Combining Machine Learning with Computational Fluid Dynamics using OpenFOAM and SmartSim

TL;DR

OpenFOAM-based CFD simulations face data-exchange and compute-heterogeneity challenges when integrating with ML at scale. The authors adopt a loosely coupled CFD+ML framework using SmartSim as an orchestration layer and SmartRedis as an in-memory data hub, bridging OpenFOAM with ML backends through a lightweight API; they also illustrate a distributed SVD workflow where the data matrix $X \\in \\mathbb{R}^{M\\times N}$ is partitioned into $S$ blocks and merged to recover the global SVD. Three educational use cases—Bayesian optimization for turbulence parameters, partitioned SVD on streaming CFD data, and AI-in-the-loop mesh motion with an ANN—demonstrate the approach on heterogeneous HPC resources. The work provides open-source implementations, an OpenFOAM sub-module, and tutorials to accelerate adoption of scalable CFD+ML workflows.

Abstract

Combining machine learning (ML) with computational fluid dynamics (CFD) opens many possibilities for improving simulations of technical and natural systems. However, CFD+ML algorithms require exchange of data, synchronization, and calculation on heterogeneous hardware, making their implementation for large-scale problems exceptionally challenging. We provide an effective and scalable solution to developing CFD+ML algorithms using open source software OpenFOAM and SmartSim. SmartSim provides an Orchestrator that significantly simplifies the programming of CFD+ML algorithms and a Redis database that ensures highly scalable data exchange between ML and CFD clients. We show how to leverage SmartSim to effectively couple different segments of OpenFOAM with ML, including pre/post-processing applications, solvers, function objects, and mesh motion solvers. We additionally provide an OpenFOAM sub-module with examples that can be used as starting points for real-world applications in CFD+ML.

Figures (7)

• Figure 1: Online ML approximation of CFD results from OpenFOAM in SmartSim. Full lines represent transitions between algorithm steps, dashed lines are database operations.
• Figure 2: A bidirectional CFD+ML algorithm with OpenFOAM and SmartSim. PDE solution block is ommitted in the CFD algorithm for brevity, full lines represent transitions between algorithm steps, dashed lines are database operations.
• Figure 3: The magnitude of the velocity from the large-eddy simulation (LES) of the Pitz and Daily case (a), the Reynolds-Averaged Navier Stokes using the default $k-\epsilon$ coefficients (b), and the optimal coefficients (c). Note that for the LES simulation, the plotted field is the large-scale U that has had the eddy components filtered out.
• Figure 4: Comparative ParaView visualization of the first 4 left singular vectors (columns of $\mathbf{U}$); each subplot shows the magnitude of the respective vector field.
• Figure 5: Mesh deformation using Artificial Neural Networks.