Project: Development of complex workflows for the simulation of materials

Our simulations of materials properties are mostly performed using our advanced all-electron density functional theory codes. In particular, we rely on our FLAPW simulation package FLEUR and the KKR-code. While these are very powerful, well established proven simulation codes, their use typically requires many steps with user intervention and the simulation of complex properties is thus difficult and error prone. Hence, we currently develop advanced high-level workflows and turn-key solutions enabling:

·      The use of the codes by non-expert users.

·      The automatic calculation of materials properties in a HTC screening approach.

·      The automatic generation of meta-data.

·      The creation of interfaces for data-analytics, searching of data.

·      The full tracking of the provenance of the calculations.

These workflows are implemented in the AiiDA framework.

So far, we implemented as open source python packages:

•       AiiDA-FLEUR plugin: Connects the FLEUR code family to the AiiDA framework

•       Source code: https://github.com/JuDFTteam/aiida-fleur

•       Online documentation: https://aiida-fleur.readthedocs.io/en/develop/index.html

•       AiiDA-KKR plugin:Connects the JuKKR code family to the AiiDA framework

•       Source code: https://github.com/JuDFTteam/aiida-kkr

•       Online documentation: https://aiida-kkr.readthedocs.io/en/stable/

Our workflows typically produce provenance graphs  as shown below.