On the development of a machine learned inter-atomic potential and a point defects analysis tool for radiation damage in crystal materials
- Date: Sep 25, 2019
- Time: 15:30 - 16:30
- Speaker: Dr. Javier Domínguez
- Location: Garching
- Room: Seminarraum D3
- Host: IPP
F. J. Domínguez and U. von Toussaint
The analysis of the damage on plasma facing materials, due to its direct interaction
with the plasma environment, is needed to build the next generation of nuclear
machines, where tungsten has been proposed as an excellent candidate for this purpose.
In this talk, I will present the development of an interatomic potential by using
the Gaussian Approximation Process, with the QUIP program, to better model neutron
bombardment processes in pristineWlattices. The potential is trained to reproduce realistic
short-range dynamics, the liquid phase, and the recrystallization process, which
are important for molecular dynamics simulations (MD) of collision cascades. The
formation of point defects and vacancies are quantified and classified by a descriptor
vector based method, which is independent of the sample temperature and its
constituents, requiring modest computational resources. Obtained results are compared
to those for MD simulations by EAM Finnis-Sinclair and Tersoff-ZBL potentials,
at sample temperature of 300 K and a primary knock-on atom (PKA) range of 0.5-10
keV, where a good agreement with the reported number of Frenkel pair is achieved.
This gives the information about the advantages and limits of the machine learned MD
simulations over the standard ones. The formation of dumbbell and crowdion defects
as a function of PKA is discussed.
 F. J. Domínguez and U. von Toussaint, arXiv:1909.00633
 J. Byggmästar et al., arXiv:1908.07330
 F. J. Domínguez, J. Byggmäster et al., in preparation for Modelling Simul. Mater. Sci. Eng.