ASTM E119 fire tests. Assemblies can be tested as either loaded or unloaded. Unrestrained assemblies are typically built of bearing or hanger supported members and are allowed to rotate and deflect freely during the test. Restrained assemblies are built tight to the testing frame and shimmed in place to restrain motion or rotation. The temperature-based failure criteria for restrained assemblies are similar to those for unrestrained assemblies; however, the difference is in the time allowed to reach those temperatures. For example, ASTM E119 states that for test specimens employing steel structural members (beams, open-web steel joists, etc.) spaced more than 4 feet on center, the test specimen shall achieve a restrained assembly classification on the basis of the temperature of the steel structural members not having exceeded 1,300F at any location and not having the average temperature recorded by four thermocouples at any section exceed 1,100F during the first hour. For restrained assembly classifications greater than 1 hour, these temperature criteria shall apply for a period of one half the classification period of the floor or roof construction or 1 hour, whichever is the greater. For unrestrained assemblies, the temperature-based failure criteria is the same, however for unrestrained beam classifications, the beam may not exceed these temperatures at any point during the test period.

The temperature-based failure criteria changes for different types of assemblies, but the result is the same: unrestrained assemblies will generally require between 50% and 100% more fire protection material than restrained assemblies. Because of how fire protection is handled, this matters most in the realm of fire protection of structural steel. Before and after photographs of an E119 fire test on spray-fireproofed structural steel can be seen Figure 1.

Since the introduction of “restrained vs. unrestrained” classifications in the early 1970s, the building community has argued over how to properly apply these classifications. Appendix X3 of ASTM E119 has provided some guidance on how to address this issue, but not necessarily clarity.

Table X3.1 in Appendix X3 of ASTM E119 provides an outline of common construction types and whether they can be described as restrained or unrestrained. However, various locations throughout Appendix X3 reference these classifications requiring “sufficient restraint” without defining what that means, practically.

This raises the obvious question: what constitutes “sufficient restraint?” How much is enough? Can it ever be enough? Is anything thermally restrained? Neither ASTM E119 nor UL263 provide guidance on what that means beyond “engineering judgment.” To quote ASTM E119, “The description provided in X3.5 requires the exercise of engineering judgment to determine what constitutes restraint to ‘substantial thermal expansion and rotation.’”

UL263 does describe the stiffness of the test assembly used to perform structural fire tests as being constructed within a nominal 14-feet x 17-feet frame of composite steel/concrete cross sections having an approximate stiffness of EI/L of 850,000 kip-in and 700,000 kip-in along the 14-feet x 17-feet sides respectively. This stiffness is assumed not to change during the test as the frame is thermally isolated from the fire environment. However, this has been incorrectly taken to mean that structural assemblies in the field must match this level of stiffness to be considered restrained. This is not an accurate interpretation of the information provided.
Some in the fire protection world would insist that, because nothing prescriptive is documented, the conservative assumption that nothing is restrained should be adopted: just apply more fireproofing or more cover over the concrete reinforcement, and you are right as rain.
While this is a valid approach, it is neither an economical nor necessary one.

How does restraint change with material type?

Consider the basics of the three most common floor and framing construction types today: wood, concrete, and steel. For the most part, if one considers how these elements are designed and utilized, the determination of restrained or unrestrained can be made using little more than an understanding of what restraint means and of the systems themselves and their connections.

Wood: In recent iterations of the building code, exposed wood has been permitted to be designed as a fire-resistant assembly through sacrificial wood or by using a char layer via the expansion of Type IV construction. Type IV is defined as “that type of construction in which the building elements are mass timber or noncombustible materials and have fire-resistance ratings in accordance with Table 601” (of the International Building Code). Further subsets of Type IV construction won’t be discussed in detail here. For further information, refer to Chapter 6 of the International Building Code. However, wood members are always considered unrestrained due to the material itself and its connections. The nature of wood connections, such as light gage metal joist hangers and bearing connections, make rotational and axial restraint difficult to accomplish. Adding to wood members’ unrestrained classification is its propensity to undergo volume change movement as it ages and to be consumed during a fire.

