The availability of reliable tire simulation models is necessary for performing accurate vehicle-handling simulations. Parameterizing of tire models, such as the Magic Formula (MF) tire model, means extensive measurement and complex fitting procedures. In addition, a general problem is that parameterized MF models are not simply adaptable to other track surfaces (e.g., dry, wet, or snowy tracks), which is a problem, especially for winter tire modeling. To face this drawback, a research project in cooperation between BMW and the Karlsruhe Institute of Technology, Institute of Vehicle System Technology, has been initiated.
The institute's internal drum test bench provides the opportunity to perform measurements on different track surfaces and various operating conditions. To identify main effects on tire performance and tire characteristics, comprehensive measurements on snow, ice, wet asphalt, and dry Safety-Walk surfaces have been carried out using three different winter tires.
Experimental designs have been worked out using the method of design of experiments (DoE) to reduce the number of measurements and to decrease measuring expenditure, especially on snow track surfaces. By using DoE, all statistic effects can be analyzed despite reducing the number of measurements.
Measurement data have been analyzed using extensive statistical methods. Thereby, effects on the tire characteristics have been empirically identified, and general predications will be presented in the article. We show identified main effects of track and ambient conditions on tire performance and tire characteristics. Furthermore, this article demonstrates the approach of using DoE to perform lean measurements as well as illustrates the realization of executing the measurements on different track surfaces on the test bench.
These results will also be a starting basis for establishing a novel empirical model for adopting tire characteristic curves and MF tire models on alternative road and ambient conditions.