agarb said:
Stuff like:
- For line voltage of 80-90% of nominal for a maximum of 20 seconds, verify that we still have sufficient torque.
- Current & torque ripple measurements
- speed vs. torque curves
- speed loop test
What exactly are you trying to achieve with a dyno test?
Nearly any standard asynchronous AC motor is capable of delivering 2.5 times its nominal torque, but that is of course without a VFD and if current can rise accordingly.
But in general:
Torque (in Nm) = Power (in kW) x 9656 / speed (in rpm)
To deliver the same amount of torque at a 20% voltage drop, current should rise 20%. (I'm not taking speed variation into account.) So you could overcome the torque issue by raising the nominal motorcurrent setting by 20%, because VFDs have the annoying habit of lowering the motor's speed first when the motor is being overloaded (just to keep motor current below the settings). 20 seconds seems very short in motor terms, but just how repetitive are these voltage drops? If the motor is already running near full load, you can not expect it to overcome 20% voltage drops every 5 minutes!
Another issue, just how critical is speed to your application? Also, dyno tests tell you nothing about load inertia. What type of load is the motor driving?
If speed drop could become a problem, check if the VFD is (sensorless) vector control capable. Using vector control, instead of standard U/f control, will improve speed stability, even under overload or undervoltage conditions.
However, chances are the VFD will trip due to a 20% voltage drop...
Now I'm assuming the motor/VFD is powered by a generator ('cause which right-in-his-mind net provider would allow a 20% drop)... Isn't overvoltage going to be something to worry about? I've seen too many VFDs trip because of a voltage overshoot (due to a generator's reaction to sudden load variations).