Onkar Sahni, Ph.D.

Assistant Professor, Rensselaer Polytechnic Institute

Massively Parallel CFD for Complex Flows Using Anisotropic Adaptive Grids

Despite the tremendous growth in the capability of computer hardware, the range of length and time scales present in most flows of engineering interest will continue to challenge the field of computational fluid dynamics for the foreseeable future. Consideration of complex geometry and or complex flows encounters numerous additional challenges: 1) the presence of solid walls introduces boundary layers requiring anisotropic grids, 2) add geometric complexity in the way of curvature and or corners and flow separations that make {\it a priori} grid definition difficult if even possible, strongly motivating the use of anisotropic adaptivity to allow the solution to discover the resolution needs, 3) the existence of time scales that may be present in the physical model but of little interest motivate further consideration of implicit time integration to substantially reduce the number of iterations to a steady state or time steps to capture the transient of interest.

A petascale solver and supporting adaptivity framework has been and continues to be developed to address these issues for arbitrarily complex geometries. The solver portion has already been shown to strongly scale to 288k processors and over 10 billion element adapted grids. Detailed simulation results will be presented in the areas of turbulence, multi-phase flows, and cardiovascular flows


Dr. Onkar Sahni is currently an Assistant Professor in the Department of Mechanical, Aerospace and Nuclear Engineering (MANE) at Rensselaer Polytechnic Institute (RPI). He is also affiliated with the Scientific Computation Research Center (SCOREC), Center for Flow Physics and Control (CeFPaC), and Center for Automation Technologies and Systems (CATS) located at RPI. Prior to his current position he was a research scientist/engineer at the Center for Predictive Engineering and Computational Science (PECOS) located in the Institute for Computational Engineering and Sciences (ICES) at the University of Texas (UT) at Austin. He received his PhD degree from Rensselaer in 2007 and Bachelor's degree from Indian Institute of Technology-Bombay (IIT-B) in 2002.

His research interests span modeling and simulation of coupled fluid flow problems. His research is focused on the development and application of simulation-based predictive tools for fluid flow problems involving turbulence and flow control. His research work also includes adaptive methods and extreme-scale computing for fluid flows in various real-world application areas, for example, aerodynamics, cardiovascular flows, remelting processes, two-phase flows, wind turbines.

Workshop Program
updated: 2011-10-19