Rupture detection in fatigue crack propagation

Romain Azaïs (1, 2) , Anne Gégout-Petit (3, 4) , Florine Greciet (5, 4, 3)

1 RDP - Reproduction et développement des plantes
2 MOSAIC - Simulation et Analyse de la morphogenèse in siliCo
3 BIGS - Biology, genetics and statistics
4 IECL - Institut Élie Cartan de Lorraine
5 Safran Aircraft Engines

This chapter focuses on piecewise‐deterministic models for fatigue crack propagation (FCP) with a particular focus on estimation issues. In this setting, the process is usually observed on a temporal discrete grid and through an additive noise, which makes unknown both the continuous trajectory and the mode. Experimental data obtained by Virkler is a well‐known source of information about the fatigue of engineering materials. The chapter describes the process of the acquisition of these data. It considers stochastic models to describe the evolution of crack propagation in fatigue. Stochastic models for propagation and rupture can be broken down into two families. The first is statistical modeling of crack propagation data, while the second includes that include randomisation stochastic models derived from phenomenological laws. The chapter discusses the use of piecewise‐deterministic Markov processes (PDMPs) for modeling crack propagation, and presents some models found in the literature.

Domaines: Méthodologie [stat.ME]; Applications [stat.AP]

Romain Azaïs, Anne Gégout-Petit, Florine Greciet. Rupture detection in fatigue crack propagation. Romain Azaïs; Florian Bouguet. Statistical Inference for Piecewise-deterministic Markov Processes, Wiley, pp.173-207, 2018, 978-1-786-30302-8. ⟨10.1002/9781119507338.ch6⟩. ⟨hal-01862267⟩