! Welcome to my homepage !

I am a computational flow modeling engineer with broad expertise in high-performance computing and multiphase flows. Currently, I am a senior research scientist at the Center of Innovation for Flow through Porous media. I was a postdoctoral research associate at the University of Wyoming and worked with Dr. M Stoellinger to implement thermal radiation models in a legacy multi-phase flow solver. I contributed to multiphysics and multiscale laboratories of Professor V Kumar at the University of Texas at El Paso. I was a guest researcher at Sandia National Laboratories and worked with Dr. W Spotz to develop an interface to integrate MFiX with Trilinos. I have worked with several other researchers at Sandia National Laboratories to incorporate tools such as Trilinos, and Dakota with other opensource software. I did my graduate work in the computational fluid dynamics group of Professor S Mittal at the Indian Institute of Technology. To date, my research results have been published in high impact journals, including Physics of fluids, Fluid Engineering, Propulsion and Power, Shock Waves, Powder Technology, Metallurgical and Materials Transactions, Numerical Methods in Fluids, Super-computing, and others.

As a postdoc at UWyo, I designed and implemented advanced spectral models for thermal radiation in MFiX. The fossil fuel community widely uses MFiX suite for simulating flow in fluidized beds. For verification, the results with those models are compared with that from OpenFoam and ANSYS Fluent. The results with those models for chemically reacting flows such as gasification and combustion of coal/biomass are validated. The chemical thermo-kinetics obtained from Cantera and C3M are incorporated in MFiX simulations.

My postdoctoral training at UTEP allowed me to develop an interface to integrate the advanced linear solvers in Trilinos with MFiX. Trilinos provides a framework for simulating large-scale, sophisticated multi-physics engineering and scientific problems. The interface is written in Fortran and C/C++. The interface has been verified and validated on various fluid bed problems. Two fluid, discrete element and particle in cell approaches are used to simulate the flow in fluidized beds. I also tested the performance of the linear solvers, which are based on the Kokkos programming model, on various computer architectures. The iterative solvers in the integrated multiphase flow solver are relatively fast compared to the built-in solvers in MFiX. Further, I designed and developed a framework to integrate MFiX with Dakota for uncertainty quantification, sensitivity analysis, and optimization of multi-physics problems. However, I also developed an exascale capable pore-network simulator and combined it with Dakota.

I have used machine learning algorithms in various python libraries to predict pressure difference/flow rate in Hagen-poiseuille flow, friction factor, metrics for characterizing laser propagation in atmospheric turbulence, and depth of penetration of molten salt into a pore network. This work has been presented at an ASME conference. I also have experience in using deep learning and convolutional neural networks in TensorFlow.

During my doctoral studies, I developed a stabilized finite-element method with higher-order interpolation functions to solve turbulent flows via Reynolds-Averaged Navier Stokes equations in three dimensions. MPI library implemented in the code for interprocessor communications allows simulating engineering flow problems. I compared the performance of 3-noded linear and 6-noded quadratic triangular elements. In 3D, the relative performance is evaluated for 6-noded linear and 18-noded quadratic wedge elements. Numerical results are compared for the solutions to Euler, laminar, and turbulent flows at subsonic, transonic, and supersonic speeds. This in-house CFD code, which is written in Fortran and C, was used to simulate flow in various components of a ramjet engine.