NOUS · Probabilistic Inverse Models for Assessing the Predictive Accuracy of Inelastic Seismic Numerical Analyses

This research proposal deals with developing probabilistic inverse models for assessing the predictive accuracy of inelastic seismic numerical analyses. Numerical models for predicting the inelastic response of structures in seismic loading are biased by so-called epistemic uncertainties that arise from our lack of knowledge: modelling errors, poor comprehension of material constitutive behaviours and of the energy dissipation mechanisms, and so on. Bringing together experts in probabilistic computational mechanics, earthquake engineering, nonlinear material science and mathematical statistics, the research project aims at providing innovative useful numerical tools to researchers, designers and analysts for decision making regarding the seismic risk and structural safety of designing and existing structures. More specifically, research will be oriented toward developing deterministic-probabilistic inverse models to quantify the epistemic uncertainties in inelastic seismic numerical analyses. Such models will allow for computing confidence regions for the quantities of interest. Confidence regions reflect the predictive accuracy of the simulations and provide useful information for defining new research orientations as well as for structural safety and risk assessment.