No abstract

The influence of welded column splice fracture on the seismic response of steel moment frames is examined. The study is motivated by pre-Northridge moment frames with welded column splices with crack-like flaws that are highly vulnerable to fracture. Costly retrofit strategies to repair these splices are usually intentioned to preclude splice fracture, without an explicit examination of its effects on global response. This study simulates post-fracture response of splices through a new material model, which is informed by fracture-mechanics based estimates of splice strength, and reproduces phenomena such as gapping and re-seating that occurs in the splices after fracture. Nonlinear response history simulations (incorporating this model) are used to examine the response of 4-and 20-story moment frames. The simulations, using 100 ground motions, and reflecting key aspects of nonlinear response are conducted within a Performance Based Earthquake Engineering (PBEE) framework, to examine global and local structural response in a probabilistic sense. The simulations indicate that neither the collapse potential nor building deformations are significantly affected by splice fracture when compared to benchmark simulations without fracture. This is attributed to a combination of phenomena; these include the mobilization of building rocking due to splice fracture, and the tendency of fractures to cascade upwards through individual columns rather than across a story. The results suggest that splice fracture may not necessarily trigger structural collapse, and retrofit strategies that consider global, rather than local response may be more cost effective. Limitations of the study are outlined.

No abstract