Optimization of an Accelerated Step-Stress Fatigue Test Plan

Abstract

The design of a component subjected to fatigue requires an experimental qualification of the couple material/process fatigue characteristics. The current industrial requirements involve that this qualification has to be carried out on the smallest possible number of tests and within the smallest possible testing time to censorship. In the same time, the accuracy of estimations has to remain significant. To do this, accelerated step-stress fatigue tests are quite efficient. Moreover, the accuracy of this kind of test can be optimized thanks to a good distribution of the steps. This paper illustrates on a particular example the process to obtain an optimized test plan by this way. Its efficiency is compared to that obtained with other kinds of plans and the advantage of this method is shown out for the design of a component. Once the advantage of a step-stress fatigue test explained, the optimization process used to define an efficient test profile is explained. The efficiency can be evaluated by stochastic simulations or real tests defined with an optimized profile. To show out the applicability of the method and to evaluate its efficiency, the method will be carried out to design an actual mechanical component (a window mechanism) with a reliability approach.

Key Words:

• Accuracy
• fatigue
• optimization
• reliability
• SSALT

Additional information

Notes on contributors

P. Lantieri

Pascal Lantieri is Doctor Engineer in mechanics, researcher in the Laboratory of Process and Materials at Arts & Metiers ParisTech (center of Angers, France). His researches are in the area of statistics applied to mechanical or electronic components. He also teaches mathematics and statistics applied to fatigue or Statistical Process Control for students in mechanical engineering.

A. Bignonnet

Andre Bignonnet is Doctor Engineer in applied mechanics, formerly Head of the Applied Mechanics and Technology Research division at PSA Peugeot Citroen, is an independent consultant for Fatigue and Structural Reliability and Associate Professor at Arts & Metiers ParisTech.