A New Fitting Method for Probabilistic Stress-Life (P-S-N) Curve with Three Parameter Weibull Distribution  
Author Ningxiang Wu


Co-Author(s) Liyang Xie


Abstract The log-normal distribution (LN) and the three-parameter Weibull distribution (3-PW) are two popular statistical distributions to describe Fatigue life. Although the LN is chosen more frequently in P-S-N curve fitting, the 3-PW is attractive since the non-zero lower limit of lifetime and its good fit effect to fatigue life test data. To reduce the number of fatigue tests required by P-S-N curve fitting greatly, this paper investigated the relationship between Weibull shape parameter and cyclic stress level by nine data series from literature, then presented an equivalent large size sample build-up method for P-S-N curve fitting based on 3-PW with a constant shape parameter. The method was verified by three different small sample data series selected from a large sample test data randomly.


Keywords P-S-N curve, Weibull distribution, small sample size, Fatigue test
    Article #:  RQD25-267
Proceedings of 25th ISSAT International Conference on Reliability & Quality in Design
August 1-3, 2019 - Las Vegas, NV, U.S.A.