high vibrations in pd pump, gearbox, motor combo on movable base 1

[jjr1111]

Have a application that flagged high for vibration and unsure how to proceeded. Vibration testing gave the following readings:
motor NDE - 1.494 ips, .11g.
pump - .365 ips, 1.02g

Pump is a PD pump is mounted to baseplate 1" thick on wheels. Gearbox and motor are direct coupled and not mounted to base plate, just overhang, and NDE motor rests on two jack screws to the attachment plate.

 

[Heaviside1925]

jjr111,
I am assuming you are trying to troubleshoot this? I hope I am reading your request correctly, from the description the motor is showing more displacement, which makes sense being that it is not affixed to the plate. The pump is showing the highest acceleration, which to me would point to it being the cause. If it's not a pulsation issue, my next thought is pump bearings.

 

[electricpete]

vibration

As Tmoose would say: What's the frequency Kenneth?

Also is the pattern directional?

NDE motor rests on two jack screws to the attachment plate.

Is it a firm attachment to the motor or as you say "rests". If rests then I'd be inclined to see what happens if jackscrews are backed away.

 

[jjr1111]

Heaviside1925 - When you say if it's not a pulsation issue can you elaborate on what you mean by that. For reference, this pump was recently added to replace a much smaller pump, and it is oversized for the application. This pump sends fluid through a ultra filtration system, essentially a series of back and forth pipes. The pipes shake violently when the system runs. I"m assuming this all stems from the pump/motor vibrations being transmitted through the pipe.

Electricpete - I'll have to go back out and collect spectrum data, I just have some screenshots from a quick analyzer that I can attach for now. Also what do you mean by is the pattern directional?

 

[jjr1111]

Motor vibration attached too

 

[jjr1111]

It is not a firm attachment for motor to base, the motor is just resting on the jack screws. I can try backing them off and see what happens

 

[Heaviside1925]

jjr1111,
Some PD pumps can need pulsation dampeners. Consider water hammer from opening and closing a valve, you are essentially doing a like action each time the pump displaces the liquid. The pump manufacture may have ones specific for their pumps or there are generic ones you can buy and set in situ. If it's been going on ever since the beginning and it's never been addressed, then you might look into it.

Quick edit: Even if there was already a pulsation dampener in the system for the smaller pump, it may not be sized or set correctly for the larger pump.

 

[jjr1111]

Oh, thanks for the info, I learned something new. Talking with others here we don't use pulsation dampers because it's on a sterile system and there's concerns for contamination, so safe to say that the prior application didn't have one either. I'll reach out to the pump MFG and see what's available and see if there are new options that are good for sterile systems. From what I hear the pump shook even when it was in a previous application with larger pipe diameters.

 

[electricpete]

There are a lot of possibilities for causes of high vibration. frequency will help narrow things down. if it is pressure pulsatioin then you'd expect to see pump speed times number of pistons frequency in the vibration. gearmesh frequency is a common source of vibration on gears. having a spectrum would be invaluable.

in the meantime we can try to make some less direct inferences about the frequency:

for one thing, your motor screenshot says "unbalance". ordinally I'd assume that means it suspects mechanical unbalance based on detecting high 1x motor speed. at least that's what i'd think, i don't know anything about the software you're using

for another thing, your motor screen shot says 0.13g / 0.941 ips. IF these magnitudes were caused by a single frequency sinusoid vibration (it is most likely not the case but it'll help us get started) then at what frequency would 0.13g's correspond to 0.941ips (assuming they are both rms or both pk/0). i don't have access to my normal unit conversion tools but using the calculator here if i did the conversion correctly then it looks like those two magnitudes would match each other if the frequency was 9hz (again IF they were single frequency, as a starting point).

[ul]
[li]To even begin interpretting the above swag at frequency (and preferably your spectrum), we'll need more info. what are the speeds of your motor and your gearbox? (and how many pistons in the pump?)[/li]
[/ul]

Also what do you mean by is the pattern directional?

There are ordinarily 3 directions in which vibration is often measured: one direction (called axial) is parallel to the axis of the shaft. the other two directions (called radial directions) are perpendicular to the shaft axis and perpendicular to each other (and preferably aligned along the principle axes of support stiffness). if vibration in one of the directions is much larger than the other that means it's directional. directionality (especially big difference in the two radial directions) is often an indication that the vibration is resonantly amplified.

 

[jjr1111]

Interesting calculator, definately helps show how velocity and acceleration are related to frequency.

The mfg info for this positive displacement pump states it's is a rotary pump using a circumferential piston operating principle. Here's the Link. to the type of pump

I'll have to double check next time on site, but believe the motor is 1800rpm, and we use a frequency drive at run it at 75%, so believe motor running speed is 1350 rpm. Then it goes through a 3.5:1 gearbox reduction so the pump rotates around 385 rpm. When I plug the motor vibration info into the calculator I get 8.49Hz, 509 rpm, So then by that logic, if the vibrations seen in the motor are reflections of pump pulsations the pump would be pulsating 1.3 times every pump revolution, however, I'm not sure I understand how many pulsations to expect per revolution for one of these pumps. The rotors are offset and each rotor has two lobes pockets, so wouldn't it be 4 pulsations per revolution?

