DARSM embeds a neural network inside an implicit algebraic Reynolds stress model derived from transport equations, trains it end-to-end via adjoint PDE optimization, and reports 2-4x average velocity error reduction plus generalization from attached to separated flows on duct and hill benchmarks.
Learned turbulence modelling with differentiable fluid solvers: physics-based loss functions and optimisation horizons
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
physics.flu-dyn 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Deep Learning-based Algebraic Reynolds Stress Closures for RANS Simulations of Turbulent Flows
DARSM embeds a neural network inside an implicit algebraic Reynolds stress model derived from transport equations, trains it end-to-end via adjoint PDE optimization, and reports 2-4x average velocity error reduction plus generalization from attached to separated flows on duct and hill benchmarks.