The behavior of steel perimeter columns in a high-rise building under fire

This study investigated the thermal and structural effects of the slab and fire protection material on the perimeter columns of a steel high-rise building exposed to fire. The beam sizes, column sizes, and applied fire remained constant in the prototype frames, but the presence/absence of fire protection material and the method of representing the slab varied. Our study indicated that in 2-D analyses it is reasonable to represent the slab in the thermal and structural model by not applying fire load to the top surface of the steel beam flange (the slab is therefore assumed to prevent heat transfer to the beam below). In the analyses with unprotected beams, the beams achieved higher temperatures, which lead to greater thermal expansions and therefore larger column lateral deformations than the cases with the beams protected. With no fire protection material on either the beams or columns (or both), the frame became unstable within an hour of the fire's ignition. If fire protection material is applied to both the beams and columns (as specified by the building's construction documents), the structure survives the full duration of the fire. The fire-induced structural response of the beams that frame into the perimeter column directly affects the perimeter column behavior and structural integrity of the frame in the following ways: (1) As these beams are heated, they expand and force the column to deform laterally, which induces large column moments that combine with large axial gravity forces to create a plastic hinge. This behavior does not necessarily result in structural instability since a collapse mechanism (the formation of several plastic hinges in the columns leading to instability) may not form. (2) If the beams do not have fire protection material, they reach their capacity due to bending and the fire-induced axial forces that develop in them. Once these beams, which brace the perimeter column, fail, the stability of the column, and structure as a whole, may be compromised and could potentially lead to structural collapse. Both this study and full-scale tests of steel frames have shown that the strength of the columns is vital, and therefore the appropriate steps should be taken to adequately protect them against fire exposure. However, especially in the case of the perimeter column, one must also carefully consider the design and fire protection of the members that brace them (the beams). It is therefore recommended that the fire protection material requirements for the beams bracing the perimeter columns be larger than that of the other beams in the building since they significantly affect the perimeter column behavior and have the potential to compromise overall frame stability. The results presented in this paper therefore support the fire protection methodology of current IBC Type IA construction (ICC, 2000), which requires beams that frame into columns to have the same level of fire protection (3 hours) as the columns to which they connect. Older construction, which required less fire protection to these critical beams than the columns (as was the case for our prototype), should be considered for upgrading to meet current codes. More research is needed to determine recommended levels of fire protection material for these beams. While fire resistive material is important in steel structures, the One Meridian Plaza event (our prototype for this study) also demonstrates the importance of automatic sprinkler systems. The floors in our prototype that were engulfed by the fire did not have operational sprinklers, which can control and extinguish fires before they reach the dangerous post-flashover stage. When a beam or column limit state was reached in the analyses, it was reached by the combined action of axial load and moment. The beams and columns examined in this study act as beam-columns (members that are subject to both axial load and moment), and therefore their capacity and behavior should be examined with this perspective.