Physically based modelling and simulation of the mechanical behaviour of metallic thin film systems and fine grained surfaces under cyclic loading - FASS

Project summary

In this project we aim at physically based modelling and simulation supported by quantified characterisation of the mechanical behaviour of microsamples (polycrystalline metallic thin film systems and micropillars) under cyclic loads. Investigation of the physical mechanisms of fatigue in these systems is motivated by the following considerations: (i) Early stages of fatigue failure in bulk systems (inception and stage-I propagation of fatigue cracks) are governed by near-surface phenomena and can be strongly influenced by surface treatment. To understand fatigue threshold and initial crack-propagation, it is necessary to model the interplay of surfaces and interfaces (grain boundaries, GBs) with fatigue-induced dislocation patterns and cracks. (ii) Samples on the micrometer scale are amenable to full simulation of microstructural processes. This allows us to directly validate models by comparing with ‘tailored’ experiments. At the same time, larger samples in this scale range already exhibit bulk-like behaviour. (iii) Fatigue is a multiscale phenomenon involving processes from the atomic to the continuum scale but, despite its huge technological importance, has rarely been addressed from a comprehensive multiscale modelling point of view. In simulations, the physics of fatigue still poses important challenges as material behaviour is governed by slip localisation and dislocation patterning phenomena which cannot be predicted by standard continuum or atomistic approaches. Present-day discrete dislocation dynamics (DDD) for the first time provides a physically based model of emergent dislocation patterning, but cannot access the large cumulative strains associated with failure under cyclic loads. We overcome this limitation by exploiting recently developed coarse-graining methods that map DDD onto continuum dislocation dynamics (CDD) simulations which represent the same dynamics in a continuum framework. These can be calibrated and validated by reference to the DDD models but extend to larger spatial and temporal scales. Plasticity simulations are combined with atomistic simulations that give access to dislocation nucleation at surfaces and interfaces and crack nucleation at surface heterogeneities or stress concentrations. The ultimate goal of the project is to provide physical foundations for compu-tational design of fatigue resistant near-surface microstructures. To this end we develop a predictive multiscale framework for the evolution of dislocation systems interacting under cyclic loads with surfaces, cracks and GBs, and evaluate damage and failure properties. Simulations will be parameterized and validated by reference to experiments carried out on model systems, using techniques which provide information on microstructural processes on scales from nanometers up to several micrometers. Investigated systems include polycrystalline thin films and single/bicrystal micropillars of varying size and grain orientation. Cyclic plasticity of these systems will be studied using tensile and compression testing. Performing selected experiments in situ in a scanning electron microscope will facilitate direct observation of deformation and failure patterns. On even smaller scales, in-situ transmission electron microscopy will directly visualise dislocation microstructure evolution under load, dislocation nucleation and interactions at the interfaces. The proposed effort will result in a simulation framework for modelling the physical processes behind a vast range of technological problems including the enhancement of fatigue resistance by surface treatment, and fatigue of microscale components. While work on any of these specific problems is beyond the scope of the project, we aim at establishing and expanding our contacts with potential ‘users’ at an early stage in order to facilitate future application and dissemination of our research findings.