Search Results

The spur timing gears in Eaton superchargers operate at low torque loads and the supercharger system is especially sensitive to gear whine noise created by minute differences in the spur gear tooth profile quality. This has necessitated the grinding of very high quality profiles on high-contact-ratio spur gears. The manufacturing operation has used subjective evaluation of profile and lead measurements to qualify grinder diamonds and audit gear quality related to noise. They have also relied on supercharger end-of-line-testers to provide a direct measurement of gear noise as the primary quality feedback to the gear manufacturing process. Since the difference in the inspection plots of very high quality profiles is difficult to determine subjectively, the inspection process assessments have been difficult to correlate to the resultant gear noise measurements.

A spectral-based substructuring approach applying linear frequency response functions (FRF) is proposed for improving the accuracy of simulating the dynamics of coupled systems. The method also applies a least square singular value decomposition (SVD) scheme to overcome the inherent computational deficiency in the basic substructuring formulation. The computational problem is caused by the magnification of measurement errors during any one of the matrix inversion calculations required for this method. The primary objective of applying this approach is to examine the possibility of analyzing higher frequency response that is normally not possible using conventional modeling technique such as the direct finite and boundary element, and lumped parameter techniques. In this study, additional concepts are also evaluated to quantify the limitations and range of applicability of the proposed substructuring approach for simulating the vibration response of complex powertrain structures.