Mechanical, fracture and termodynamic modelling of reactor materials
Abstract
The purpose of this project is to (i) investigate the impact of impurities and transmutation products on the mechanical and fracture mechanical properties fusion reactor materials and (ii) produce phonon data and generate thermal scattering laws for fission reactor materials and fuel, e.g. UHx and LiH.
The first project is a collaboration with the ICAMS-group (at Ruhr-University, Bochum) in which we aim to use classical and quantum mechanical modelling to probe the effects of transmutation elements on the grain boundary strength of tungsten alloys. Owing to the fact that empirical potentials for tungsten (W) are notoriously difficult to generate, with severely limiting transferability and predictability, our aim is to generate a machine learning potential for W within the atomic cluster expansion (ACE) formalism, which can be expanded to the ternary W-Re-Os and binary W-He systems to enable classical modelling with close to ab initio accuracy. This requires the systematic establishment of a large database comprising DFT data that covers a large range of low to high coordination configurations and accounts for different local bond orders. The potentials will be used (i) to explore the role of precipitation hardening on the brittle to ductile transition of W while in operation and (ii) to model the fomation of helium bubbles and their impact on the grain boundary cohesion. Such data will be generated to train a scale-bridging machine-learning surrogate that describes the mechanical properties of interfaces and can be used in continuum modelling to investigate the fracture of grain boundaries.
For the final part of the project, we will compute the phonon spectra of UHx and LiH. This serves two purposes: (i) to support experimental activities associated to devising a new method to use ultra-cold neutrons for characterizing the hydrogen content in hydride phases and (ii) to generate thermal scattering laws (TSLs) for neutron interaction. The experimental activities will be performed under the supervision of Dr. Takeyasu Ito at the Neutron Science Center at Los Alamos Science Lab, while the researchers in Lund and Malmö will support the experimental activities with first-principles DFT data regarding phonon and thermodynamic properties. Owing to their relevance in nuclear reactor applications, we are particularly interested in probing the vibrational properties hydrides (UHx and LiH). Because of the difficulty to reproduce experimental hydrogen phonon spectra using conventional force-constant methods, we need to use ab initio molecular dynamics (AIMD) modelling to probe the dynamics of the system. Specifically for LiH, due to the low mass, we most likely have to resort to path integral approaches, which are expected to be numerically costly. For postprocessing purposes we will rely mainly on the TDEP package, which enables the extraction of finite temperature phonon data, including anharmonic effects, and has proven reliable to extract hydrogen phonon density of states. For the TSL generation, we have established a collaboration with another research group at Los Alamos and a group at Idaho National Laboratory, which we will support with vibrational data.
