Computational Fluid Dynamics Laboratory
School of Mechanical and Aerospace Engineering, Oklahoma State University

We are developing advanced computational methods for nonlinear dynamical systems driven by multiscale and multiphysics processes. The systems studied in our lab range from engineering flow problems to geophysical flow problems with different levels of complexities. There is an emphasis on modeling and analysis of turbulent flows across a variety of spatial and temporal scales. Our current efforts are centered in the development of hybrid modeling & analytics approaches in LES and ROM frameworks. Integrating  and  aligning  different  ways  of  physics-based and machine learning/data-driven modeling approaches, hybrid modeling framework embraces flexibility to use most effective tools depending on the problem and its challenges. Currently, we are working on both functional or structural closure modeling ideas to take into account subgrid-scale effects. Closure modeling refers to the process of including the truncated scales  (due to the limited numerical resolution & computational resources)  into the resolved dynamics (the scales captured in our underlying numerical model) to account for the missing physics, which is quite important if the underlying dynamical process is nonlinear with strong interactions between small and large scales. More information about our research activities can be found here.  We are located at the third floor of Advanced Technology Research Center building on OSU's Stillwater campus (ATRC 303).
Closure modeling analogy between large eddy simulation (LES) and reduced 
order modeling (ROM) where higher m index refers to smaller scales (figures adapted from 

A snapshot of the density field for the stratified turbulence problem where we assess the validity of the proposed blind deconvolution approach. Maulik and San, J. Fluid Mech. 2017; 831: 151-181

Evolution of the x-component vorticity iso-surfaces for the Taylor-Green problem, exhibiting well-known vortex stretching mechanism in isotropic, homogeneous turbulent flows and the consequent production of small eddies. San, Staples, and Iliescu, Int J Comput Fluid D 2015; 29: 40-66 

A high resolution snapshot of the double shear layer problem obtained by the Roe solver, where we test the implicit and explicit LES models. Maulik and San, Fluids 2017; 2:14

Rayleigh-Benard natural convection: Streamfunction (top), vorticity (middle) and temperature (bottom) fields for Ra= 10,000 (left) and Ra=100,000 (right). Maulik and San, Int. J. Heat Mass Transfer, 2017; 108: 1656-1675

Evolution of temperature field for the Marsigli flow problem (also known as the lock-exchange problem), a strong shear flow problem exhibiting the Kelvin-Helmholtz instability. San and Borggaard, Int J Numer Meth Fluids, 2015;  78: 37-62

Instantaneous Q surface plot for a turbulent channel at Re_t=395, colored by the magnitude of the absolute vorticity. Presented in the Coalition for Advancing Digital Research & Education (CADRE) conference, Oklahoma State University, Stillwater, April 17, 2018.