Small sample theory for reliability design

Abstract

This paper outlines a new technique to address the paucity of data in assessing fatigue strength reliability. New formulations are presented for scatter factors, dealing with conditions where the population standard deviation is unavailable and where fatigue test results are incomplete. Reduction factor formulae for these two cases are also given. Finally, the concepts are applied to two sets of experimental data, demonstrating the practical and economic use of the proposed technique. It is shown from these examples that, when the standard deviation of the population is unknown, then safe fatigue life and fatigue strength can be obtained realistically according to the small sample test data using the new formulae for the fatigue life scatter factors and for the fatigue strength reduction factor.