SNIC
SUPR
SNIC SUPR
Atomic scale modeling using cluster expansions
Dnr:

SNIC 2018/5-105

Type:

SNAC Small

Principal Investigator:

Mattias Ångqvist

Affiliation:

Chalmers tekniska högskola

Start Date:

2018-08-16

End Date:

2019-09-01

Primary Classification:

10304: Condensed Matter Physics

Webpage:

Allocation

Abstract

This project will: 1. Distribute and train an ensemble of cluster expansions using a wide variety of fitting methods in the object of procuring the optimal cluster expansion model. 2. Use the constructed models to sample a wide variety of materials with a focus on thermoelectric materials with the goal to ultimately optimize these materials in terms of the thermoelectric conversion factor. 3. Evaluate the potential use of the available GPU nodes at Kebnekaise for sampling the high dimensional integrals associated with thermodynamical integration of atomistic systems. In the end we hope to be able to provide a rationale for finding composition and stoichiometry that optimize materials properties.