Abstract
Power transmission gears are usually used under fluctuating loads. But in the case of strength design of such gears we cannot get sufficient fatigue data and therefore the effectiveness of Miner's rule on fatigue strength is uncertain. This report deals with a new gear tooth pulsator for program and random fatigue tests and with the results of a test which was carried out under a programmed 16 steps loading. The frequencies of loads are distributed uniformly and the variation of mean load is like a saw tooth wave with positive and negative slopes. Normalized S38C and carburized SCM21 gears were tested. Every test gear had 27 teeth, 4mm module and 8mm width. Under the above condition the effects of loading sequence, changing rate and mean level of program load, gear material and heat treatment were investigated. And, though Hardrath has shown a program loading sequence effect for 7075-T6 aluminium alloy, the author's experiment indicates that its effect is not significant at the 5% level for both S38C and SCM21 gears. By arranging the cumulative damage statistically, it is concluded that Miner's theorem is not valid in the strict sense and further investigation is necessary to conclude whether Miner's rule is applicable to the fatigue design of gears.
