A Framework for the Integration of Models in Structural Design and Nonlinear Analysis for Performance-based Seismic Design of Buildings

Abstract

The increasing application of Performance-Based Seismic Design to design and retrofit buildings has been possible due to ongoing research, growing computing power, and development of numerical models to simulate cyclic structural behaviour. Performance assessment through nonlinear structural analysis can be cumbersome due to the time and expertise needed to develop a nonlinear model, running analyses, and the subsequent data post-processing, which can prevent a broader adoption of the approach by engineers. Furthermore, the volume of information on nonlinear modelling of structural components is vast and widespread. This thesis explores the formulation of a computational framework for semi-automatic conversion of an elastic structural model, developed with a commercial platform, into a nonlinear analysis model. The conversion method allows the user to focus on choosing the most suitable model for the structural components involved, from a knowledge base built of component models, while avoiding cumbersome work when defining the model.