Gear noise can be a source of intense annoyance. It is often the primary source of annoyance even when it is not the loudest noise component. This is because of the way it is perceived. Gear noise is a collection of pure tones which the human ear can detect even when they are 10dB lower than the overall noise level. Another reason for our sensitivity to transmission noise is that we associate it with impending mechanical failure.

In the design of any new gear drive, the performance of previous similar designs is very carefully considered. In the course of evaluating one such new design, the authors were faced with the task of comparing it with two similar existing systems, both of which were operating quite successfully. A problem arose, however, when it was realized that the bending stress levels of the two baselines differed substantially. In order to investigate these differences and realistically compare them to the proposed new design, a three-dimensional finite-element method (FEM) approach was applied to all three gears.

Gears are currently run at high speed and under high load. It is a significant problem to develop lubricants and gears with high load-carrying capacity against scoring. The particles of molybdenum disulfide have been considered to increase the scoring resistance of the gears. The wear characteristics and the scoring resistance of the gears lubricated with MoS2 paste and MoS2 powder have been investigated. (1) However, there are few investigations on the performance of the gears coated with MoS2 film with respect to scoring.

The most conclusive test of bevel and hypoid gears is their operation under normal running conditions in their final mountings. Testing not only maintains quality and uniformity during manufacture, but also determines if the gears will be satisfactory for their intended applications.

Anyone involved in the design, manufacture and use of gears is concerned with three general characteristics relative to their application: noise, accuracy, and strength or surface durability. In the article, we will be dealing with probably the most aggravating of the group, gear noise.

It has previously been demonstrated that one gear of an interchangeable series will rotate with another gear of the same series with proper tooth action. It is, therefore, evident that a tooth curve driven in unison with a mating blank, will "generate" in the latter the proper tooth curve to mesh with itself.

Gear shaving is a free-cutting gear finishing operation which removes small amounts of metal from the working surfaces of the gear teeth. Its purpose is to correct errors in index, helical angle, tooth profile and eccentricity. The process can also improve tooth surface finish and eliminate, by crowned tooth forms, the danger of tooth end load concentrations in service. Shaving provides for form modifications that reduce gear noise. These modifications can also increase the gear's load carrying capacity, its factor of safety and its service life.

On gear drives running with pitch line velocities below 0.5 m/s so called slow speed wear is often observed. To solve some problems, extensive laboratory test work was started 10 years ago. A total of circ. 300,000 h running time on FZG back-to-back test rigs have been run in this speed range.

Your May/June issue contains a letter from Edward Ubert of Rockwell International with some serious questions about specifying and measuring tooth thickness.

Much of the information in this article has been extracted from an AGMA Technical Paper, "What Single Flank Testing Can Do For You", presented in 1984 by the author