Development and construction of a formula student wind tunnel model : experimental and numerical simulation correlation

This bachelor’s thesis presents the end-to-end development of a scaled wind tunnel model for Formula Student applications, with a primary focus on achieving experimental-numerical correlation in aerodynamic testing. The research addresses three core objectives: the systematic design and construction of a modular wind tunnel model, the evaluation of its aerodynamic performance through wind tunnel experimentation, and the validation of corresponding results using Computational Fluid Dynamics (CFD) simulations. The theoretical framework is based on aerodynamic principles, wind tunnel methodology, and CFD fundamentals, including the Navier–Stokes equations and turbulence modeling. The design process followed the VDI 2221 methodology and incorporated project management tools such as milestone planning, a bill of materials, and functional decomposition. Among the evaluated concepts, BASIC, BASIC-PLUS, and HIGH-TECH, the BASIC-PLUS variant was selected for its optimal balance between cost, performance, and modularity. The final model was physically assembled using machined components, custom aerodynamic elements, and integrated metrology systems. It was designed as a modular platform to support iterative development and future upgrades, allowing for component-level modifications and the integration of new features over time. The experimental phase involved force measurement and flow visualization techniques, gathering data on aerodynamic forces, pressure distribution, and velocity profiles across various test conditions. Numerical simulations were performed using CFD software, encompassing geometry preparation, mesh generation, solver configuration, and post-processing to extract aerodynamic coefficients and flow regimes. Comparative analysis revealed agreement between experimental and simulated results. For the wind tunnel model, deviations were 13.3% in lift and 20.9% in drag when comparing experimental data with its own CFD counterpart. For the full-scale car, the lift coefficient (cL) differed by 9.5% between experiment and simulation, while the drag coefficient (cD) showed a 17.1% deviation. Differences were primarily observed in regions with high turbulence and complex boundary layer behavior. These differences are attributed to scaling limitations, mesh resolution, and the influence of the blockage factor, which was not corrected despite the wind tunnel’s 15% blockage ratio. The discussion addresses model design efficiency, CFD fidelity, and limitations resulting from simplified physical modeling and environmental influences. The study concludes by affirming the model’s reliability for aerodynamic investigations and highlights its suitability as a modular platform for iterative, long-term development.