Seismic Response of an Instrumented 13-Story Steel Frame Building Damaged in the 1994 Northridge Earthquake

Abstract

This paper summarizes a case study of a 13-story welded steel moment frame (WSMF) building affected by the 1994 Northridge earthquake. The building, which was instrumented, sustained extensive damage to its welded connections. Ground motion records from the basement and response records from the sixth and twelfth floors were available. Damage data was collected with post-earthquake inspection and testing of each joint. The primary objective of the study was to compare modeled behavior with recorded response in order to assess the value of present analytical tools and modeling techniques for predicting the distribution and severity of connection failures. Calculated elastic time-history displacements matched well with recorded displacements in the E-W direction, less so in the heavily-damaged N-S direction where the elastic model was unable to simulate fractured moment connections. In the elastic analyses, joint demand was represented by beam demand-capacity ratios (DCRs). The highest beam DCRs were concentrated between the second and seventh floors; these locations correlated strongly with observed damage. Inelastic time-history analyses improved the displacement match in the N-S direction. They also indicated that panel zone yielding would have controlled the intended ductile response. This study suggests that for a regular structure, current modeling and analysis tools for both elastic and inelastic analysis, while unable to simulate premature brittle fractures, can be useful for predicting in a probabilistic way the intensity and distribution of damage expected in moderate seismic events.