Aerodynamic drag force is identified as the main cause of vehicle energy consumption at high speed, so aerodynamics becomes the main vector to be addressed to improve vehicle efficiency in highway driving conditions. To do so, three different study techniques can be used: wind tunnel, real driving tests and CFD simulations and the equivalency between all three techniques needs to be certified in order to allow combining all three methods along vehicle programs and have a robust decision-making process. The authors show in this paper a comparison of all these three assessment methods, it presents a methodology to compare wind tunnel, coast-down and CFD results and ultimately demonstrates an acceptable equivalency between all these techniques with appropriate boundary conditions. This paper also presents a new measurement device to characterize incoming airflow profiles in different driving conditions, such as crosswind or vehicle following conditions. The authors undertook a wind-tunnel testing campaign with a baseline vehicle and tested several vehicle add-on parts that modified the vehicle aerodynamics, then performed coast-down tests of all these configurations to obtain the F2 forces, which are deemed equivalent to the aerodynamic forces, to confirm the consistency between the wind tunnel and coast-down methods and finally simulated the same scenarios by means of CFD computation. Besides comparing the overall aerodynamic drag forces on the vehicle with these three methods, the authors undertook a deeper flow field analysis including velocity probes, pressure sensors, tomography measurements, etc. and performed constant speed tests in the proving ground to get better understanding of real driving aerodynamics. Vehicle aerodynamics CFD simulation tends to be a tough task for engineers and computers, this difficulty comes from the fact that modelling the chaotic behaviour of turbulence around bluff bodies like cars is still one of the main challenges in computational mechanics. In this context, to assess the accuracy of standard CFD methods utilized by the automotive industry, the authors generated a DOE that included 3 different vehicle configurations, then tried with different turbulence models (steady state & transient), as well as open-road and wind-tunnel tests conditions and compared all these results with coast-down and wind tunnel tests. A digital version of the S2A wind tunnel model was created and correlated, first in empty wind-tunnel configuration, to validate different wind-tunnel features, such as: boundary layer suction system, moving belt, nozzle, collector, etc., and then with the real vehicle tests, whose results are presented in this paper. This document will describe a reasoned methodology to compare wind tunnel results, proving ground tests and CFD simulations, while considering key phenomena, such as, wind-tunnel blockage, pressure gradients, wheels ventilation drag, electric powertrain push effects or the nondimensionalisations of wind tunnel simulations. Ultimately, this paper demonstrates the importance of using realistic wind-tunnel models for improving the accuracy of aerodynamics CFD simulations.
Mr. Enric Aramburu, Fluid Engineering Product Manager, IDIADA Automotive Technology SA