Hysteretic Models for Reinforced Concrete Shear Walls in Nuclear Facilities Using NSGA-II and Data-Driven Techniques

Abstract

The current thesis provides a thorough exploration of the seismic behaviour of reinforced concrete (RC) shear walls, with a particular focus on the performance characteristics of squat RC shear walls, which are pivotal for the seismic resilience of safety-related structures in nuclear facilities. The thesis is rooted in the application of the Bouc-Wen-Baber-Noori (BWBN) model, an advanced hysteretic model that captures the complex nonlinear response of materials under cyclic loading. The primary objective is to simplify the predictive aspect of the hysteretic response of squat RC shear walls through a multifaceted framework that integrates the BWBN model with data-driven techniques. Specifically, the thesis adopts the non-dominated sorting genetic algorithm (NSGA-II) for the optimization of the BWBN model parameters through a dataset of 100 squat RC shear wall specimens that were collected from previous relevant experimental programs. The thesis then utilizes the BWBN model results through genetic programming to develop equations for the different model parameters. The developed framework is expected to provide a practical tool for engineers and practitioners, simplifying the incorporation of complex hysteretic behaviours in the seismic design and assessment of squat RC shear walls. The extended analysis and findings presented in the current thesis underscore the critical importance of adopting sophisticated computational techniques in the field of earthquake engineering. By advancing our understanding of the seismic behaviour of RC shear walls and improving the tools available for their analysis, this research contributes significantly to the ongoing efforts to enhance the resilience of nuclear facilities in the face of extreme seismic events.