科学研究

国际与国内交流

XIAO Heng: Physics informed machine learning for turbulence modeling

时间:2019/08/16浏览:13

报告时间:2019818日(星期日)14:10

  

报告地点:校学术会议中心

  

人:XIAO Heng 助理教授

  

工作单位:弗吉尼亚理工大学

  

举办单位:土木与水利工程学院

  

报告人简介:

  

肖恒于2003年毕业于浙江大学土木系,2005年获瑞典皇家理工学院科学计算方向硕士,2009年获普林斯顿大学博士,2009年至2012年在苏黎世联邦理工学院流体力学研究所从事博士后研究,2013年加入弗吉尼亚理工大学任助理教授。目前的主要研究方向是结合传统物理模型与现代数据科学的建模方法和数据驱动的湍流模拟。其他研究领域包括离散元方法,泥沙输运,海洋可再生能源等。

  

报告简介:

  

Turbulence is among the last unsolved problems in classical physics, and it impacts many issues of societal importance including energy, environment, and climate. Accurate predictions of turbulent flows are of vital importance for the design and operation of mission critical systems such as aircraft, gas turbine engines, and nuclear power plants. Currently, RANS simulations are still the workhorse simulation tool for industrial turbulent flows, as direct numerical simulations and large eddy simulations (LES) are still too expensive computationally. RANS simulations rely on turbulence models to represent the unresolved physics. These models introduce large uncertainties into the results, severely impairing their predictive capabilities. Such difficulties have been highlighted in recent reviews.




友情链接/ Links
合肥工业大学土木与水利工程学院@版权所有