Observation and modeling of strain gradients in AA6016 – Influence of length-scale, microstructure, and strain path

Abstract

Structural aluminum alloys are often less-than ideal materials for studying sub-grain strain gradients via EBSD, at typical resolution settings. Sharply defined slip bands are not generally observed due to cross-slip, and second-phase particles formed during solidification of work-hardened alloys provide obstacles that disrupt potential structure development, leading to what can seem like random distributions of geometrically necessary dislocations (GNDs). This study considers the roles of length-scale and second-phase particles in sub-grain distributions of AA6016-T4 following deformation. Second-phase particles are shown to play a stronger role than grain boundaries (GBs) in local GND accumulations. The net Burgers vector is used to show the transition from crystallographic-level slip to macro-scale slip as length scale increases, with a corresponding transition in the GND vs. step size graph. A strain gradient crystal plasticity model is applied to assess predictability of the observations. Real 3D structures were extracted, via serial sectioning, following application of different strain paths. Predicted GND and total dislocation evolution closely follows observed values. The model is then used to study the relative contributions of GBs and second-phase particles to GND localization, leading to the conclusion that second-phase particles must be included in the model to reflect observed behavior.