Forward Gradients for Data-Driven CFD Wall Modeling
This addresses the computational bottleneck in data-driven CFD wall modeling for industrial applications like gas turbine design, though it is incremental as it applies an existing gradient method to a specific domain.
The paper tackles the high computational cost of training machine learning wall models for Computational Fluid Dynamics by applying forward gradients, which compute unbiased gradient estimates in a single forward sweep to avoid back-propagation bottlenecks, potentially enabling efficient surrogate models for wall-bounded flow simulations.
Computational Fluid Dynamics (CFD) is used in the design and optimization of gas turbines and many other industrial/ scientific applications. However, the practical use is often limited by the high computational cost, and the accurate resolution of near-wall flow is a significant contributor to this cost. Machine learning (ML) and other data-driven methods can complement existing wall models. Nevertheless, training these models is bottlenecked by the large computational effort and memory footprint demanded by back-propagation. Recent work has presented alternatives for computing gradients of neural networks where a separate forward and backward sweep is not needed and storage of intermediate results between sweeps is not required because an unbiased estimator for the gradient is computed in a single forward sweep. In this paper, we discuss the application of this approach for training a subgrid wall model that could potentially be used as a surrogate in wall-bounded flow CFD simulations to reduce the computational overhead while preserving predictive accuracy.