Dislocation based modelling of engineering alloys

The plastic deformation of metals and alloys is generally governed by the generation and motion of dislocations through the crystal lattice. Microstructural features such as grain boundaries, precipitates and inclusions impede the dislocation motion causing strengthening but often limiting ductility. Dislocations often accumulate in the vicinity of these microstructural features and their mutual interactions and reactions lead to further increased hardening and local hot spots in stress than can in turn lead to failure initiation. Understanding the behaviour of the dislocation ensemble is complex due to the many body interactions that take place. This project will continue development of discrete dislocation plasticity simulations with the particular aim of extending present capabilities from simple single phase, single crystal models to more complex geometries incorporating the grain boundaries and precipitates that are more representative of real engineering alloys.

The project will involve assisting in the development of a 3D discrete dislocation plasticity code. This will require learning Matlab, C and CUDA for acceleration of the computation using parallelisation on graphics processing units. The student will also learn the finite element method (FEM), discrete dislocation dynamics (DDD) and how to couple these together so that FEM captures external boundary conditions while DDD captures the shorter range dislocation-dislocation interactions responsible for the material hardening. Training will be provided to the student in these areas although self-study will also be required. The main tasks are as follows:

1) Accelerating the code through parallelisation on graphics processing units
2) Performing comparisons with available analytic results for simple well-defined sample and dislocation line geometries to check for errors
3) Debugging and development of the code to accurately simulate micromechanical testing
4) Simulating micro-cantilever single crystal bend tests and comparing the model prediction with measured load-displacement curves, the predicted elastic strain fields and lattice rotation fields with high resolution electron backscatter diffraction (HR-EBSD) measurements and the predicted dislocation structure with diffraction contrast observations in the transmission electron microscope (TEM), including 3D dislocation tomography results if available.
5) Incorporating complex microstructure such as precipitates or grain boundaries. We will begin with simple building block configurations of single planar impenetrable grain boundaries in bi-crystals, tri-crystal geometries to investigate triple junctions, and quadruple points at the junction of four grains before moving on to larger groups of grains. We will also consider isolated precipitates and groups of precipitates incorporating them in different ways: as impenetrable to dislocations, as elastically different but penetrable, and with an initial misfit strain to the matrix lattice. This will require an elegant approach to incorporate the geometry. Different possibilities will be investigated such as the level set method. This will also require dislocation remeshing algorithm will also need to be developed to account for dislocations interacting with the geometry.

This project falls into the Engineering theme