Beating Phenomenon at Plant

Hi

i believe most of you have been experience dealing with beating vibration at your plant
How do you guys plan/ and control it
any maintenance strategy..

How about the vibration database and trending
Since the amplitude can be as high as reaching alarm..

Thanks for sharing

regard,

miz@l

Original Post
If your beat frequency is in a motor, you can verify or rule out soft foot by loosening one motor foot hold down bolt on the run and see if the beat ceases. Tighten foot again then repeat procedure on another foot. If you find the foot that is soft you can put shim under to fix. Shim required can be determined by measuring gap under the foot with feeler gauge.
I think you may be asking how to get around the problem of a beat or fluctuating vibration causing basic overall alarm threshholds to be breached and creating lots of alarms? It is always recommended I think that vibration amplitude is recorded at the highest amplitude. Easy to say. I have tons of machines whose vibration fluctuates alot due to process, vfd or other variables and it is a problem for sure. Generally the beat or fluctation occurs at frequency so band alarms can help but requires a lot of set up and order based data collection works best
At our facility, we have a pump noise monitor (a simple overal vibration meter) that monitors vibration on critical coolant pumps on a daily basis.
This system is the first line of defense for these machines.
At certain pump combinations, a beat phenomenon occurs due to the modal interaction between two pumps on a common loop. The beat is ALWAYS there....however, it is not always noticed by the watchstander since the beat can be a few seconds long (and noted by the watchstander)or it can be over several minutes long and, as far as the watchstander/log taker is concerned, non existent.
This occurs because the pump's speed (split bus driving AC induction motors) is not constant.

The point is, the only way to properly set the baseline vibration limits is:
1) realize that a beat frequency exists and
2) look at data over a long period of time, convince yourself there is no degrading condition and then statistically set the upper limit.

Hope this helps
Jim P
I can confirm the beating time waveform modulation pattern and by using high resolution on the spectrum to separate the peak of interest

So you guys are using either band alarm on the identical peak or statistical alarm.

I have a thought that the high vibration from the beating might causing any secondary damage.
Maybe it is better to avoid any vibration transmit between each machine,
The best is to have individual base for each machine or strengthen the base perhaps.

how about adding additional damper to the machine.Would that works?
Quote,
"I have a thought that the high vibration from the beating might causing any secondary damage."

No
By itself, beating is not a problem.

In the case I mentioned, the beating phenomenon merely caused problems with our overall alarm setting. The only way we could deal with it was to statistically evaluate the highest levels and set our baseline appropriately. The beating exists, it causes NO DAMAGE and there is nothing else to do except adjust our alarm baselines to account for it.

Quote
"Maybe it is better to avoid any vibration transmit between each machine,
The best is to have individual base for each machine or strengthen the base perhaps."

Perhaps I misunderstand your concern, but why would you bother to go to that extreme? Other than the fact that the beating is ocurring, it is not causing secondary problems.

Regards
Jim P

Add Reply

Likes (0)
×
×
×
×