Bright Spark Lectures
| Bright Spark Lecture 1 | Values of Monte Carlo Samples for Geotechnical Reliability-Based Design |
| Date / Time | 13 December 2019, Friday / 16:30 - 18:00 hrs |
| Venue | Room IB-101 |
| Speaker | Zi-Jun Cao Wuhan University, China |
Biography
Dr. Zi-Jun Cao, is currently Professor in School of Water Resources and Hydropower Engineering at Wuhan University, China. He earned his PhD from City University of Hong Kong in 2012 and was a visiting scholar at University of California, Berkeley, USA. His research interests include uncertainty quantification in geotechnical engineering, simulation-based geotechnical reliability analysis and design, and probabilistic observational method. He is the member of ISSMGE TC304 (Engineering Practice of Risk Assessment and Management) and TC309(Machine Learning and Big Data), Geotechnical Safety Network, and Engineering Risk and Insurance Research Council of China Civil Engineering Society. He is presently serving on Assistant Editor and Editorial Board Member of Geo risk (Taylor & Francis). He is a recipient of the Outstanding Young Researcher Fellowship, CAST (2017), Highly Cited Research Award, Engineering Geology, Elsevier (2017), the First Prize of Natural Science of Hubei Province, China (2017), Luo-Jia Young Scholar Award, Wuhan University, (2015), and Chu-Tian Scholar Award (2014).
Abstract
With the advent of computing technologies, Monte Carlo simulation (MCS) has been gaining wide applications in geotechnical engineering. Various MCS techniques (e.g., direct MCS, Subset Simulation (SS), and Generalized SS (GSS)) have been successfully applied to developing full probabilistic design methodologies for geotechnical reliability-based design (RBD). Nevertheless, geotechnical practitioners are reluctant to adopt full probabilistic design methodologies due to conceptual and mathematical hurdles of reliability algorithms. This paper presents recent attempts aiming at improving the practical applicability of MCS-based full probabilistic RBD approaches by decoupling reliability algorithms from deterministic design calculations and implementing MCS-based reliability calculations in an EXCEL spreadsheet
environment. Besides reliability calculations, post-processing MCS samples allows achieving efficient RBD updating under different design scenarios, providing reliability sensitivity of uncertain parameters to identify the most significant uncertainty sources, and determining design points of uncertain parameters from their failure samples for bridging full and semiprobabilistic RBD methodologies. More attentions shall be paid to these values of MCS samples in future to take advantage
of fast development of high performance computing.
Full Text (PDF)