Measurement Location Optimization of Component Transfer Path Analysis Method for Road Noise 2020-01-1581

This paper compares the application of a non-intrusive on-site component Transfer Path Analysis (TPA) method used to identify interface forces to a previously performed direct blocked forces measurement. The latter is the current practice for determining the load cases related to isolated vehicle components. Force transducers are placed between the investigated source component and a specific rigid measurement rig. The comparison presented aims to answer the question of whether the faster and cheaper TPA method can produce accurate enough interface forces. The TPA method used in this work calculates the force component contributions without disassembly of the interfaces and through the local stiffness of multiple indicator positions per interface combined with operational measurement. The method is based on the application of an inverse matrix model. This approach is applied to a vehicle road noise investigation carried out on a roller bench at different roller speeds. A rough surface single axle roller bench in the same conditions as for the blocked forces method is used. The matrix inversion is carried out by using an overdetermined set of linear equations using an optimum number of degrees of freedom (DoFs). The paper mostly shows the results for the wheel spindle forces and moments at the wheel centre virtual point. The comparison between the inverse method and the direct measurement of loads shows very similar results and phenomena likely to appear in road noise excitation, as component resonances and tyre cavity modes are clearly visible. The matrix inversion method is observed to be a useful alternative if indicator DoFs and reference position selection are performed with care.