I've got a 150HP AC Variable motor with .8+ips in the horizontal, .07 vertical and .09 axial 3x dominant (1586cpm 529 motor rpm)The high peak is only there at this speed which matches 1000m/min for the machine. When they slow to 950, the amplitude at 3x drops to .179ips showing one additional harmonic at about 25% of the first peak. I suspected the coupling because it's a 3 jaw Rotex with the spider element but they tell me it's OK. The resolver coupling is good. Any thoughts? My thought was that the coupling halves are jamming into each other.

Since it's directional and very speed dependent, maybe resonance in the horizontal direction at that frequency?

Ditto, I agree with Pete... Directional, speed dependent.
I am assuming the Harmonic you speak of at 25% (amplitude?) is 6x at .04475" ampl, which seems inconsequential compared to the main event (.8"@ 3x)

Since this is variable speed, you can check for the Phase shift through the resonance while it is running...

Is this something new that has just shown up or has it been there a while?

Ralph, it just showed up. Resonance? Why at 3xrpm? I would expect resonance to be at 1x. Tell me if I'm missing something with that. Thanks.

I think resonance can be triggered by any frequency force if the resonace frequency is at or near that frequency, like your 1586 cpm area.

With a 3 jawed coupling, I would be looking for misalignment. But would not rule out an impact test and a search for something that could have "broken" or come loose to give it a resonance frequency that had never been there before.

Thanks. I'll have a closer look at the support stand and the floor bolts.

Ron,

How is the motor mounted? If it is a sugar scoop type mount (where the motor support is bolted to the front of the gearbox), they are very prone to that kind of behavior and will often switch directions with bracing. Yours sounds like it might be too big for that type of drive.

I would look around with a strobe set to just above or below 1586 and run my finger along all the joints (welds, bolted steel connections, bearin feet, grout) feeling for looseness. Be careful!

Good Luck,

Danny

You are right that high vibration from resonance requires both a machine natural frequency and some exciting force at that frequency. It might be that the exciting force is the 3-jaw coupling. Even though the alignment is good and insert is in good condition, the resonance may magnify the effect of any small misalignment. Just like a machine resonant at 1x is sensitive to even small levels of unbalance, a machine resonant at 3x with a 3-jaw coupling would be sensitive to even small levels of misalignment.

If you have a peak standing out at 6x, I would say coupling is very likely involved in some way (3x/6x/9x pattern). I think that's what you said above when you referred to "one additional harmonic"?

If neither 6x nor 9x stands out above their neighbors, and you have only series of harmonics of running speed with 3x the highest and nothign special at 6x/9x, then looseness might be another possible exciting force to consider that could be exciting a 3x resonance as well.

Thanks for your help so far. I have carefully checked all the mounting bolts and joints and can find nothing loose. The motor sits on a reinforced table above the floor. The table itself is not vibrating very much. I have attached a file with some data I just took at 2 different speeds. It sure looks like resonance because the peak is so clean. Maybe it's a drive control issue?

5WDRREELMTR.doc (114 Kb, 40 downloads) Motor Data 2 Different Speeds

When you are operating at 950 m/minute with 3x peak of 0.2 ips at frequency 1510, there is some raised noise floor to the right of that peak.

When you are operating at 1000 m/minute with 3x peak of 0.9 ips at frequency 1584. the raised noise floor is centered around that peak.

Both of the above tend to confirm resonance in the neighborhood of 1584 cpm. That resonance is amplifying broadband noise present within your system to give the raised noise floor in vicinity of the resonant frequency at 1584.

==============
Sidebands spaced at 45cpm away from 3x with magnitude 0.04 ips at the lower speed. At the higher speed when the 3x peak increase, it doesn’t look like the magnitude of the sidebands doesn’t increase. I don’t know what those sidebands are. You could check for pole pass I guess (can’t tell what your slip is on a vfd). Also perhaps there are other frequencies in this machine based on the nature of the load? Whatever they are, they are probably being amplified by the resonance as well.

Take a look at the attached data. I captured this as the machine sped up. I think I'll do a waterfall next time. (tomorrow) there is more going on here than I think. I got them to go as high as 1050 but there was no change in the 3x except it got higher peaking at .89ips.

5WDRREELMTR1.doc (66 Kb, 33 downloads)

The new information in your most recent file:
1 - You sped up the machine to push 3x up to 1667 and magnitude of that peak came down. That part makes sense and is Cconsistent with resonance at ~ 1584.

2 - Vibration stands out at 3x, 9x, 15x (first, third, fifth harmonics of 3x forcing function). I think this further confirms the forcing function is at 3x and most likely associated with the coupling or the alignment.


3 – There are raised noise floor around 3x, 9x, 15x. This contradicts my suggestion that the raised noise floor and sidebands might have been an effect of the resonance. The raised noise floor and/or sidebands around 9x and 15x can’t be blamed on simple resonance and have some other cause... which I don’t know. (Any suggestions?)

I would say a bump test while secured or attempt temporary bracing while running to confirm resonance. If you can get rid of that high vibration 0.8 ips through bracing, you will be improving machine reliability. If not, further investigation is required. I'm not sure if any further action is required to address those sidebands.

Are the sidebands 45 cpm no matter what speed you check the 3x at?

I got a thought. If this is a VF drive, what is the base speed of the AC motor? If it is 1200 rpm, then the frequency of the drive is ~26.45 Hz. That is a little over 1587 rpm. Could the VF drive be the forcing function exciting a natural frequency?

Gary B

Pete & Gary,

Most VFD's have a frequency and RPM meter on them, if they arent't working try an online motor tester, that'll give it to you, also.

Good point. We can't tell what the slip is, but Ron can.

Checking whether these are pole pass frequencies would be a good step for further investigation. If they are, current signature analysis would be a good next step to look for pole pass sidebands around line frequency.

The base speed is 575rpm. The 45 appears to be consistent regardless of speed.