In many simulations in fluid and solid mechanics but also in molecular simulations there are many sources of uncertainty, e.g., associated with boundary conditions, material properties or equations of state and constitutive laws. These uncertainties may contribute to large errors in the simulation, typically much larger than the spatio-temporal discretization errors, leading to erroneous dynamics or performance predictions. In this talk, we will present several such examples from fluid mechanics but also across various disciplines and subsequently we will discuss different probabilistic and deterministic numerical approaches to quantify the effect of such uncertainties. Such stochastic simulations also serve for a better validation with the experiments but also through sensitivity analysis they can steer the experimental effort to a more effective sets of measurements.
报告人简介：George Karniadakis received his S.M. (1984) and Ph.D. (1987) from Massachusetts Institute of Technology and did his postdoc at the Center for Turbulence Research at Stanford /Nasa Ames. He is a Fellow of the Society for Industrial and Applied Mathematics (SIAM, 2010-), Fellow of the American Physical Society (APS, 2004-), Fellow of the American Society of Mechanical Engineers (ASME, 2003-) and Associate Fellow of the American Institute of Aeronautics and Astronautics (AIAA, 2006-). He received the CFD award (2007) and the J Tinsley Oden Medal (2013) by the US Association in Computational Mechanics. Karniadakis is the lead PI of an OSD/AFOSR MURI on Uncertainty Quantification and Director of a new DOE Center of Mathematics for Mesoscale Modeling of Materials (CM4).
