The objective of this paper is to show that a number of key features of the Dang Van
high cycle fatigue criterion can be observed using simple polycrystalline
computational models.
This paper presents a series of numerical computations for an inclusion consisting
of 156 grains embedded in a homogeneous matrix. The grains are modeled using a
polycrystalline single slip elastoplastic model, whilst the matrix is considered as
elastic. As expected the numerical simulations confirm the theoretical prediction on
which the Dang Van fatigue criterion is based, that if a large enough number of
grains is considered under uniform loading, a grain with the least favourable lattice
orientation will always be present. This grain will constitute the weakest link in the
assembly and thus its fatigue life largely determines the fatigue life of the bulk
material.
Next the question of stress-gradients in the high cycle fatigue regime is
addressed. An example of stress gradients appears around notches as they create
stress concentrations in structures. It is a well known problem that fatigue
criteria have to be locally arranged using stress-factors or critical distances
in order to give satisfactory predictions. The work presented here shows
that an analysis of the problem at the grain scale explains the apparent
discrepancy when using classical fatigue criteria. The discussion is based on a
numerical model of single slip crystal plasticity and the Dang Van fatigue
criterion.
Keywords
high cycle fatigue, fatigue criterion, polycrystalline
plasticity, stress gradient, notch