Concrete: Concrete is the original “fireproof” construction material. In days gone by, fire protected steel construction was encased in either concrete or masonry. Thinking back to what is required to be considered restrained and how concrete members are designed, it is easy to see how most concrete construction can be considered restrained. Negative-moment reinforcement, integral reinforced concrete slabs, or continuous topping slabs all serve to help restrain both expansion and rotation of members.

Simple-span, wall-bearing concrete slabs or precast concrete systems will tend to be unrestrained unless the walls themselves are designed to resist thermal movement.

Steel: Most of the debate is, and has been for the better part of 50 years, in the realm of structural steel. Here, fire protection is always provided by applying some type of fire-resistive material. Steel, although not combustible, does not have substantial fire resistance due to its high thermal conductivity and its less-than-ideal temperature-versus-strength/stiffness behavior.
Starting in the 1960s, many research programs, both analytical and empirical, have sought to clarify and improve the understanding of the effect of thermal restraint on the fire resistance of structural steel buildings. A literature review by Gewain and Troup provided an in-depth review of research published on this topic. The resounding consensus offered by numerous researchers and institutions is that standard steel construction, even with the simplest of shear connections, provides adequate restraint of structural elements and thus can be classified as thermally restrained. While these connections do not (and should not) provide the same level of thermal restraint as a UL263 standard test frame setup, that does not matter. If the assembly meets or exceeds the failure criteria, as indicated earlier in the article, then how it gets there is of little consequence.

A caveat to the observation above: open-web steel framing, such as bar joists and joist girders, is typically designed as unrestrained, with some exceptions. Open-web steel framing requires a concrete slab system to provide the necessary end restraint to develop a restrained classification. Otherwise, where open-web steel framing bears on walls or supports bare metal deck, these conditions are considered unrestrained.

If there is a question about whether the structure is restrained, an analytical approach can always be used. A useful description of this approach was investigated by Ioannides and Mehta in Restrained vs. Unrestrained Fire Ratings: A Practical Approach. This method describes analytical procedures for calculating and evaluating thermally induced forces and actual stresses on composite concrete slab-on-metal deck steel structures subjected to thermal loads.

Most guidance on this topic falls back on the age-old adage, “you can never be too careful.” However, in the realm of structural design, that approach may result in structures that are needlessly expensive or overdesigned for no purpose other than to say the designer has done so.

Current and previous versions of the IBC and its predecessor model building codes have taken the stance that it is better to be safe than sorry and to assume structures are always unrestrained unless you can convince the building official otherwise.

While caution is the better part of valor, extensive research supports the position that many structures can and should be designed as restrained. Knowledge is the best tool a design professional has and using it will result in buildings that are more economically designed and as safe as if the designer had taken the path of least resistance. ■

About the Author

Rick Way, PE., SE, is a project engineer at DeStefano & Chamberlain. Way is also the President of the Structural Engineers Association of Connecticut (SEAConn) and serves on the Building Code Advisory Committee. He is a member of the Timber Frame Engineering Council where he serves on the Technical Activities Committee.

References

ASTM International. (2026). Standard test methods for fire tests of building construction and materials (ASTM Standard E119-24). ASTM International.

Underwriters Laboratories. (2011). Standard for fire tests of building construction and materials (ANSI/UL 263-11).

Coalition of American Structural Engineers (CASE). (2008). Structural engineer's guide to fire protection. American Council of Engineering Companies.

Gewain, R. G., & Troup, E. W. (2001). Restrained Fire Resistance Ratings in Structural Steel Buildings. AISC Engineering Journal, 38(2), 78–89. https://doi.org/10.62913/engj.v38i2.754

Ioannides, S. A., and Mehta, S., (1997). Restrained Vs. Unrestrained Fire Ratings: A Practical Approach, Modern Steel Construction, May 1997. 30-37 https://cloud.aisc.org/msc/archive/1997/1997v05.pdf

Ruddy, J., Marlo, J. P., Ioannides, S. A., Alfawakhiri, F. (2003). Fire Resistance of Structural Steel Framing, AISC Steel Design Guide 19, American Institute of Steel Construction, Chicago, IL