you mean because of missing material we have high vibration? but we should doing field balanc...
When motors are manufactured, the rotor is balanced with the missing material in the keyseat area replaced by a "dummy" keyway. Often the weight of this missing material is very carefully calculated and the amount used is carefully matched.
When you test a motor in a solo condition you have to replace the missing material. This is often done by using the final fitted key and coupling hub to hold it in place. If you don't do this the motor is essentially out of balance but if it passes through the 1st flexural mode (a possibility with a large motor) then it may enter a region of speed where it "self" balances. The behavior, without the dummy" key, is high vibration up to the critical, then the rotor may operate at a lower level after passing through the critical. You need to identify where the flexural mode may be; this is the reason why I asked if you can acquire shutdown transient data, something like a polar plot or bode plot.
You should examine some past threads that have similar discussion:
Another good resource is at https://www.industry.usa.sieme...r-vibration-prob.pdf. Check the paragraph on page 7 of 16, left column, specifically the paragraph just above the words "Thermal Unbalance".