Column-base connections in steel moment-resisting frames (SMFs) in seismic regions are commonly designed to develop the capacity of adjoining column with an intent to develop a plastic hinge in the column member, rather than in the connection (i.e., a strong-base design). Recent research has shown base connections to possess high ductility, indicating that this practice may be not only expensive but also unnecessary. This suggests that concentrating inelastic rotations in the base connection may result in acceptable performance. Motivated by
this finding, the performance of steel moment resisting frames with a weak-base design is investigated to examine the relationships between base-connection strength, deformation capacity, and structural performance. The main scientific basis of this study is nonlinear static pushover and nonlinear time history simulations on SMFs designed for high seismicity.
These simulations interrogate parameters including frame height (4-, 8-, 12-, and 20-story), base-connection strength, base flexibility, and base rotation capacity, resulting in a total of 160 parametric combinations. The performance of each of these is assessed to estimate the probabilities of failure or collapse corresponding to a 2/50 (2% probability of exceedance in 50 years) intensity of shaking. The key finding is that for all frames, acceptable performance (i.e., comparable to current practice with capacity-designed bases) may be obtained by designing the base connections for overstrength seismic loads (i.e., based on the Ω factor) rather than based on column capacity; this has the potential to offer significant cost savings. For a low-rise (i.e., four-story) frame, acceptable performance is achieved even without significant rotational demands in the base connection. For the other frames, a baseconnection rotation capacity of approximately 0.05 rad is necessary to achieve acceptable performance. Implications of these findings for prospective design practices, as well as future research, are outlined, and limitations are discussed. D