NTN Corporation, Kuwana,
Pages: 20 Published: Jan 2010
In this report, several methods for both rolling contact fatigue (RCF) life test design and result interpretation are introduced. These methods generate results using random numbers followed by Weibull distribution (i.e., Weibull random number). The first method illustrates a relationship between the minimum number of test specimens and the suspension time in a fixed time test required at L10 and L50 lives with an arbitrary reliability. This relationship is useful to maintain the qualitative reliability and avoid excessive quantitative testing. The second method can clarify a relationship between the given number of test specimens and resultant significant differences at L10 and L50 lives in an accelerated test with an arbitrary reliability. This relationship is also useful to estimate the appropriate number of test specimens based on statistical logic. Of note, calculations employing the Weibull random number can apply to not only RCF life test design but also estimations of the test results. The third method enables the determination of a range of L10 and L50 lives with an arbitrary reliability even if the number of test specimens is too small to estimate L10 or L50 lives from the Weibull plots. The fourth method can determine significant differences of L10 and L5 lives between any two given lots and allow a quantitative estimation of the minimum difference between their lives from data obtained by experiments. These methods provide techniques that are easier to understand as compared to the recent mathematical model, and they show enough flexibility to apply to almost all types of testing. These systems will therefore eliminate the need for qualified experiences related to the statistical design and result interpretation for RCF life testing.
rolling contact fatigue, Weibull, statistics, Weibull random number, Monte Carlo simulation
Paper ID: STP49129S