Motor vibration 3

How can differentiate between electrical and mechanical frequency component in  a motor spectra?

 

[brin]

I recommend contacting the Vibration Institute for training courses or literature.&nbsp;&nbsp;Also many of the major vendors such as Entek have progressive series of training seminars which would help you out.<br><br>Back to your question though-&nbsp;&nbsp;you would have to give us a bit more info on the application.&nbsp;&nbsp;Diagnosis of motor electrical vibration can vary depending&nbsp;&nbsp;on what kind of motor we are talking about, eg:&nbsp;&nbsp;DC, induction motor, synchronous motor, etc.<br><br>

 

[Hatch]

I dont' know your setup, but one tool is to monitor the vibration spectra in real time and shut the power off to the motor.&nbsp;&nbsp;At the instant of the shutdown, the electrical components usually disappear, assuming they were present or are visible in the first place.<br>

 

[Franko]

If motor is a 3600 rpm induction motor, zoom with a spectrum analyzer to see if the frequency is at exactly at 3600 cpm (or 7200 cpm for 2x). If frequencies are slightly below, a mechanical problem is indicated. Also, contact CSI or SKF for help with motor current spectrum analysis to check for broken rotor bars, etc.

 

[ArneLindholm]

If you need something real basic introductory with pictures rather than formulas, I could mail you our basic .ppt pictures. Some english text, but mostly picture says more than words.<br>You can contact me at <A HREF="mailto:arnespamlind@algonet.se">arnespamlind@algonet.se</A>. Remove spam to get the address right.

 

[thierrybelge]

Some electrical vibrations are easy to detect, they will be in the neighbourhood of the electrical netfrequency or its following in the mathematical row.F.E. 50 Hz net look at 50,100,150,...
Mechanical frequenties will be more in the neighbourhood of the eigenfrequencies of the mechanical structure. Those can be calculate by a finite element program or in rough by hand.

 

[Franko]

Be careful in some countries where line frequency fluctuates. Recently had another case in India where line frequency varied between 49 and 51 Hz.

 

I have several Dc motors that i monitor on a regular basis. Three of them have 360 hertz with several harmonics which I suspect this is caused from the DC circuit SCR's . I need additional information on what causes the DC circuit 360 hertz to be produced in the spectrum and what changes to cause the 360 hertz to increase in amplitude ????

jameswhanna@yahoo.com

 

[jil01]

The Finite Element can be time consuming. If you are equiped with accelrometers and impulse hammer and can perform a modal test, you can perform the following:
1. With the motor off: obtain the frequency response function (Accerleration/Force),with accelerometer mounted on the shaft and impacting the motor housing. This should provide the structural (mechanical) frequencies of the motor.

2. You can also get the power spectrum (narrow band frequency data) when you turn the power on with a holding torque (Acceleration response over a selected frequency range). your frequency maximum domain should be selected larger that the power cycle, >120 Hz or > 400Hz. 1600Hz will show for example a 400Hz and its harmonics of 800Hz.
Here you might get more acceleration response data by turning the shaft every 45 degrees fro 0 to 360, and the response reflect inner mechanical characteristics of the coil/armature/magnetic field interaction. Electrical frequency should be that of the power cycle.

3. The third motor behaviour you would like to get would be with motor running (steady state speed), and that would give you the operational modal characteristics. Here, bearing frequency should show clearly.

By no means the task is easy and requires good repeatable data sampling of an isolated motor (to avoid external interaction with motor support).

I hope this can be help to you.

 

[brin]

Just a quick note on the SCR vibration- this is a typical characteristic of DC motors, and worthy of comparing against baselines (try to ensure same load). I've seen considerable fluctuations in amplitude and number of harmonics, etc. for ski lift drives. As a consultant, I recommended checking the rectification system, but have yet to get any worthwile feedback on faults, etc.

Anybody have good case histories on SCR faults/diagnosis?

 

[ArneLindholm]

A 60 Hz current has two force peaks or maxima per period. You have 120 such force humps per second. With three phases, there are three rows of peaks nicely split over each second. That sums up to 360 forces per second. Yields 360 Hz. In Europe we suffer from 300 Hz. Some rectifiers have a filter and the noise is smaller. I suggest that you measure the DC-current and the AC-current component and ask you electrician if all is ok with the rectifier. A ripple like 2 - 5 percent is not unusual. Above 10 percent? - well somebody pays the bill - all AC on a DC-motor is a lot of heat for nothing. Best regards Arne