Numerical Simulation of Walls and Seismic Design Recommendations for Walled Buildings

Abstract

Structural walls are a commonly used lateral force-resisting system in low, medium and high-rise building construction in seismically active regions throughout the world. The stiffness and strength provided by structural walls are ideal for resisting service level seismic events; additionally, walls that are capacity designed to suppress shear failure are expected to perform well in design level seismic events by dissipating energy through ductile inelastic flexural response. To achieve this objective, seismic design procedures for walls are required that 1) adequately estimate shear demands expected to develop in walls yielding in flexure and 2) limit inelastic flexural demands to levels which ensure an acceptable collapse risk can be achieved. Validation of such procedures for walls can be aided by the use of numerical tools such as finite element analysis; however such tools must be capable of accurately simulating inelastic wall response, including deteriorating flexural response and loss of lateral load-carrying capacity. The research presented discusses development of a recommended modeling approach for simulating inelastic wall response and the use of the proposed simulation method to evaluate existing US seismic design procedures for slender walls. The recommended modeling approach utilizes material energy regularization to allow for mesh-independent prediction of post-peak behavior and accurate simulation of wall ductility (in terms of system drift) using fiber-type distributed-plasticity line elements. Material energy recommendations for both confined and unconfined wall regions were developed using an experimental dataset of 21 cyclically loaded wall test specimens. Existing US seismic design procedures were evaluated in terms of collapse risk and assessment was performed using inelastic time history analysis (ITHA) and incremental dynamic analysis (IDA) in accordance with the Federal Emergency Management Agency's (FEMA) P695 methodology. Preliminary assessment using P695 identified that current US design procedures, which do not include capacity design provisions for wall shear, significantly underestimate wall shear demands when compared to shear demands predicted for building models subjected to design level ground motion records and that walls designed using current US procedures may be unable to develop the assumed flexural mechanism. To address this problem, a modified modal response spectrum analysis, which is verified by P695 analysis to adequately suppress shear failure and allow fully development of the desired flexural mechanism, was developed. Results of P695 analyses are used to establish recommended flexural force reduction factors and design envelopes for slender walls.