ALMA Simulation Engine for Applications in Magnetohydrodynamics and Computational Fluid Dynamics*

TIME: 10:00 AM

LOCATION: Dean’s Conference Room, ENG 203E

SPEAKER: Juan D. Colmenares, Mark Kostuk, Federico Halpern, Akshay Deshpande and Pranav Puthan, DIII-D National Fusion Facility, General Atomics

ABSTRACT:

General Atomics (GA) has been at the forefront of fusion energy research for more than 50 years and operates the DIII-D National Fusion Facility for the U.S. Department of Energy. By pioneering the science and innovative technology in fusion energy, GA has acquired expertise in scientific and high-performance computing (HPC) to tackle some of the most pressing challenges in diverse fields of science and engineering, including plasma physics and fluid dynamics. For example, developing liquid metal blanket technology is a crucial component to enable fusion as a power source, which requires advanced modeling capabilities. In this talk, we will discuss the development of the Anti-symmetric, Large-Moment, Accelerated (ALMA) simulation engine for solving problems in magnetohydrodynamics (MHD) and computational fluid dynamics (CFD), which can run efficiently on GPUs and scale to thousands of nodes [1]. ALMA uses an anti-symmetrized form of the MHD and Navier-Stokes equations, which leads to improved stability and conservation properties at extreme scales. The immersed-boundary method enables simulations of flows over complex geometries while making use of computationally efficient (structured) grids. A wall-modeling approach, developed by our group, introduces a forcing term to account for the under-resolved velocity profile near solid boundaries, allowing high Reynolds number simulations with coarse grids [2]. Recent developments include the addition of curvilinear grid capabilities, implementation of the elastic equations and surface tracking methods for fluid-structure interaction (FSI) problems.

*The ALMA numerical toolkit has been developed using internal research and development support at General Atomics. Plasma physics applications of ALMA are based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, Theory Program, under Award No. DE-FG02-95ER54309.

[1] Halpern, F. D., Sfiligoi, I., Kostuk, M., Stefan, R., & Waltz, R. E. (2021). Simulations of plasmas and fluids using anti-symmetric models. Journal of Computational Physics, 445, 110631.
[2] Colmenares, J. D., Kostuk, M., & Halpern, F. D. (2023). A Wall-Modeling Approach for Immersed Boundaries Based on Momentum Forcing. AIAA Aviation Forum, June 2023, AIAA 6.2023-4335.

Speaker Biography: Dr. Juan Diego Colmenares is a computational scientist with expertise in high-performance computing (HPC), computational fluid dynamics (CFD), and magnetohydrodynamics (MHD). He currently works as a Scientist in the Advanced Computing Group, at General Atomics (GA). He is one of the primary developers of the ALMA simulation engine, designed to solve MHD and Navier-Stokes equations in anti-symmetric form using GPU acceleration, scaling on leadership class supercomputers. Before joining GA, he worked as a postdoctoral appointee at Argonne National Laboratory, where he worked on modeling high-pressure fuel injection sprays using the high-order spectral element method (SEM). Dr. Colmenares holds a Ph.D. in Engineering from the University of New Mexico. His doctoral research focused on designing and performing direct numerical simulations (DNS) of turbulent mixing layers using SEM. He received his B.S. and M.S. in Mechanical Engineering from the Universidad de los Andes, Colombia.

HOST: Xiaofeng Liu