670725 The Reliability Evaluation of Transmission Gears W. E. Schilke Allison Div., General Motors Corp. THE DESIGN AND DEVELOPMENT OF heavy-duty transmis­ sions includes, as one step in the design procedure, the eval­ uation of component reliability. These transmissions are used in a wide variety of duty cycle conditions, ranging from earthmoving vehicles (such as scrapers) and working vehicles (such as front-end loaders) to military tracklaying vehicles. Correlation methods have been established for various categories of transmission components such as shafting, gears, clutches, bearings, and so on, which allow a priori estima­ tion of the reliability of that component operating with a given duty cycle. This discussion will be limited to reliability evaluation as it applies to gears. The results of Allison gear testing will be presented, with particular emphasis upon studies of the pitting fatigue mode of failure. The analysis of transmission duty cycles, as applied to gears, will be discussed, along with methods for determining the reliability of an assembly made up of a number of com­ ponents, and tradeoff methods for improving the reliability of a transmission assembly. 1. GEAR RELIABILITY 1.1 GENERAL FORM OF RELIABILITY FORMULA - A wide variety of components have been shown experimentally to have a cumulative life-reliability distribution of the form (1) *: Gear testing to be discussed later has shown that Eq. 1 also applies to gear pitting fatigue failure phenomena. In Eq. 1, R(that is, reliability) is defined as the probability that the component will perform its specified task under specified conditions for a specified time (or number of cy­ cles) N without failure. In Eq. 1, commonly known as the Weibull cumulative distribution, 8 is termed the charac­ teristic life and N the minimum life. If this relation is o plotted on special graph paper having a vertical scale pro­ portional to log/log R/and a horizontal scale proportional to log (N-Nq), the slope of the resulting straight line, b, provides a measure of the dispersion or scatter in N values of the data. 1.2 MODES OF FAILURE IN GEARING - A number of modes of gear failure have been described in the literature (2). * Numbers in parentheses designate References at end of paper. — ABSTRACT Pitting fatigue of gear teeth, if allowed to progress through continued operation, will result in tooth breakage and loss of drive. Laboratory test gears were run to determine their probability of failure under this mode. It was found that the dispersion or scatter of life values depends upon the gear surface contact stress; thus, there is less scatter at higher stress and, conversely, more scatter at lower stress. A cumulative damage method of evaluating transmission duty cycles is presented, with particular application to gear pitting fatigue, in a form adaptable to digital computer cal­ culation. 2317 Downloaded from SAE International by North Carolina State Univ, Saturday, September 15, 20182318 W. E. SCHILKE Included in the list are bending fatigue, pitting fatigue, scoring, case crushing, and so on. Certain of these modes, such as pitting and bending fatigue, tend to be of a progres­ sive nature, appearing only after relatively long hours of serv­ ice. Other modes of failure have been found to be more of a "go/no-go" nature. Scoring, for example, either appears very early in the life of a pair of gears, or it never occurs; the success of a particular gear design with respect to this failure mode is usually established during the design process by a pass/fail type of analysis, rather than by a progressive failure type of evaluation. Two modes of gear failure which are progressive in na­ ture, and therefore warrant a study of failure rate, are bend­ ing fatigue and pitting fatigue. 1.2.1 Bending Fatigue - The bending fatigue of gear teeth, like other bending fatigue phenomena in ferrous materials, is characterized