Search Results

This paper presents a validation method for indoor testing of a passenger car suspension. A study was done to design a supporting modular structure with comparable inertances with respect to a vehicle's actual suspension and body connection points. For the indoor test, the rear axle is positioned on a rotating drum. The suspension system is excited as the wheel passes over cleats fixed on the drum and transient wheel motions are recorded. The indoor test rig outputs (i.e., wheel and chassis accelerations) were compared with experimental data measured on an actual vehicle running at different speeds on the same set of cleats along a flat road. The comparison results validate the indoor testing method. The forces and moments acting at each suspension and chassis connection point were measured with a set of patented six-axis load cells. The forces, moments, wheel and subframe accelerations were measured up to 120 Hz.

As vehicle development timelines continue to shorten and more emphasis is put on simulating vehicle dynamic phenomena; the importance of having physically correct inputs increases. For modeling the road noise phenomena, there are some methods used in the industry such as application of experimental spindle forces or vertical displacements applied to the tire patch. Each of these has limitations with respect to absolute accuracy or dependency of the input on suspension characteristics. For accurate evaluation of new designs, an invariant input which can reproduce measured vehicle cabin response is beneficial. Specifically, it is desired that significant improvement can be made over the spindle force method. To this end, a tire model derived from experimental data has been developed, along with three degree of freedom tire patch input displacements.

As vehicle development timelines continue to shorten, it is necessary for the full vehicle NVH engineer to be able to predict performance without actual prototypes. There has been significant advancement in the accuracy of finite element modeling techniques of trimmed bodies; however accuracy is still low in the road noise mid frequency range from 150-400Hz. Also, calculation times for these frequencies are long, with very large results files in some cases. To alleviate these limitations, a Hybrid approach has been used, where a finite element suspension and drive train model is coupled with a test based Frequency Response Function (FRF) model of the trimmed body. The predicted road noise level was compared to actual vehicle tests and exhibited excellent correlation.