Validation of CFD predictions for flow over a full-scale formula student vehicle using PIV in real conditions

Abstract

Computational Fluid Dynamics (CFD) predictions are becoming an industry standard. They allow for making accurate predictions of complex problems without requiring extensive real-world testing, as well as saving time and money. However, it has been proven many times that the classic Reynoldsaveraged Navier – Stokes (RANS) approach has its flaws and fails to provide highly accurate predictions. Even though CFD only approaches a physical solution, which can be reached only in very specific applica-tions, it usually provides enough precision for engineering purposes. To reach a convergence with real-world physics, plenty of factors must be taken into consideration like mesh, boundary conditions, and turbulence models. In order to obtain a CFD simulation that accurately represents real physics, some kind of real-world validation must take place. For aerodynamics, it is usually done in wind tunnels, which are expensive to run but provide controllable conditions to match those specified in CFD. One of the many methods used to validate the calculations is Particle Image Velocimetry (PIV). This study tries to validate CFD of a Formula Student car using PIV, but in realworld conditions, without wind tunnel. The compact size of equipment required for PIV testing and flexibility of CFD boundary conditions allow for that.