Design of an automated OpenFOAM workflow for external-aerodynamics simulation

Abstract

The central research question of this thesis is: Can the aerodynamic simulation workflow for automotive applications be fully automated using open-source tools such as Open- FOAM, while maintaining accuracy and mesh quality suitable for engineering evaluation? This question is significant in the automotive sector, where highly competitive development cycles demand rapid, cost-effective aerodynamic testing across multiple geometries. Traditional manual workflows limit the number of design iterations, whereas automation offers the potential to increase throughput while reducing human error and operational costs. The workflow developed in this project integrates Python scripting to automatically generate OpenFOAM case files from YAML configuration inputs. The meshing pro- cess leverages SnappyHexMesh, and simulations were performed using simpleFoam, potentialFoam, and turbulence models including k-ε, k-ω, and laminar flows. Post- processing was conducted using ParaFOAM, allowing automated extraction of aerodynamic coefficients. This approach was successfully applied to three benchmark cases: the AeroSUV, unitCube, and Ahmed Body, demonstrating the feasibility of a fully automated simulation pipeline. Key challenges were observed in the meshing stage, where SnappyHexMesh presented limitations in achieving consistently high-quality meshes, particularly in regions requiring refined boundary layers. In commercial workflows, tools like ANSA provide more robust control over mesh quality; however, their unavailability during this project necessitated fully open-source solutions. As such, the developed workflow stands as a proof of concept: it successfully automates the aerodynamic simulation process, reduces manual setup time, and enables testing of multiple geometries, but further refinement of mesh control and integration with advanced tools would be required for production-level deployment.