Correlation of Cloud Based Computational Fluid Dynamics Simulations to Wind Tunnel Test Results for a NASCAR XFINITY Series Vehicle

Abstract

The cost of setting up and maintaining a high performance computing cluster for large scale CFD usage is too expensive for many smaller motorsport organizations, and so the turn to cloud based computing resources is an attractive one. Cloud based computing centers allow users access to a shared computing cluster and charge based on the amount of resources used by each account. Efficient use of a cloud based computing center necessitates optimizing the CFD simulations to maximize accuracy and minimize cost due to the charge structure in place. This paper attempts to optimize steady state RANS simulations through systematically altering the refinement settings within the simulation mesh. These simulations are conducted using OpenFOAM on two NASCAR XFINITY Series vehicles and are validated using wind tunnel data. The effects of mesh refinement near the surface of the model and the refinement level within a bounding box around the vehicle on the aerodynamic forces of the vehicle are studied and related to the cost of running each simulation. A more computationally intensive transient simulation was also conducted and was not found to have a significant influence on the accuracy of the results beyond that of the steady state simulations.