Uncertainty Analysis of Aerodynamic Coefficients in an Automotive Wind Tunnel

This paper presents an uncertainty analysis of aerodynamic force and moment coefficients for production vehicles in an automotive wind tunnel. The analysis uses a Monte Carlo numerical simulation technique. Emphasis is placed on defining the elemental random and systematic uncertainties from the tunnel’s instrumentation, understanding how they propagate through the data reduction equations and under what conditions specific elemental error sources are or are not important, and how the approach to data reduction influences the overall uncertainties in the coefficients.

The results of the analysis are used to address the issue of averaging time in the context of maintaining a maximum allowable uncertainty level. Also, a maximum error requirement in the vehicle’s installation is suggested to allow the use of rapid but approximate vehicle alignment methods without incurring errors that exceed the data uncertainty.

Observed reproducibility results are presented spanning a 16 month period. The uncertainty of the drag coefficient over this period is shown to be ±0.002 at the 95% confidence level, with similar results for the other key coefficients. As the analysis presented in the paper focuses only on those uncertainty sources that can be captured explicitly by the data reduction equations, the observed variations are higher than the calculated uncertainties.