Effect of modeling assumptions on predicting seismic responses of a three‐story reinforced concrete shear wall structure

The behavior of short‐period reinforced concrete (RC) shear wall structures is often complicated and hard to predict accurately, even when the structure behaves in the elastic region, due to significant uncertainties in the material and the environment. Modeling assumptions used in finite element (FE) analyses often influence the accuracy of the dynamic response predictions. This paper discusses the numerical modeling of shaking table tests of a 3‐story RC shear wall specimen, which was carried out by Korea Atomic Energy Research Institute in July 2020. The experimental program is briefly introduced in this paper. Through nonlinear time history analyses using ABAQUS, the effect of modeling assumptions on the accuracy of the FE methods in predicting the linear and moderately nonlinear behavior of the RC structure is presented. Two commonly used modeling/model updating assumptions are considered: concrete Young's modulus and foundation flexibility. Influences of such modeling assumptions in predicting beyond design dynamic behavior (i.e., nonlinear responses and damage development) of the testing structure are also studied. The results showed that the accuracy of the dynamic response prediction of the structure could be improved significantly after calibrating the models against the white noise test results. Nevertheless, models established with different modeling assumptions can only capture the behavior of the structure at certain seismic intensity levels. Different models give results with a considerable variation in the structure's peak acceleration, floor response spectrum, acceleration amplification profile, damage pattern, and damage severity.