Presented "Symmetry-aware Reynolds-averaged turbulence modeling with equivariant neural networks" at the APS Division of Fluid Dynamics meeting.
Aaron Miller
PhD Candidate in Applied Mathematics
Harvard University
I study the statistical description and modeling of turbulent flows. I'm advised by Petros Koumoutsakos in the Computational Science and Engineering Lab.
Previously, I studied physics and applied math at UNC-Chapel Hill, where I worked with Joaquín Drut.
Recent updates
Began a DOE CSGF research practicum at Sandia National Laboratories on reduced-order models for turbulent reacting flows.
Awarded the DOE Computational Science Graduate Fellowship on the Mathematics/Computer Science track.
Papers
Simulating rapidly distorted turbulence statistics using spherical t-designs
Sahil Kommalapati, Aaron Miller, Robert Moser
Structure tensor Reynolds-averaged Navier-Stokes turbulence models with equivariant neural networks
Aaron Miller, Sahil Kommalapati, Robert Moser, Petros Koumoutsakos
Conferences
Symmetry-aware Reynolds-averaged turbulence modeling with equivariant neural networks
APS Division of Fluid Dynamics · 78th Annual Meeting · Contributed talk