Optimized random sampling for the load level method in Wöhler tests

Abstract

AbstractFor the optimal planning of Wöhler tests, DIN 50100 offers tables with different methods, such as the load level method, the pearl-string method and staircase method, to identify the required minimum sample size. Based on this norm, the sampling planning by means of the load level method was optimized in this work. Analytical formulas were developed with statistical methods. These formulas enable a fast estimation of the sample size for arbitrary use cases. Another optimization was the implementation of varying standard deviation within the area of fatigue strength in the load level method corresponding to DIN 50100. This optimization decreases the sample size compared to a constant standard deviation. Both optimizations were implemented via a MATLAB® based tool which also contains the methods of DIN 50100.