Dynamic Shake-Table Testing and Analytical Investigation of Self-Centering Steel Plate Shear Walls

Abstract

Recent research has shown that self-centering steel plate shear walls (SC-SPSWs) offer an enhanced seismic performance over conventional steel plate shear walls by providing an additional self-centering capability using steel frames detailed with posttensioned (PT) beam-to-column rocking connections. As with other previously proposed self-centering frames detailed with similar beam-to-column connections, this detailing facilitates the incorporation of replaceable energy dissipation components, as part of the lateral force resisting system (LFRS). In doing so, by design, the gravity frame components of the LFRS are also protected from damage during an earthquake. To investigate the dynamic seismic response of this proposed structural system, one-third scaled SC-SPSW specimens were subjected to ground motions during a series of dynamic shake-table tests. The experimental investigation results presented in this paper are the first shake-table tests conducted on SC-SPSWs. This test program was composed of two three-story single-bay SC-SPSW frames, each with a different PT beam-to-column connection. For one frame type, connections rock about both beam flanges; for the other, connections rock about the top beam flanges only (referred to as the NewZ-BREAKSS connection). The latter connection essentially eliminates PT boundary frame expansion (a.k.a., beam-growth) that occurs with connections that rock about both beam flanges. Furthermore, both an infill web plate and a bidirectional infill web strip layout were investigated as alternative infill configurations. Results show that the presence of infill web plate compression strength has no significant effect on recentering of the frame (contrary to what has been reported in the literature for quasi-static tests). Furthermore, presented analytical expressions that describe the drift induced infill web plate strains and posttension demands for beam-to-column rocking joints are found to compare well with the experimental results, but conservatively overestimate the PT demands. Numerical models were able to reasonably estimate the peak roof drift and maximum base shear demands. This paper presents information on the actual seismic response of SC-SPSWs detailed with a flange-rocking and the NewZ-BREAKSS posttensioned rocking joint connections, along with analytical equations that could be used to inform some aspects of design.