Frequency-Based Substructuring Decoupling for Tire Characterization

Due to the vehicle electrification trend and the consequent phasing out of internal combustion engines, tire-road interaction is nowadays one of the main contributors to inner cabin noise. To handle the ever-increasing number of components that contribute to noise, the current industry trend is to characterize the full vehicle Noise, Vibration and Harshness (NVH) profile through the virtual assembly of various subcomponents. The level of prediction of the current coupling methodology is strictly dependent on the quality of information at the interface of each subcomponent. Furthermore, when determining realistic component interface transfer functions, especially in real operational conditions, the significant difference between experimental measurements and finite element models is an increased difficulty. Noise, positioning errors, inaccessibility of such locations and difficulties in recreating some operational conditions are the primary causes in tests, therefore characterizing the behaviour of tires, under different conditions (static, rolling) is of primary interest. Frequency-based Substructuring (FBS) is a potential technique for indirectly characterizing the dynamics of an unknown component (tire) when coupled to a known structure (suspension and body), also in such a difficult condition as rolling. In this thesis the main goal is the evaluation of experimental FBS Decoupling performances at tire level, investigating and applying this technique to identify tire dynamic behaviour when coupled to a variety of supporting structures, ranging from a bulky test rig to a vehicle, and in a variety of boundary conditions, including static and rolling. In fact, FBS decoupling is the most suitable method in an experimental environment because it deals with frequency response functions commonly measured in NVH applications. The number and quality of measurements provided to FBS decoupling strongly affects the level of prediction. A secondary experimental setup in which the preloaded tire is free at the hub but constrained at the contact patch (fixed-free condition) is used to validate the static component level prediction. Successively, considering both interface and external degrees of freedom (internal to the known structure), various optimization techniques have been used to improve target prediction. Finally, some conclusions and recommendations to develop and improve future potential projects will be provided.