This content is not included in your SAE MOBILUS subscription, or you are not logged in.

Computational Prediction of a Vehicle Aerodynamics Using Detached Eddy Simulation

Journal Article
2013-01-1254
ISSN: 1946-3995, e-ISSN: 1946-4002
Published April 08, 2013 by SAE International in United States
Computational Prediction of a Vehicle Aerodynamics Using Detached Eddy Simulation
Sector:
Citation: Castro, N., Lopez, O., and Munoz, L., "Computational Prediction of a Vehicle Aerodynamics Using Detached Eddy Simulation," SAE Int. J. Passeng. Cars - Mech. Syst. 6(1):414-423, 2013, https://doi.org/10.4271/2013-01-1254.
Language: English

Abstract:

In the present paper, Computational Fluid Dynamics (CFD) simulations of the aerodynamics of a station wagon using DES (Detached Eddy Simulation), based on the Spalart-Allmaras turbulence model, are discussed and compared with experimental results. DES is a non-zonal hybrid turbulence model that uses both Reynolds Averaged Navier-Stokes (RANS) and Large Eddy Simulation (LES), enabling a better equilibrium between the accurateness and the computational cost of the solution.
Simulations were run in parallel using the commercial software ANSYS/FLUENT v13.0, and required a computational grid of approximately 50 million cells. Some flow characteristics such as boundary layer separation, recirculation zones, and the entire pressure and velocity field were also obtained and analyzed. Computational results of drag force were compared with experimental results based on the SAE J1263 recommended practice (which describes a way to calculate the resistive loads that act on a vehicle during non- forced deceleration). Changes were made in the experimental estimation of the drag force, compared to the procedure stated in SAE J1263 (e.g., a differential formulation of the system was used, and the mass of the vehicle was altered by loading extra bodies into the vehicle). A selection among different sets of experimental conditions (vehicle velocities and vehicle weights) was made to improve the accuracy of the prediction.
Even though computational and experimental approaches involved different models and simplifications, results from each of them showed a fair agreement. Congruence was driven by the iterative improvements made on the boundary conditions and meshing parameters on the CFD approach, and the careful experimental design performed. The aerodynamic drag coefficient calculated by CFD was of 0.460 compared to 0.404 when calculated by experimental means, which represents a 13.8% difference or 56 drag counts.