The Influence of Different Unsteady Incident Flow Environments on Drag Measurements in an Open Jet Wind Tunnel

Abstract

Aerodynamic development for road vehicles is usually carried out in a uniform steady-state flow environment, either in the wind tunnel or in Computational Fluid Dynamics (CFD) simulations. However, out on the road, the vehicle experiences unsteady flow with fluctuating angles of incidence β, caused by natural wind, roadside obstacles, or traffic. In order to simulate such flow fields, the Forschungsinstitut für Kraftfahrwesen und Fahrzeugmotoren Stuttgart (FKFS) swing® system installed in the quarter scale model wind tunnel can create a variety of time-resolved signals with variable β. The static pressure gradient in the empty test section, as well as cD values of the Society of Automotive Engineers (SAE) body and the DrivAer model, have been measured under these transient conditions. The cD measurements have been corrected using the Two-Measurement Correction method in order to decouple the influence of the unsteady flow from that of the static pressure gradient. The investigation has determined that the static pressure gradient in the empty test section varies greatly with different excitation signals. Thus, it is imperative to apply a cD correction for unsteady wind tunnel measurements. The corrected cD values show that a higher signal amplitude, as in, signals with large β, lead to higher drag forces. The influence of the signal frequency on drag values varies depending on the vehicle geometry and needs to be investigated further in the future.

An Estimation of the Effect of Turbulence from the Natural Wind and Traffic on the Cycle-Averaged-Drag Coefficient

“…This approach was pioneered by Cogotti, [16], with the Turbulence Generating System, (TGS), in the Pininfarina full-scale wind tunnel, while similar model scale facilities have been built at FKFS, [17], and Durham University, [18]. A recent investigation by Fei et al, [19] using the model wind generator (SWING) system at FKFS with two DrivAer and two SAE model configurations found that for three configurations the unsteady drag coefficient variation with turbulence intensity was predicted using a quasi-steady approach based on the yaw angle distribution. For the other configuration the drag rise with turbulence was correctly predicted but was higher than the experimental results by approximately ΔCD = 0.004.…”

“…This paper is based completely on the assumption that the effect of turbulence in the atmosphere can be predicted using a quasi-steady approach as suggested from the study by Fei et al, [19]. The authors of that paper make the comment that the turbulence generated by the moving vanes in their wind tunnel may lack some high frequency Cooper and Campbell, [21], found that their quasi-steady theory, when compared with drag data for model scale trucks in grid generated turbulence in the wind tunnel, gave qualitatively correct results but underestimated the magnitude of the changes due to turbulence.…”