Parameters for Modeling a High-Rise Steel Building Frame Subject to Fire

This paper examines the level of detail and complexity that one needs to incorporate in a computational finite element (FE) model to predict the thermal and structural response of steel high-rise building frames to fire. Comparisons are made between these models in terms of accuracy and efficiency. Performance related to three parameters was examined: (1) the representation of the structural system as a 3-D full frame model versus a 2-D plane-frame model, both of which include the steel frame and the floor slab; (2) the representation of the slab in the 2-D plane frame model; and (3) the effects of modeling the temperature profile of each steel member cross-section as non-uniform (i.e. allowing a thermal gradient to develop) versus uniform. Results indicate that the 2-D plane frame model can be reasonably used in some cases to predict the performance of the perimeter column and floor beams framing into them in a fire-exposed high-rise moment-resisting frame (MRF) with a significant savings in analysis run time. The slab has little influence on the structural analysis of a 2-D plane frame; however, the slab influences the thermal profile through the depth of the beams, and these temperature changes will produce a non-negligible change when calculating the behavior of the frame and should be accounted for. Results also indicate that models whose members have uniform temperature can be used to obtain reasonable estimates of the interaction between connected beams and columns. However, thermal gradients produce significant changes in the deflection mechanics and plastic P-M limit state behavior exhibited by non-uniformly heated beam-columns that experience a severe decrease in capacity; therefore, it is recommended that thermal gradients be included in models that are used to predict deflections or plastic limit state behavior.