Parking-Specific Parameterization Method for FTire

Abstract

ABSTRACT Virtual steering system layout in the early development phase requires adequate tire models to predict realistic steering rack forces. An accurate representation of parking is particularly important, as the largest steering rack forces occur during this maneuver. Physical tire models are mainly parameterized for rolling conditions. Since the tire exhibits different mechanical behavior under nonrolling conditions, this article introduces a new parameterization procedure for the physical tire model FTire that characterizes the conditions during parking maneuvers. To this end, an additional full vehicle measurement setup is used to understand the tire motions, forces, and torques during parking. It is also shown that a tire model based on a standard parameterization procedure results in steering speed-dependent parking torque deviations of up to 17.5% when compared with component measurements. Thus, new measurement methods are developed to help parameterize the tire model for this maneuver. A linear friction tester is used to determine the friction interaction between tire and road at the relevant relative velocities. In addition, measurements are performed on a tire stiffness test rig, in which translatory and rotatory movements are overlaid. Furthermore, the contact patch shape, ground pressure distribution, and tire outer contour are digitalized and added into the model. A tire model based on the new parking optimized parameterization is then compared with the standard tire modeling approach and component measurements as well as the full vehicle measurements. In conclusion, improvements of up to 12% for drilling torque, up to 15% for longitudinal force, a more realistic lateral stiffness, a more realistic pressure distribution, and improvements of up to 8% when simulating the steering rack force can be stated. After the results are evaluated and interpreted, recommendations for future developments of this parameterization procedure and an extension of the virtual tire model are discussed.