[haitao-ZCU]

• Replace the design variable f with w = 2*pi*f in motor2's output variable formulas (dIA_dt--dIE_dt) and excitation formulas (I_A--I_E), and in oper_params. f collides with Ansys's reserved intrinsic variable name and gets misread when used inside output variable expressions.
• computation.py's create_setup() still hardcodes Nper/f and 1/(f*(PointPer-1)) for StopTime/TimeStep, shared with motor1. Extracted these into stop_time_expr()/time_step_expr() so motor1 keeps the original f-based formulas unchanged, while motor2 overrides them with w-based ones.
• Verified that w reads back correctly through an output variable (equal to 2*pi*50), and that dIA_dt/Torque results are unchanged compared to a previous test where f was renamed to fr everywhere.
• Updated test_design2.py to match the new oper_params.

[sadda]

@haitao-ZCU Please resolve the conflicts created by the merged PR:29. Basically, just copy the changes from motor2.py and test_design2.py into the new files.

[sadda]

@haitao-ZCU We do not want to use the default expression "Nper/f" because it contains f, which will be disallowed in a subsequent PR. I think we should remove the definition of the children stop_time_expr and time_step_expr and move them to the parent.

[haitao-ZCU]

@sadda
Agreed.

I have moved stop_time_expr/time_step_expr into the parent class as suggested, and it uses w now.
I also verified this modification does not change the results for motor1, I ran the old f based formula and the new w based one side by side across 3 structures (60 points total) and compared TorAvg and TorRippleRms. The results are same between f and w. I have attached the csv.
f_vs_w_motor1.csv