Development and experimental validation of a novel nonlinear layered shell element for RC structural walls

Increasingly widespread use of nonlinear response history analysis methods for performance-based design/assessment of buildings with complex reinforced concrete (RC) wall systems has underlined the need for development and implementation of more robust nonlinear modeling approaches that is capable of accurately estimating the nonlinear response of structural walls at both global and local response levels, while also reflecting important (yet commonly-ignored) nonlinear response characteristics, such as shear-flexure-torsion interaction, plane sections not remaining plane, warping, out-of-plane behavior, nonlinear shear deformations, and shear stress migration across wall cross-section. Furthermore, such modeling methodologies for walls need to be implemented in widely-accessible computational platforms that are capable of efficiently simulating the dynamic response of entire three-dimensional structural systems under multi-directional seismic excitations. With this motivation, a novel nonlinear layered shell element, referred to as the Layered Fixed-Strut-Angle Finite Element (LFSAFE) model, was developed and implemented in OpenSees. In this study, the constitutive behavior and working principles of the LFSAFE model element are presented, the effectiveness and accuracy of the model is experimentally-validated, via simulation of test results for non-planar wall specimens subjected to quasi-static multi-directional lateral and torsional loading conditions. Additionally, nonlinear dynamic response simulation capabilities of the model are also evaluated, via comparison of model estimations with shake table tests results of a U-shaped wall specimen and a ten-story wall-frame building specimen. Comprehensive response comparisons provided in this study demonstrate that the LFSAFE model can accurately simulate salient response attributes of non-planar walls, at both global and local response levels, under different loading conditions, while also efficiently replicating experimentally measured dynamic response of shake table tests, with reasonable accuracy.