You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
CosseratRodIO currently does not support multiple rod IO. The user instead has to create an IO object for each rod in order to properly render each rod as a rod in Paraview, with separate xmf files for each rod. A preliminary workaround was to append all the nodes of Cosserat rods into a single array, and then add the array as lagrangian field into base IO, altogether bypassing the need for CosseratRodIO. This is slightly undesirable since the rendering in Paraview will render the rods as points and not as polylines.
Instead, the add_as_lagrangian_fields_for_io() function in the base IO class should handle multiple rods (i.e multiple grids) just like how add_as_eulerian_fields_for_io() handles multiple fields (albeit on the same grid). CosseratRodIO can then leverage that and add rods to the IO as needed, resulting in only one file for paraview rendering.
I will push an updated patch of this implementation in sopht-mpi (taking care of the parallel hdf5 writing). This should provide a better idea on how we can refactor and merge MPIIO to the base IO in sopht as planned in #146
The text was updated successfully, but these errors were encountered:
CosseratRodIO
currently does not support multiple rod IO. The user instead has to create an IO object for each rod in order to properly render each rod as a rod in Paraview, with separate xmf files for each rod. A preliminary workaround was to append all the nodes of Cosserat rods into a single array, and then add the array as lagrangian field into baseIO
, altogether bypassing the need forCosseratRodIO
. This is slightly undesirable since the rendering in Paraview will render the rods as points and not as polylines.Instead, the
add_as_lagrangian_fields_for_io()
function in the baseIO
class should handle multiple rods (i.e multiple grids) just like howadd_as_eulerian_fields_for_io()
handles multiple fields (albeit on the same grid).CosseratRodIO
can then leverage that and add rods to the IO as needed, resulting in only one file for paraview rendering.I will push an updated patch of this implementation in
sopht-mpi
(taking care of the parallel hdf5 writing). This should provide a better idea on how we can refactor and mergeMPIIO
to the baseIO
insopht
as planned in #146The text was updated successfully, but these errors were encountered: