We use computational and experimental tools across length scales to examine complex phenomena such as plasticity and hydrogen embrittlement, with data-driven methods to keep the calculations affordable.
Rhenium alloying improves the plasticity of tungsten, yet the dislocation-scale mechanisms responsible for Re-induced softening remain unclear. To address this, dislocation-based plasticity in W–Re alloys is examined using a multiscale computational framework.
Hydrogen absorbed from the environment changes how dislocations move, and that is why a ductile steel stops being ductile. This work tracks the effect from first principles up to a crystal plasticity model, to say what a given hydrogen concentration does to plasticity in polycrystalline α-Fe.
An atomistic simulation is only as good as its interatomic potential: the function that says what each atom feels from the atoms around it. Get it wrong and every prediction built on it is wrong too. Potentials are fitted here for W–H and Fe–C–H, the systems behind plasma-facing tungsten and ferrous structural materials.
Hydrogen has to move through a steel before it can embrittle it, and how fast it moves depends on the phases it meets, the stress in them and the shape of the microstructure: contributions an experiment cannot separate. This work leverages a microstructurally informed, coupled diffusion–mechanics finite element model to separate them.
Hydrogen collects at the boundaries between grains, and that is where cracks start. Not every boundary is equally vulnerable. A large set of boundaries in α-Fe was screened to find the ones hydrogen stays away from, and simulation used to measure how much harder a boundary holding hydrogen is for a dislocation to cross.
Two samples of the same alloy do not deform identically: the microstructure varies, and so does the local strain it produces. Quantifying that variability properly takes a dataset far larger than anyone can afford to compute, so a low-rank approximation is fitted to a small one and used to predict the full local strain field.
Corrosion, hydrogen and wear all reach a component through the same place, its surface. Modifying that surface is the common idea here, carried across alloy systems: cathodic plasma electrolytic nitriding (c-PEN) of a low-alloy ferritic steel, and a cerium conversion coating on a magnesium alloy.
If the electrochemical response of an alloy can be predicted from first principles, corrosion becomes something to design against rather than something to measure afterwards. The approach is applied here to second phases in magnesium alloys, which form micro-galvanic cells with the matrix, and to alloying additions in ferrous alloys in acidic media.
A component in service carries load and corrodes at the same time, and neither process leaves the other alone: stress alters the electrochemistry of the surface, and metal lost to corrosion alters the stress. The two are solved together in a single framework, across a wide range of mechanical and electrochemical conditions.
An alloy is strengthened by the precipitates a dislocation has to get past, and there are two ways past: cutting through, or bowing around and leaving a loop behind. Atomistic simulation of Al–Cu resolves which happens for both Al2Cu phases, and finds that for θ′ it is the character of the dislocation, edge or screw, that decides.
Openings in the group
A grounding in solid mechanics and a willingness to learn are the prerequisites.
