报告人：Prof. Andrew Ooi（The University of Melbourne）

报告内容：

Numerical methodologies based on Computational Fluid Dynamics (CFD) are now commonly used to predict the flow field in many engineering applications. However, CFD data from Direct Numerical Simulation (DNS), Large Eddy Simulation (LES), Reynolds Averaged Navier-Stokes (RANS) are expensive to compute, even with modern day supercomputers. In using any of the CFD methods, the users need to provide information about the flow configuration of interest such as geometry of the object, upstream boundary conditions, fluid properties etc. In real engineering applications, often accurate information for all these flow quantities are difficult to obtain. For example, the geometry of a bluff body is usually not known exactly due to variations in the manufacturing process, there are usually uncertainties about the upstream boundary conditions especially if the flow field is turbulent, and in many instances the velocity of the object is not known in advance. Such uncertainties cause difficulties in interpreting the data from CFD and could cause doubts in the reliability of the results obtained from numerical simulations. In order to understand the effects of these uncertainties, one can use a brute force approach and just conduct the CFD simulations for all values of the uncertainties. This approach is usually not very efficient and maybe even not feasible due to the resource requirements of CFD simulations. In recent times, more efficient methods such as Polynomial Chaos Expansion (PCE) and bifidelity approaches have been used to quantify uncertainties, in an attempt to understand the variation and reliability of the simulation results. In this seminar, results from applying these methods to quantify the uncertainties for the flow around two cylinders, the flow past a cylinder confined in a duct and gravity current flows will be presented. Advantages and disadvantages of these different approaches will be discussed.  

报告人简介：

Professor Andrew Ooi is an academic staff member in the Department of Mechanical Engineering at The University of Melbourne and Associate Dean (Academic) in the Melbourne School of Engineering. Prof Ooi graduated with a BEng in 1993 and PhD in 1997 from the University of Melbourne. Prior to his current appointment, Andrew worked at the Center for Turbulence Research (CTR) at NASA Ames, Stanford University and as a research scientist at the Defence Science and Technology Organisation (DSTO). His current research interests include numerical simulation and the application of artificial intelligence methodologies for fluid flow applications such as bluff body natural convection and multiphase flows.

主持人：赵耀民 研究员（北京大学应用物理与技术研究中心）

时  间：2025年10月16日（周四）12:20

地  点：北京大学工学院1号楼210会议室