Lifetime of a mechanical system

In a series of three dossiers, we'll be looking at different approaches to the analysis of random loading, with the aim of enriching life calculation methods.

This first dossier [BM 5 030] is devoted to analysis methods. The second deals with the modeling of random stresses. A stress, or the combination of several stresses, is considered to be the main cause of the reduction in strength of the mechanical components under consideration. Methods for taking into account the consequences of random loading on the service life of a mechanical component will be presented in the third part [BM 5 032] .

The motivation for this paper is based on the observation that over-simplification of a random load can lead to a significant loss of content and, consequently, to the loss of the right information derived from real conditions of use. The analysis of a random load is carried out in several ways and according to several approaches, with the aim of subsequently assessing the uncertain nature of the service life of a mechanical component.

Statistical and frequency analyses are two complementary approaches. Statistical analyses have the advantage of leading to probabilistic models , which can be used to model the natural dispersion of the stresses studied and their consequences (cracking, fatigue, damage, service life, etc.). The disadvantage of these statistical analyses, however, is that they ignore event histories. On the other hand, frequency analyses attempt to remedy this drawback, by using relationships that exist between, on the one hand, the frequencies contained in the stress under consideration and, on the other hand, either the average amplitudes measured (studied with the Fourier transform, FT), or their dispersions (studied with the power spectral density, PSD)