Principle of sample polymerization and method of P-S-N curve fitting

In sense of probability percentiles, there exists a correspondence between specimen fatigues lives associated with different cyclic stress levels. Based on such a correspondence, the principle to fit P-S-N curve with small sample test data is investigated, which related with information fusion techniques, and the method to establish P-S-N equation. By polymerizing fatigue life test data under different cyclic stress levels and integrating the life distribution information, small sample experiment data can be efficiently applied to acquire the necessary information to estimate the parameters of the probability distributions of fatigue lives associated with any stress levels. According to the relationship between life distribution and cyclic stress level, fatigue life test data at different stress levels can be equivalently converted to an arbitrarily specified baseline stress level, and the distribution of the equivalent lives is identical with that of the test data at the baseline stress level if the relationship is correct. With the mean life-stress level relationship fitted by small sample test data, the correct relationship between life standard deviation and stress level can be found by statistical test oriented perturbation technique. Test results illustrate that the P-S-N curves fitted by this method with 20-30 specimens are very close to those obtained by fitted by conventional method with 40-60 specimens. ©2013 Journal of Mechanical Engineering. ©2013 Journal of Mechanical Engineering.