An interval analysis method for fatigue crack growth life prediction with uncertainty

This paper presents an interval analysis method for the fatigue crack growth (FCG) life prediction with uncertain-but-bounded parameters. The method can effectively address the difficulty in the uncertainty modeling of the FCG life due to the lack of experimental samples. The uncertain parameters are quantified by interval variables which are classified into correlated and uncorrelated ones. All the intervals are uniformly transformed into an independent standard space, in which the standard intervals are then divided into subintervals to deal with the large uncertainty problem involved in the calculation of the FCG life. The subinterval analysis method is then employed to predict the upper and lower bounds of the FCG life. The proposed method is well validated by the numerical examples and experimental test. It is demonstrated that this method possesses an acceptable computational accuracy even there exist large uncertainty and correlation in parameters. Moreover, the correlation coefficient between crack propagation rate constants shows a considerable effect on the predicted FCG life. (C) 2018 Elsevier Ltd. All rights reserved.