Effectiveness of physics-based collapse criteria in seismic risk assessment of code-conforming low-rise RC framed structures

Abstract

The evaluation of structural collapse capacity is an integral part of the estimation of collapse risk in performance based earthquake engineering. Several methods and procedures have been proposed in the past to quantify collapse capacity. However, there is no clear consensus on which method is most appropriate. Recently physics-based collapse criteria have been proposed, but their effectiveness is yet to be compared with codes of practice. In order to ensure safety and consistency, different code/standard based recommendations enforce the use of thresholds predominantly in terms of the engineering demand parameter (EDP) for the performance level of collapse prevention. These thresholds serve as limit-state criteria to evaluate the collapse capacity of a structure. This paper compares several performance measures that are derived using different criteria to understand their impact on the final collapse risk estimates. Four different criteria are studied, two of which are based on standards, and the other two are physics-based, which use energy formulations. Three different ASCE 7-16 code conforming RC buildings are designed and analysed. The effects of modelling uncertainties and of the ground motion spectral shape have been considered to impart more confidence in the results. The collapse risk estimates are interpreted in terms of different collapse performance measures. The estimates derived from all the four criteria are compared. It was found that the collapse risk is significantly affected by the choice of the criterion. For the same collapse risk, the physics-based criteria allow higher ultimate deformation at collapse. On the other hand, the code/standard based criteria tend to be conservative as they censor the deformation response of the structure using upper bound thresholds. It was also found that the physics-based criteria could underestimate the collapse risk, yet they can still be employed to produce more economical designs.