As of now the user has to manually toggle on the "Perform strain rosette recovery during dynamics simulation" for the FE part owing that strain rosette. This is easily forgotten, and instead we get an mis-leading error message from solver that the strain rosette does not exist. This is fine if the strain rosette results are only used for plotting, but if they are also used as function arguments, the recovery must be enabled for the model to be consistent.
As of now the user has to manually toggle on the "Perform strain rosette recovery during dynamics simulation" for the FE part owing that strain rosette. This is easily forgotten, and instead we get an mis-leading error message from solver that the strain rosette does not exist. This is fine if the strain rosette results are only used for plotting, but if they are also used as function arguments, the recovery must be enabled for the model to be consistent.