ESRA Webinar: Prof. André Teófilo Beck, University of São Paulo
Robust, Reliability-Based and Risk-Based Optimization under Aleatory and Epistemic Uncertainties: Concepts and Applications to Progressive Collapse of Structures
Date: February 5th, 2024, 01:00 PM (Central European Time).
Speaker Bio: Prof. André Teófilo Beck graduated in Mechanical Engineering from Universidade Federal do Rio Grande do Sul (Brazil, 1996), obtained a master’s degree in Mechanical Engineering from Universidade Federal de Santa Catarina (Brazil, 1998). He obtained a Doctoral degree (Ph.D.) in Civil Engineering from University of Newcastle (Australia, 2003), and works as Associate Professor at the Department of Structural Engineering, São Carlos School of Engineering, University of São Paulo. He is further a member of the Brazilian Association of Computational Methods in Engineering (ABMEC) and of the Uncertainty Quantification and Stochastic Mechanics Committee of the Brazilian Association of Mechanical Sciences (ABCM) and serves as a peer-reviewer for the main Brazilian Research agencies FAPESP, CAPES and CNPq. His research interests include Structural Mechanics and Structural Safety, in particular: structural reliability, stochastic processes, structural optimization. Finally, he is a member of the Editorial Boards of the Journal of the Brazilian Society of Mechanical Sciences and Engineering, ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Reliability Engineering & System Safety and Structural Safety
Abstract: It is nowadays well accepted that engineering systems need to be designed considering the inherent uncertainties in demands and capacities. Robust, reliability-based and risk-based optimization are three modern approaches to address the optimal design problem under objective aleatory uncertainty, that is: uncertainty which can be quantified using probability theory. Yet, subjective epistemic uncertainties, related to limitations of knowledge, social, political, financial, organizational, and behavioral factors, human errors, etc., cannot be quantified probabilistically, but also need to be considered in design optimization. In this talk, the "Latent Failure Probability" concept is proposed to address epistemic uncertainties in optimal design of structural systems. An optimal redundancy allocation problem in structural engineering is addressed. Application to progressive collapse of frame structures is also addressed.