CAN WE AVOID BEATS IN TWO NEAR BY SIMILAR PUMPS MOUNTED ON COMMON CIVIL STRUCTURE , PUMPING FLUID FROM COMMON SUMP? WE CAN'T PHYSICLLY ISOLATE THE MOUNTINGS.

You have two options to try,
Reduce vibration of one or both by precision balancing, tighter tolerance alignments, etc. Your vibration spectra identifies the predominant vibration (I don't know if it is 1x, since you didn't specify) and do what is necessary to greatly reduce that component. If the high vibration is vane pass frequency, then you will have to evaluate your piping and pump curves.
The other option is to put the pumps on variable frequency drives and move their speeds apart.

quote:

Originally posted by Ron Brook:
You have two options to try,
Reduce vibration of one or both by precision balancing, tighter tolerance alignments, etc. Your vibration spectra identifies the predominant vibration (I don't know if it is 1x, since you didn't specify) and do what is necessary to greatly reduce that component. If the high vibration is vane pass frequency, then you will have to evaluate your piping and pump curves.
The other option is to put the pumps on variable frequency drives and move their speeds apart.

BROOK, we have already done balancing and alignment part. since both the pumps are discharging in common header, while running in parallel , can we go for different speeds. I think, no.

Akhtar,
You didn't say what the predominant vibration frequency is. This is very important for the purpose of alleviating the vibration levels causing the beat. There is no reason why two pumps couldn't operate at different speeds just because they continue into a common discharge header.

Can you clarify more why you consider the beats to be a problem?

Are the beats bothering people?

I don't think the equipment effects of adding vibration at two different frequencies far apart is any different than the effects of adding two frequencies close together (beating), except that when close together, the low beating frequency is evident to humans.

One thing to consider if the machines share vibration through their bases is whether false brinneling might occur during a period when one pump is off and the other running. Realistically, I personally probably wouldn't worry about that until if/when I saw bearing faults, and then consider it carefully during the failure analysis. The defect pattern from false brinneling is easy to recognize when failed bearings are cut open for inspection.

This message has been edited. Last edited by: electricpete,

AKHTAR,

To answer your question directly, the only way to get rid of beating or any other background vibration transmission is reducing this transmission by mechanical isolation. If this is that important (see electricpete's post) then you can go into this trouble, but it is feasable.

David

quote:

Originally posted by electricpete:
Can you clarify more why you consider the beats to be a problem?

Are the beats bothering people?

I don't think the equipment effects of adding vibration at two different frequencies far apart is any different than the effects of adding two frequencies close together (beating), except that when close together, the low beating frequency is evident to humans.

One thing to consider if the machines share vibration through their bases is whether false brinneling might occur during a period when one pump is off and the other running. Realistically, I personally probably wouldn't worry about that until if/when I saw bearing faults, and then consider it carefully during the failure analysis. The defect pattern from false brinneling is easy to recognize when failed bearings are cut open for inspection.

PETE,
Is gradually increasing and decreasing vibration amplitude not a cause for concern? Vibration is varying from 2.5 mm/s to 8 mm/s (RMS). Apart from false brinelling , it may also cause true brinelling and also cause damage to pump bearings .

Akhtar,

Whether vibration change is gradual or not has no significance. Neither has closeness of frequencies. What does have significance is whether or not additinal vibration due do background vibration causes significant damage. If pumps had worked independently, their respective vibration would have been around 2.75 mm/sec and 5.25 mm/sec - not that much lower then 8 mm/sec.

Secondly, brinnelling occuring due to vibration transmission regardless of severity is by definition called "false brinelling".

I agree with David. Vibration that slowly varies between 2mm/sec and 8 mm/sec is certainly no more damaging than vibration continuously at 8 mm/sec.

If we assume that the vibration comes from two closely spaced peaks with magnitude 5 mm/sec rms and 3 mm/sec rms (which are alternately in-phase and out-of-phase to cause variation between 2mm/sec and 8 mm/sec), then the total rms value is sqrt(5^2+3^2)~ 6mm/sec. Then your vibration varying from 3mm/sec to 8mm/sec is equivalent on an rms basis to a steady vibration at 6 mm/sec.

Also, although it's difficult to quantify, I would suspect in most cases the bearing loading associated with a given amount of vibration on a machine bearing housing originating from an outside source is probably less than the bearing loading associated with the same amount of vibration originating from within the machine itself. (assuming the machine is running.... again false brinneling is a concern when the machine is subject to vibration while secured).

This message has been edited. Last edited by: electricpete,

quote:

Originally posted by David_G:
Akhtar,

Whether vibration change is gradual or not has no significance. Neither has closeness of frequencies. What does have significance is whether or not additinal vibration due do background vibration causes significant damage. If pumps had worked independently, their respective vibration would have been around 2.75 mm/sec and 5.25 mm/sec - not that much lower then 8 mm/sec.

Secondly, brinnelling occuring due to vibration transmission regardless of severity is by definition called "false brinelling".

DAVID
Brinelling on a stationary equipment bearings due to vibration of nearby running equipments is by definition called false brinelling. But brinelling in bearings due to actual loading during running (due to its own as well as near by running equipment vibration)is called true brinelling.

quote:

Originally posted by Ron Brook:
Akhtar,
You didn't say what the predominant vibration frequency is. This is very important for the purpose of alleviating the vibration levels causing the beat. There is no reason why two pumps couldn't operate at different speeds just because they continue into a common discharge header.

Brook
Two pumps can run in parallel only if their shut off head is same otherwise both pumps will not be sharing same load.Also there is danger of one pump running continuously under shut off
and subsequent damage. If we go for different speeds, we will have to maintain same shut off head. This will be possible only by change of impeller size.This means a total new installation and scraping the old one. Clearly not a very fancy idea.

quote:

Originally posted by AKHTAR:

DAVID
Brinelling on a stationary equipment bearings due to vibration of nearby running equipments is by definition called false brinelling. But brinelling in bearings due to actual loading during running (due to its own as well as near by running equipment vibration)is called true brinelling.

Achtar,

The point I was trying to make is that when external vibration is being transmitted into the idle equipment (false brinelling )it is more damagingto the bearing then that when it is running in respect to the external vibration since unlike the former there is a lubricating film when it is running.

David