LightForm-group/UoM-CSF-matflow

Ability to use predefined .geom and material.yaml files with matflow-damask

Opened this issue · 4 comments

In order to perform replicate simulations of dual phase microstructure experimental loading, the ability to use predefined damask input files in a matflow-damask workflow requires implementation

Note:

  • Newest damask version will use .vtr file format for geometry files.
  • Damask-parse already contains methods to do this

Task name: generate_volume_element ( already used here; task_schema l:335 and here; func mappers l:43 )
Method name: from_damask_input_files
Implementation name: damask

Inputs

  • geometry_filepath
  • material_filepath

Function steps

  1. Read input files.
  2. Write a volume_element dict that can be passed to a simulate_volume_element_loading task. (already functionality to do these individually in damask-parse)

Output

  • volume_element dict

Implementation

  • function mapper as part of the generate_volume_element task.

Thanks Guy. Might as well throw the option to use an existing load case file in there as well!

EDIT: actually never mind about the load case file for now, that is easy enough to include directly in the simulate_volume_element_loading task.

  • Task_schema for this in commit 53f660a to UoM-CSF-matflow branch: master
  • func_mapper for this in commit ffe6322 to matflow-damask branch: ve_from_input_files

Thanks Guy. Might as well throw the option to use an existing load case file in there as well!

EDIT: actually never mind about the load case file for now, that is easy enough to include directly in the simulate_volume_element_loading task.

No I do think it would be better to keep separate as a task so as to keep the workflow as modular as possible? Unless you mean defining a filepath to the .load file in the simulate_volume_element_loading task?

If you want to add a generate_load_case task with method from_DAMASK_load_file, that would be nice, but it's not vital at the moment. We would need to add a new function damask_parse.readers.read_load to parse DAMASK's new YAML load file format.