Design against fatigue failures: Lower bound P-S-N curves estimation and influence of runout data

In the present work, three methodologies for estimation of fatigue design P-S-N curves are compared. The approximate Owen with staircase method, a method based on likelihood ratio confidence interval, and an original bootstrap approach considered. Three experimental datasets employed validation, showing that methods permit to have reliable number data and, accordingly, test duration can be significantly reduced proper testing strategy.