Rocket flow: Understanding Drag Reduction through Geometry Analysis

Advanced evaluation methods such as wind tunnel testing and high-fidelity Computational Fluid Dynamics (CFD) simulations are often exceedingly expensive or computationally intensive for most researchers in aerospace engineering, particularly in rocket design. This project presents an accessible, scalable, and accurate approach to optimizing rocket geometry for drag reduction. A Python-based system was developed to generate thousands of parameterized rocket models using OpenSCAD, which were simulated in OpenFOAM to extract aerodynamic features. The resulting dataset was analyzed using ensemble machine learning models, including AutoGluon, achieving an R^2 of 0.989. SHAP analysis identified fin height, fin base, and body radius as the most influential geometric features determining drag. This integrated CFD-ML framework enables rapid, interpretable, and cost-effective aerodynamic optimization.