Reproduce “An Invariant Neural-Network Eddy-Viscosity Closure with Iterative Solver Coupling for Molten Salt Fast Reactor Thermal-Hydraulics” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Zaid Abulawi, William Oliveira, Akzhol Almukhametov, Jun Fang, Ling Zou, Rui Hu, Stefano Lorenzi, Yang Liu
The authors develop a neural turbulence closure for SAM that uses flow and geometry from neighboring mesh elements to estimate eddy viscosity, with iterative retraining on states produced by SAM. In a 120-second pump-coastdown, its time-averaged velocity-vector error was 17.73%, versus 37.6% for the mixing-length model. The reference is a two-dimensional axisymmetric RANS solution, and the test covers one trajectory.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.