Dear all

Below are just some questions for more discussion:

1- using accelerometers, what is the best measurement unit for spectrum (not overall)? Until now, I use velocity regardless the speed of machineries. My justification was that if needed I can integrate and attain displacement. Now I think acceleration is the best unit both for low and high speed machineries. Because I can attain both velocity & displacement if needed. Is my justification correct?

2- What is the best Fmax for machinery with AFB? In a paper, James Berry mentioned that natural frequencies of AFB components occur in 500-2000 Hz and become apparent in second stage of failure, so is it good idea to set Fmax=2000Hz for AFB and regardless of machine speed? Have anybody experience in this issue?

3- For AC motors it is necessary to set Fmax in a way to include slot pass. freq., and also with high resolution for separation of pole pass freq. side bands. this leads to a time consuming data collection, so is this really necessary for a condition monitoring program (not analysis) ?

Regards,F

Dear Farakhani,
1- With most standards and evaluation material based on velocity, you would rarely go wrong with velocity.
2- The best Fmax for any machine is a frequency that is slightly higher than the maximum expected fault frequency. My standard scheme is to take 100XRPM. For almost all machines this is correct and satisfactory. However, there is one exception: with a gearbox I take a frequency that is slightly higher than the 3rd harmonic frequency of the gearmesh with the highest frequency. For a reduction unit, input gearmesh has the highest mesh frequency. For a speed increaser, highest mesh frequency rests on the output gear mesh.
3- For electric motors rarely if ever do you need an Fmax higher than 100XRPM.

Please note how I stick with "orders" rather than CPM or Hertz. As you see it is much eaqsier to work with 100X rather than saying for 1500RPM machines take Fmax as 15000CPM, for 3000RPM motors take Fmax as 30000CPM etc.

Regards
Ibrahim