Co-authors​ :

 François LAMMING (FRANCE), Nacera BEDRICI , Stéphane GILLET , Sébastien JOANNÈS (FRANCE) 

To meet the challenges of global warming, optimising the energy efficiency of vehicles is an incentive to take greater advantage of composite materials with an organic matrix, in particular a thermoplastic one. These easily moldable materials significantly reduce structural weight while offering good recyclability and high production rates. Initially intended for low-stressed components, thermoplastic composites, such as PA66GF50, are now well-suited to meet the requirements of parts subject to severe stresses and harsh environments. To guarantee user safety and optimise maintenance, it is essential to effectively predict the service life of components under operating conditions. However, these materials are particularly sensitive to environmental conditions such as temperature and humidity. In addition, they are highly anisotropic due to the injection manufacturing process. To characterise the fatigue life of these complex materials within reasonable time and budget, a model is needed. Although the fatigue behavior of such materials has been extensively studied and is gradually being understood, these design models have not yet been deployed for industrial applications. Initial work has led to the establishment of a unified fatigue criterion based on mean strain rate. The latter is applicable to both creep and fatigue (tension/stress) for short durations (less than 10^4s). This direct link highlighted between fatigue and creep opens up a newresearch perspective for a fatigue design approach based on creep studies.
The aim of the presented thesis work is to propose a behavioural model associated with this criterion. To do so, the first step is to deepen the understanding of the link between fatigue and creep, both from a macroscopic perspective (S-N curves) and microstructural mechanisms (fractography, cryofractography and X-ray tomography observations). To explore the limits of this criterion, the research methodology involves a comprehensive experimental setup, encompassing various load ratios and parameters corresponding reflecting real-world operating conditions, ranging from compression-compression to tension-tension. These tests will extend an existing database set up previously and will enable the criterion used to be fully characterised. Finally, this criterion will support an innovative numerical approach to (oligo-cyclic) fatigue design, based on the lessons learned from creep tests. These efforts contribute to a better prediction of the fatigue behaviour of structures.