Comparative Investigation of Uncertainty Analysis with Different Methodologies on Fatigue Data of Rebars

Stochastic modeling of uncertainties for fatigue strength parameters is vital step as basis for reliability analyses. In this paper, the Maximum Likelihood Method (MLM) is used for fitting the statistical parameters in a regression model for the fatigue strength. Furthermore, application of the Bootstrap method is investigated. The results indicate that the latter methodology does not work well in the considered case study because of runout tests within the test data. Moreover, use of Bayesian inference with Markov Chain Monto Carlo implementation is studied. These results indicate reduction in the uncertainty and thus better parameter estimations