Tooth Bending Research Articles

Paper 16: Gear Errors and Their Resultant Noise Spectra

The effects are considered of base circle eccentricity, pitch error, dropped tooth, tooth bending under load, and of involute error on the transmission errors of gears. Measurements accurate to ± 1/4 μm show that adjacent pitch errors are frequently randomly distributed. When subject to Fourier analysis, they can give rise to unexpected harmonics both in the errors of a gear and the noise produced by it. A cause of ‘phantom’ gear noise is suggested.

Technique for Measurement of Axle-Gear Dynamic Tooth Stiffness

This paper describes a program aimed at obtaining correlation between axle-noise levels and variations in stiffness of gear-teeth pairs during meshing and loading in an operating-gear set. The work described involves only the technique developed to measure the effective dynamic gear-tooth stiffness for each combination of pinion and gear teeth. In principle, this was done by measuring the change in time for passage of reference locations on the drive and driven shafts (due to tooth bending under load) relative to the unloaded time interval. The incremental time is related to incremental deflection angle by a precise measurement of average shaft speed. An optical-triggering system was used for the electronic timers, which had a resolution of ±1 μsec. This system was capable of resolving 1% variations in tooth stiffness. The data-acquisition procedure included a technique for averaging ten consecutive deflection measurements for each of the 645 tooth-mesh combinations involved. A Fourier analysis was performed on the 645 data points from two axles, thereby giving some indication of the noise spectra that might be expected from the gear sets.