13th International Conference on Fracture June 16–21, 2013, Beijing, China -1- Analysis of the multiaxial fatigue strength at the mesoscopic scale using 3D microstructure modeling and extreme value statistics Anis Hor1,*, Nicolas Saintier1, Camille Robert2, Thierry Palin-Luc1, Franck Morel2 1 Arts et Métiers ParisTech, I2M - CNRS, Esplanade des Arts et Métiers 33405 Talence, France 2 Arts et Métiers ParisTech, LAMPA, 2 bd du Ronceray 49035 Angers, France * Corresponding author: anis.hor@ensam.eu Abstract Fatigue life computing methods are generally based on putting into equation the mechanical quantities calculated at the micro or meso scale, the relevance of these selected quantities being validated by the capacity of the models to reproduce experimental results at the macroscopic scale. Although the scaling of the damage mechanisms involved in fatigue crack initiation processes are relatively well identified (grain scale, slip bands), their explicit consideration in fatigue criteria is still not well-developed. Furthermore, the existing methods do not consider the microstructure-sensitivity. The aim of this paper is to present the computational strategies developed to account for the microstructure-sensitivity in the calculation of fatigue strength. This work is based on three parts: (1) the development of 3D microstructure modeling tools (2) the analysis of the dispersion induced by the microstructure heterogeneities on the critical fatigue damage indicators and (3) the development of a statistical approach which provides a framework for analyzing calculation results in the HCF (High Cycle Fatigue) regime. In this context, a method based on the construction of statistical extreme value distributions from FEA calculation results was developed. The evolution of the scaling parameters of these distributions for different loading conditions is related to the effect of non-proportional loading and microstructure. A design method based on these extreme value statistics is presented to obtain a new mesoscopic criterion sensitive to microstructure parameters. Finally, surface effects are discussed. Keywords HCF, crystal plasticity, extreme value probability, FE simulation. Abbreviation and designation HCF: High Cycle Fatigue FIP: Fatigue Indicator Parameter RVE: Representative Volume Element SVE: Statistical Volume Element GEV: Generalized Extreme Value distribution Microscopic length scale: corresponding to the integration points Mesoscopic length scale: Corresponding to the average density in a grain Macroscopic length scale: Corresponding to the elementary volume average 1. Introduction In literature, methods for determining the fatigue behavior based on multiscale modeling estimate that the fatigue strength of metals depends on the extreme value statistics of a single microstructure
RkJQdWJsaXNoZXIy MjM0NDE=