Added
- Writing of loadcase defined by a velocity gradient.
Changed
- Allow passing deformation gradient (rate) as a nested list to `writers.write_load_case`.
- Updates for use with latest alpha version of damask (v3a3) - writing yaml load file and vtr geom file.
- Updates to parsing damask simulation results back to a volume element response dict - added new data specification types:
- volume_data - Data from a given field over entire VE with transformations applied (to partly replace incremental_data)
- phase_data - Data from a given field over a single phase of the VE with transformations applied (to partly replace incremental_data)
- field_data - Data from a given field in correct order/shape as the VE dimensions. Can also output a grain and phase map.
- grain_data - Data from a given field averaged over each grain in the VE
All use a consistent definition of the increments to extract from, based on the increment number in the simulation.
Fixed
- Add boolean option `use_max_precision` to orientations dict. If `True`, in `utils.validate_orientations`, cast quaternions to an array of type `np.longdouble` prior to normalisation, and then write out these quaternions (in `writers.write_material` via a new function `utils.prepare_material_yaml_data`) to the largest precision supported by `np.longdouble` on that machine, to ensure maximum precision, and hopefully avoid the DAMASK error: "invalid orientation specified".
- Fix issue where `write_geom` would raise if volume element `size` or `origin` is an array rather than a list.
- Fix issue [15](https://github.com/LightForm-group/damask-parse/issues/15)