[SY-E11] Symposium E-11

[SY-E11] A computational infrastructure for multiscale materials simulation

^○Xiaoyu Yang (Computer network information center of the Chinese Academy of Sciences, China)

Bridging materials models and passing materials-related data and information across length scales is critical for the quantitative and predictive modelling of materials development. However, lacking of acceptable linkage software and tools is an obstacle. It is recommended in the report by the Minerals, Metals & Materials Society that “establishing an infrastructure for multiscale materials data is the first programmatic recommendations”. Currently we are developing such an infrastructure to address this need based on a high-throughput computational infrastructure MatCloud[1]. The core of this multiscale simulation infrastructure is a linking software tool and technologies that contains a workflow designer and a workflow engine. The workflow designer aims to support the design of simulations involving running quantum mechanics scale ABINIT simulation and molecular scale LAMMPS simulation by drag & drop approach. The workflow engine is responsible for parsing and processing the workflow of multiscale materials simulation created by the workflow designer. As a test case, the multiscale materials simulation linking tool will be used to develop a particular interatomic potential through data-driven approach. In this talk, we will report our progress and challenges encountered.

Reference:

[1]. Yang X. et al , MatCloud: A High-throughput Computational Infrastructure for Integrated Management of Materials Simulation, Data and Resources, Computational Materials Science, Volume 146, 15 April 2018, Pages 319-333, https://doi.org/10.1016/j.commatsci.2018.01.039