Abstract:

Due to the effects of thermal expansion, many members in a fire-exposed steel building frame will experience a combination of axial load (P ) and moment (M ), thus acting as beam-columns. When the frame heats, adjacent members will push on one another and introduce new loading demands. At the same time, the strength and stiffness of the frame's steel material will weaken. Current practice in the US and Europe for the fire-resistant design of steel high-rise structures typically calculates the behavior of these members under the assumption of uniform heating. However, members that are exposed to fire on fewer than all four sides, such as perimeter columns (exposed to fire on three sides) and floor beams (whose the top surface is shielded by the slab), will develop a thermal gradient through their cross-sectional depth. These members will experience a combination of P and M as they encounter restraint to both thermal expansion and thermal bowing in addition to their gravity loads. Thermal gradients may also produce a shift of the sections' effective centroid (i.e. the center of stiffness), resulting in additional bending moments due to eccentrically applied axial load. The plastic capacity of these members to resist combinations of P and M may be altered by the thermal gradients and may be conservative or non-conservative when compared to capacities calculated under the assumption of uniform temperature.

The research presented in this dissertation describes (for the first time) the fundamental behavior of steel beam-columns with thermal gradients, with emphasis on the case of the perimeter column. The study focuses on the performance of wide-flanged member types that are used as perimeter columns in typical North American steel frame construction. Both computational and experimental research results are used to (1) demonstrate the unique changes in capacity and demand experienced by beam-columns that develop thru-depth thermal gradients when exposed to fire, (2) examine the parameters that affect the use of computational models to predict the behavior of beam-columns in a high-rise steel moment frame exposed to fire, and (3) develop performance-based tools and methodologies to predict the demand and capacity of these members.