Using Surface Texture Parameters to Relate Flat Belt Laboratory Traction Data to the Road 2015-01-1513

Indoor laboratory tire testing on flat belt machines and tire testing on the actual road yield different results. Testing on the machine offers the advantage of repeatability of test conditions, control of the environmental condition, and performance evaluation at extreme conditions. However, certain aspects of the road cannot be reproduced in the laboratory. It is thus essential to understand the connection between the machine and the road, as tires spend all their life on the road. This research, investigates the reasons for differences in tire performance on the test machine and the road. The first part of the paper presents a review on the differences between tire testing in the lab and on the road, and existing methods to account for differences in test surfaces.

The second part of this paper, presents a case study detailing the causes for difference in measurements and correlating the tire performance measured on the National Tire Research Center (NTRC) Flat-Trac® LTRe, Alton, VA and at Cooper Tire and Vehicle Test Center, Pearsall, TX. The case study presents the dry braking traction tests conducted on the machine and on the road for the same set of tires. The friction-slip ratio (μ-S) curves from both the test methods are compared, a model is then proposed based on the Dugoff tire model and surface texture parameters to relate tire performance from the lab to the road. The main differentiating factor which alters the friction levels between the tests on the machine and in the field is surface texture. The contribution of surface texture to tire-road friction is very significant, thus the lab to road model is developed with consideration of macro and micro texture parameters of the surface. Other distant differences, such as the effect of temperature, effect of road crown, and effect of dust particles are not accounted for the analysis. Surface texture parameters of asphalt and sandpaper are estimated using nonlinear optimization. In conclusion, the model predicts the tire performance on asphalt based on the measured tire performance on sandpaper with an acceptable percentage error. The validation of the proposed concept resulted in a good match between the peak friction value and the longitudinal stiffness of the tire, however a slightly high percent error was obtained when comparing the slip ratio at peak friction. The proposed concept is thus promising to relate tire performance measured on different surfaces, and future work is proposed that presents the possible enhancements to the lab to road model and make it more comprehensive and complete.