Dislocation-density-function dynamics simulation for crystal plasticity : a full-dynamics, all-dislocation approach

Abstract

Current strategies of computational crystal plasticity that focus on individual atoms or dislocations are impractical for real-scale, large-strain problems even with today’s computing power. Dislocation-density based approaches are a way forward but a critical issue to address is a realistic description of the interactions between dislocations. In this thesis, a new scheme for computational dynamics of dislocation-density functions is proposed, which takes full consideration of the mutual elastic interactions between dislocations based on the Hirth-Lothe formulation. Other features considered include (i) continuity nature of the movements of dislocation densities, (ii) forest hardening, (iii) generation according to high spatial gradients in dislocation densities, and (iv) annihilation. Numerical implementation by the finite-volume method, which is well suited for flow problems with high gradients, is discussed. Numerical examples performed for a single-crystal aluminium model show typical strength anisotropy behaviour comparable to experimental observations. Furthermore, this approach has been applied to three engineering problems and discussed in detail: (i) Application on small-scale crystal plasticity successfully captures a number of key experimental features, including power-law relation between strength and size, low dislocation storage and jerky deformation. (ii) Crystal softening and enhanced cell formation are predicted by applying oscillatory loads. The simulations reveal the main mechanism for subcell formation under oscillatory loadings to be the enhanced elimination of statistically stored dislocations by the oscillatory stress, leaving behind geometrically necessary dislocations with low Schmid factors which then form the subgrain walls. This is the first simulation effort to successfully predict the cell formation phenomenon under vibratory loadings. (iii) Tensile deformation of tri-crystals with grain size ranging from 200𝑛𝑚 to 500𝑛𝑚 can be divided into three stages. The results indicate different controlling mechanisms of the flow stress at different stages of deformation and grain sizes. Changing the middle grain tilt angle with respect to the outer grains is found to affect the stress-strain relationship and the distribution of plastic strain in the three grains.

A refined meso-scale scheme based on the full dynamics of dislocation-density functions is also proposed aiming to bridge across the meso scale. In this scheme, the evolution of the dislocation-density functions is derived from a coarse-graining procedure which clearly defines the relationship between the discrete-line and density representations of the dislocation microstructure. Full dynamics of the dislocation-density functions are considered based on an “all-dislocation” concept in which statistically stored dislocations are preserved and treated in the same way as geometrically necessary dislocations. Elastic interactions between dislocations are treated in accordance with Mura’s formula for eigen-stress. Dislocation generation is considered as a consequence of dislocations to maintain their connectivity, and a special scheme is devised for this purpose. The model is applied to simulate a number of intensive microstructures involving discrete dislocation events, including loop expansion and shrinkage under applied and self-stress, dipole annihilation, and Orowan looping. This is the first successful attempt to capture such intensive dislocation microstructures using a simulation scheme that is based on dislocation-density functions. The scheme should also be able to handle high densities of dislocations present in extensive microstructures.