[haitao-ZCU]

• currently, we use solutions.data_real(expr, convert_to_SI=True) to extract the values from Ansys, this can help us obtain the correct vars quickly instead of manually hard-coding and guessing unit scaling factors (a big mess).
• corresponding tests are also updated, and all passed, including test_ansys_regression_synrm_5f_40s.py.

[sadda]

Good, that makes it much simpler. I would add something like

@staticmethod
def extract_expression(solutions, expr: str) -> np.ndarray:
    return np.asarray(solutions.data_real(expr, convert_to_SI=True))

to ComputationBase so that we do not forget the conversion to SI units by accident later. Please verify that this method is always called and that there are no direct calls to data_real.

[sadda]

Also, this duplicates #54, which should be closed.

[haitao-ZCU]

@sadda
good idea. I have added it and replaced data_real() with extract_expression(), and all tests passed.