Experimental Evaluation of Theoretical Impact Models for Seismic Pounding

Abstract

ABSTRACT One of the identified consequences of the aftermath of strong quakes is that structural pounding may trigger substantial damage including the collapse of colliding structures. This situation is attributed to inadequate distance between the adjacent structures. In order to address this concern, various numerical and experimental studies were conducted simulating earthquake induced pounding using different models of impact force. In this study, shaking table tests were conducted to assess the impact of seismic pounding between steel frame structures under near-field and far-field earthquakes. Five theoretical pounding force models, namely: linear spring, linear viscoelastic, Hertz, non-linear viscoelastic and Hertzdamp models were employed to replicate the pounding force. Various components of the theoretical models such as contact element stiffness, coefficient of restitution and impact damping ratio were calculated using the experimental results. Comparing the results of the Experimental Impact force time history with the prediction outcomes showed that the models over-predicted the pounding response. However, the prediction outcomes were congruent with the experimental results. Lastly, the linear viscoelastic model generated the least variance and most accurate predicted results compared to the other models. Hence, it can be inferred from the data that among other models, linear viscoelastic is the most accurate model to be used for the testing of pounding response.