 

[jjr1111]

When I put the velocity and acceleration data info into the calculator for the pump (1.02 in/s and .385g) I get 23.19 Hz, which is 1391rpm, and that is close to 4X pump speed so that does make sense that what I'm seeing is pulsations from the pump assuming that you do get 4 pulsations per revolution.

Strange that the frequency is so much higher at the pump and so much lower at the motor. I think the readings from the tester are RMS not peak readings, so maybe that's what's causing the discrepancy with frequency at motor.

I did reach out to a vendor and he has a solution for a dampener to use on a sanitary system, so that could be a good next step

 

[Heaviside1925]

OP,
Based on pump type and the fact the issue follows the pump (so likely not fluid or process driven), I would not suspect pulsation to be the issue (that's a pricy little guy). I took a look through the IOM (I can link it here if you need) and the mounting you describe in your original post gives me pause when comparing it to the IOM. Basically, what I read in the IOM is that the pump and drive both need to be rigidly mounted and shimmed to the base plate for proper alignment with the coupling. It does not sound like the drive and motor are rigidly mounted, just resting on jack screws, which I am guessing was done to perform the alignment in leu of a rigid mount and shims. I other words, it was not installed correctly per the IOM. There are some pretty basic steps the IOM provides for checking the bearings and backlash by hand. I would suspect at a minimum an alignment issue and a possibly damaged pump bearing. A positive note, the tear down and rebuild look pretty straight forward. Hopefully electricpete can provide you some further guidance based on what your frequency calculator outputs show.

 

[jjr1111]

When converting assumed RMS to pk-pk now get the attached, not sure what to make of this. Probalby just need to go back and remeasure and look at the frequency data..

 

[jjr1111]

Thanks for the feedback, I wish I would have taken a picture of the gearbox, it might have been mounted to the plate in addition to the pump. I know for a fact the motor was not mounted and just resting on jack screws, but I'm starting to think maybe the gearbox was mounted to the plate too. I'll be out there next Monday and will be able to get some more clarity there, and will have a read though the IOM in the meantime.

 

[jjr1111]

So had a look at the pump this morning, it is mounted as it calls out in the manual. Both the pump and gearbox are mounted to the plate, the motor is directly mounted to the gearbox. The motor around 1350rpm, and is gear box reduced to around 347rpm. One thing I'm not sure on is the pump has 4" inlet's / outlets, but the process piping is 2". The pump inlets/outlets are reduced down to 2" directly before and after, and I'm not sure how to tell if this is acceptable or is causing issues.

The pump is .521gal/rev of displacement, so at 347 rpm we're around 180 gpm of flow at 60 psi from process. The process fluid resides in a tank, exits the bottom of the tank, 90's, goes a few feet, then is expanded to 4", then to the pump, then 180 degree turn, then reduced to 2", then goes through a filter, then returns to the tank. Viscosity is similar to water. 2" sanitary line so 2.067" ID at 180gpm would be around 17.2 ft/s, which seems fast, I though typical was around 7-10ft/s in piping, but not sure what effects of high velocities are. Perhaps there's a NPSH issue on the inlet where going from 2" ID to 4"ID right before the pump is causing cavitation. I'm not sure how to tell.

 

[BrianE22]

I assume there's a flexible coupling between motor and gear box?

 

[jjr1111]

Pic attached up the setup

 

[Heaviside1925]

OP,
Could it be process related? Maybe. Consider this issue followed the pump in a different service. (I am assuming this, maybe it was the exact same set up). If you are looking at NPSH and cavitations issues, consider the vapor pressure of the process fluid as a driver of that.
You mentioned in an earlier post that the motor was connected to a VFD, so I am assuming you can adjust it. Can you determine if the vibration follows the frequency / rpms? If it does, it's likely mechanical issue.
You could also experiment with different process conditions to see if it affects the vibration for more confirmation.

Back to your original post. motor NDE - 1.494 ips, .11g./ pump - .365 ips, 1.02g. I am not a vibrational analysis expert but seeing these readings indicates to be that the most displacement is showing up in your motor but with low acceleration, whereas the highest acceleration is at the pump, which I would also assume has the most mass. My thought would be that the pump has the highest force in the system. In any spring mass dampener system, i.e. vibration, there is a forcing function. So, if I follow the forces, that leads me to the pump being the issue. I could be totally wrong but this is what I thought when I read your original post. Another thought is your pump has 4 lobes and as you noted your vibrational frequency aligns with 4x the rpm. The check for slack and bearing condition can be done by hand per the IOM. Just consider trying to eliminate this from the possible causes before going down too many other rabbit holes.

 

[jjr1111]

Hey Heaviside, operating the pump at full speed gave the initial velocity and acceleration reading I posted, but it also runs at reduced speeds for short periods of time, when at these reduced speeds, the vibration is significantly reduced. I was able to capture pump vibration and acceleration data during one of those events today as well as velocity and acceleration spectrum data at 100%. Here is the data for at 100%, I'll look to get the reduced speed data posted soon.

 

[jjr1111]

Acceleration