Efficient Unmanned Aerial Vehicle Design: Automated Computational Fluid Dynamics Preprocessing from Geometry to Simulation

Current trends in the aerospace and UAV sectors emphasize integrating Artificial Intelligence (AI) technologies into the design process. AI technologies necessitate extensive data to capture the non-linearities in fluid phenomena. To address these needs, this work focuses on automating the data aggregation process for fixed-wing platforms, ranging from Micro–Mini to HALE-Strike UAVs, as classified by NATO. Specifically, this paper presents a framework for automating the tedious tasks required for geometry generation, mesh generation, and solution setup in a commercial Computational Fluid Dynamics (CFD) solver, for any arbitrary wing within the aforementioned design space. By combining various well-established open-source suites and commercial software via Python scripting, the preprocessing steps up to the solution require only a few minutes on a typical laptop workspace. Despite the rapid geometry acquisition, mesh generation, and solution setup through the pipeline, the guidelines and common practices for subsonic external flow simulations are still strictly followed. This results in solutions with a deviation of merely sub 5% from those of an experienced designer, even for the extremes of the flight envelope. The proposed framework significantly reduces design iteration times, enabling more efficient and innovative UAV development. Additionally, the framework’s ability to accumulate high-quality data for machine learning enhances predictive modeling and optimization capabilities across UAV design practices.