Effect of dislocation density-twin interactions on twin growth in AZ31 as revealed by explicit crystal plasticity finite element modeling

Abstract

In this work, we employ the recently developed framework for the explicit modeling of discrete twin lamellae within a three-dimensional (3D) crystal plasticity finite element (CPFE) model to examine the effects of dislocation densities in the twin domain on twin thickening. Simulations are carried out for 1¯012101¯1 extension twins in a magnesium AZ31 alloy. The model for the twin lamellae accounts for the crystallographic twin-matrix orientation relationship and characteristic twin shear transformation strain. The calculations for the mechanical fields as a result of twinning consider that one of three types of twin-dislocation density interactions have occurred. One case assumes that the expanding twin retains in its domain the same dislocation density as the parent. The second one considers that twin expansion has lowered the dislocation density as the twin thickens, and the last one, the Basinski effect, assumes that when twin sweeps the region, the dislocation density incorporated in the twin domain is amplified. In the modeling approach, the twin is thickened according to a criterion that maintains the stress state in the vicinity of the grain at a pre-defined characteristic twin resistance. The calculations show that most of the averaged properties, such as the rate of dislocation storage in the entire twin grain, the twin growth rate, the stress field in the twinned grain and neighboring grains, and the slip activity in the parent matrix are not significantly altered by dislocation storage in the twin. The results indicate that, however, the slip activity in the twinned domain is affected. In particular, in the increased dislocation density case, the rate of dislocation density in the twin domain increases at low strains when the twin is first growing from 2% to 5% volume fraction. This initial boost in the dislocation density storage rate causes the newly expanded dislocation twin to contain more stored dislocations than the other cases for all strain levels. Another interesting difference concerns the preference for one or two twins for the same total twin volume fraction; for the increased dislocation twin or twin that retains the dislocation density as it grows, formation of two twins is favored. For a twin that removes dislocation density, only one twin is preferred. The results imply that in the case with reduced dislocation density leads to lower stored dislocations and dislocation storage rates, and lower pyramidal slip activity.