Journal of Mechanical Science and Technology 26 (6) (2012) 1663~1670 www.springerlink.com/content/1738-494x DOI 10.1007/s12206-012-0413-8 An innovative experimental on-road testing method and its demonstration on a prototype vehicle † José C. Páscoa 1,* , Francisco P. Brójo 2 , Fernando C. Santos 1 and Paulo O. Fael 1 1 Electromechanical Engineering Department, Faculty of Engineering, University of Beira Interior, Covilhã, 6201-001, Portugal 2 Aerospace Sciences Department, Faculty of Engineering, University of Beira Interior, Covilhã, 6201-001, Portugal (Manuscript Received July 2, 2011; Revised February 9, 2012; Accepted February 9, 2012) ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- Abstract Ground vehicle drag coefficient is herein obtained using an unconv entional on-road test in real scale. At low-Re numbers, and a s a function of velocity variations, transition introduces changes on the vehicle’s drag coefficient. Therefore, the drag coefficient must be obtained as a function of velocity. Traditionally, only an average drag coefficient value is usually obtained using the coast down method. To obtain the on-road, velocity dependent, drag coefficients we introduce a new approach. The aerodynamic resistance coefficient is obtained by towing the vehicle with and without an aerodynamic shield, in order to eliminate the rolling resistance component. A detailed description of the method, its associated techniques, and related errors is presented. We conclude that the present experimental procedure is needed when comparing the experimental drag coefficient against computational results, since numerical computations are usually performed in a velocity dependent framework. Further, the same on-road test procedure is herein used to obtain the rolling and aerody- namic drag coefficient for a prototype vehicle working in the transition regime. Keywords: Ground vehicle; On-road test; Experimental method; Aerodynamics ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 1. Introduction Generally speaking, for high-performing cars weight reduc- tion and engine efficiency are usually the two most important bottlenecks affecting fuel consumption, considered typically more important than aerodynamic drag reduction. However, a reduction in drag coefficient remains an important matter to tackle when designing these vehicles. Actually, aerodynamic resistance will certainly result on a measurable gain in per- formance, even if the drive cycle does not comprise high- speed roads. Another consideration, even more important, is that the designer must insure that the aerodynamic perform- ance improvements are transposed to the road conditions and not only achieved on controlled wind tunnel conditions, or in computational fluid dynamics (CFD) simulations. The ex- perimental, or numerical, modeling of the flow around ground vehicles is inherently complex, in particular due to boundary layer separation and ground effects [1]. This triggered the need to develop the means to obtain an accurate drag coeffi- cient in ground vehicles [2]. Most often, aerodynamic flow optimization for ground ve- hicles is usually performed in a wind tunnel, but this approach is associated to a series of similarity and dimensionality prob- lems. Even if we can ensure that the wind tunnel provides controlled and repeatable conditions, it cannot mimic in full the road conditions. Even in the most realistic case, when us- ing a moving floor wind tunnel, the boundary layer is not completely representative of road conditions. The moving belt floor must be synchronized with free stream, and boundary layer suction must be performed in front of the vehicle. This must be carefully matched, which is very difficult and can introduce difficulties in achieving good dynamic similarity conditions. Besides, the vehicle tires must be rotating in order for taking into account the energy losses due to their rotation. Additionally, blockage effects in full-scale tests for these bluff bodies also strongly affect the achievement of similarity con- ditions. Very often, experimental results obtained in diverse wind tunnels, for the same geometry and at the same Reynolds number, result in a scatter of aerodynamic coefficients by around 5% [3]. Albeit these deficiencies we can still resort to wind tunnel testing in order to improve the aerodynamics of ground vehicles. Considering that the resultant on-road drag coefficient will be slightly different from wind tunnel, but that the performance trends are correlated to the real conditions. This introduces us to the problem of obtaining the drag co- efficient from road testing. Track tests are complex, time con- suming and introduce problems of controlling the environ- mental conditions. For this kind of testing, coastdown is the * Corresponding author. Tel.: +351 275 329 763, Fax.: +351 275 329 972 E-mail address: pascoa@ubi.pt † Recommended by Editor Yeon June Kang © KSME & Springer 2012