Have you heard a split-case pump knock before? This is random and when I listen to it I tend to hear it from the shaft end bearing area. But operations and others in management have heard this knocking at other locations but never in the motor. Vibration readings are low .025 to .07 IPS range across the entire unit. Axial is .02 IPS. Frequencies measured 0 to 20,000 CPM is no higher than .015 IPS. Slow motion study of shaft and coupling did not show any abnormal movement. Valving in the system seems to temporarily change the noise but it comes back later. Shaft end bearing is a roller and unit is VFD. Coupling is an Omega (orange tire).
Motor nameplate speed is 1185 but with VFD it typically operates around 350 to 700 RPM. Lower the RPM the less knocking is heard.

Looking for opinions?

This message has been edited. Last edited by: Don Jones,

Don,

Is the knocking random? You might see it better in the acceleration time waveform, but unless it is periodic that won't tell you much.

What is the likelihood of their being something in the pump?

Pump recently rebuilt. Knocking was heard and a vibration test was requested. Pump opened up and shaft end bearing inspected. Both times it was reported that inside the pump was clean but we know through inspection that it has wear from flow or cavitation from the past. Currently cavitation sound is not present and has not been heard the past few years. The higher the RPM the more prominent the knock.

This message has been edited. Last edited by: Don Jones,

Don, sounds like this is not your problem, but will provide food for thought. I have seen 2 good-sized pumps that were running fine with shafts broken completely in half inside the impeller. In both cases I felt a sporadic knocking. The strobe light diagnosed both situations.

In the first case, the drive end shaft was smooth as could be, but the outboard shaft had a sporadic "twitch" (torsional) when frozen with the strobe light.

In the 2nd case, the drive end shaft was also smooth and steady, but the shaft at the outboard end showed sporadic, axial "shuttling."

In the first instance, they chose not to believe what I was telling them, and it "swarmed" a couple of days later. They shut the 2nd one down right away.

We had a induced draft fan with a slow (1-2 times per sec) knock at the inboard bearing. It was not continuous. Machine 1X is 1780. We found a broken pedestal stud.

Several times at our facility when we have had unusual vibes that are on a VFD it has been the drive that caused the problem. It won't take much to get the electricians check the fault codes.

Don,

Have you verified that knock is in pump and not from coupling? I have encountered incorrect shaft end gap and spool piece on disk-pack couplings to cause knocks.

I like Rusty's comments. I found a "knock" in a gearbox that was due to a broken shaft. A torsional sensor or strobe light may be usefull for torque/torsional issues. A temporary proximity probe for thrust movement may also be usefull. If possible, test pump with different flow conditions, since rotor thrust load can change with flow.

Walt

Try using phase to identify shift from point to point. Across a crack will be noticed.

quote:

we know through inspection that it has wear from flow or cavitation from the past

Are you sure the knock isn't from cavitation? Just not bad enough to be constant. Where is the pump running on the pump curve?

quote:

Slow motion study of shaft and coupling did not show any abnormal movement.

Sorry, I missed that comment. Omega coupling.... have you checked that the hub band bolts are all tight. Also, could the element be torn? If you can't 'see' the coupling (because of the guard) a digital camera with flash will allow you to visually inspect it... I hold it down at the bottom, below the guard and just point it up at the coupling and snap away.

I'd check what Ron said. Look at the pump curve.

Rusty,

It must be nice to chairman of the board. Reaching under a coupling guard can get you in trouble at a number of plants.

Yes, it's nice being your own boss. I used to be a "slave to security" (corporate employee) but I decided life is just too short to live under someone's thumb, with your job - or where you live and work - at the mercy of the next reorganization. So I took a 'package' during yet another reorganization and went out on my own... no equipment, no customers, no prospects really. And it's worked out well.

No I don't reach under the coupling guard where I could possible get my hand into a coupling. Guards are usually open at the bottom, a good 18"-24" away from the coupling. Sticking a camera under there to snap a picture is no more dangerous than standing next to the guard taking vibration readings. Most guards are not going to contain a coupling if it fails at 3600 rpm (or faster).

I bet you stand next to such coupling guards all the time on turbines, compressors, etc. I don't even want to think about what would happen if a turbine were to actually fail while you're standing next to it. And, yes, it happens all the time, doesn't it?

Bill, I appreciate your concern for my safety, but why don't you give it a rest....

Many of the people on the board are not chairman of the board and/or don't own the plants in which they work. Saftey rules must be followed to keep your job, even for most consultants.

My father put his hand on a small heatplant turbine after gathering the measurement equipment since the vibrations was for that time not abnormal and it crumbeled and failed, it was a cast casing so nothing came out but he was not welcome to visit no more as the story goes. Olov

Thanks for all your comments.

Coupling inspected with a strobe. The Omega is a new coupling and all looked acceptable, all though the element did have a minor wobble. Plan is to check the shaft and coupling again with varied loads.

The pump system is odd in that this pump feeds 5 other pumps. Not all operate at the same time but typically 2 and sometimes three operate together. One thought is that this is flow related and it is being investigated.

Thanks again!
Don

Again, I think it is more likely a control issue from the VFD and not a flow problem. Keep us informed as to what you find.

Hey! It's Memorial Day Weekend. Enjoy and relax. I'll do my part.