We have a fan pump that is driven by two motors. About two months ago a vibration showed up at the inner race frequency on the pump. We changed the bearings and put back in service. Five weeks later the bearings failed again. Now anytime the pump is run above 500 rpm there seems to be a beat pattern every 45 seconds.

We took a five minute wave form and verified the time. If we drop below 500 rpm the vibration levels drop dramatically. The motors are driven by 4 vfd's. One vfd is not strong enough to pull a motor so 2 vfds are required for each motor. There are tachs on each motor (or so I am told).

I suspect one motor is not pulling its share of the load. It seems if it were a mechanical problem there would not be a dramatic drop off at 500 rpm.

Any ideas on which way I should jump from here?

Aubrey

Aubrey
Not sure I understand the connection between two bearing failures and beat pattern.
Do you suspect some sort of cause/effect here?

"Usually" beat patterns are not problematic.

Was the second failure similar to first ?? Same bearing location ?

Perhaps some more information would help us help you

Regards
Jim Powers

I'm trying to figure out why you would use two VFD's for a motor. Why not size the motor and VFD together?
You mentioned you thought there was 2 tachs on each motor. I would asssume they were for feedback to the VFD comparitor board.
Depending on where the problem is (bearing, or beat), I would suspect the 4 VFD's would have trouble staying in exact sync to drive the motors. You could very well get a beat freq out of that setup.

WHat bearing failed? What was the failure? Was the iner race chewed up (spalled), or did you just get the BPFI?

Dave

vibr8

I think the "beat" every 45 seconds is causing the bearing failures.

Dave

I was told that one vfd would not run a motor this size. They had to have two vfds per motor. This is from our motor foreman so I assume he knows what he is talking about. Initially the west bearing on the pump failed, then it failed again. Inner and outer race looked eat up. Todays reading seems to indicate the east inboard motor bearing may be failing. There is one tach per motor.

Aubrey

What size motor? I have not seen a motor in which you cannot size a drive.

Parallel drives are not a good idea with a motor. I suspect that you will find, if you have the bearings analyzed, that this is a classic shaft current problem.

If so, there are a number of solutions including insulating the bearings and coupling and/or an Aegis shaft brush.

In the meantime, I think I would investigate why anyone would have designed the system that way. Too unstable and difficult to control if you are using it for any precision.

Howard

quote:

Originally posted by MotorDoc:
Aubrey

Parallel drives are not a good idea with a motor.

In the meantime, I think I would investigate why anyone would have designed the system that way. Too unstable and difficult to control if you are using it for any precision.

Howard

Hear Hear
Dave

quote:

vibr8

I think the "beat" every 45 seconds is causing the bearing failures.

Aubrey
You sound fairly convinced that the "beat" is the cause of the bearing failures. I have never heard of that before. R U sure its not coincidental ??
I suggest you look a little harder at other possible causes.
Regards
Jim P

motordoc,
Many large horsepower applications require dual drives. Not uncommon at all.
I do wish you would stop telling people that insulated bearings and Aegis brushes will prevent EDM. They do NOT!
The true prevention is circumferencially shielded inverter duty cable.
Insulated bearings are required to prevent low frequency circulating currents in motors above a 445 frame, due to the possible potential difference from end to end in the windings, but the Aegis brushes, as tested by numerous motor manufacturers are failures.
Why would anyone want to add a maintenance item (is the brush working this week?) to a motor, if the power cable eliminates the need for them?
Europe has already changed their code. You cannot put a motor and drive in without the proper power cable. Not the US. So, to this day, we have marketing leading customers, who don't know any better, to buy the motor and drive and wire it with only a concern for 60 Hz grounding.
Your response is welcome.
Ron Brook

Aubrey,

I have to agree with MotorDoc that feeding a motor with two parallel VFD drives, although may not be uncommon, is a bad idea. ( Recall effects of connecting two DC batteries in parallel. ) If you have a chance, take motor current draw readings ( use hi res) and you are likely going to see beating at 45 sec period due to slightly different frequency output. You may also see higher current distortion that ultimately leads to accelerated bearing etching. There is a possibility that you would've had bearing etching problem even with just one drive as well.

Dave

This is an interesting thread. We still haven't heard more detail about the size of the motors. It's tough to waste any more time on this subject until we hear more about the installation, unless we just want to argue about side subjects.

Dual parallel motor/drive combination such as this would have to have a lower reliability factor. We definately have more VFD failures than motor failures. Also, a high percentage of the motor failures we have, which aren't very many, are driven by a VFD and were installed before VFD rated motors became the norm.

I have to agree that we need to adopt VFD cable standards in the US.

J-

I have heard of connecting the DC rectifier sections of drives in parallel. Could this be the case? Are you sure the outputs are being syncronized for parallel operation?

J-

Sorry guys about not posting back earlier. Right now I am having to get caught back up on our routes and reports and such. When I get a breath I will answer questions as I hunt out the answers.

BTW, the motors are only 250 horse. To the PdM team we dont understand why they would use two small motors on one pump.

I hate it when work gets in the way of fun!!

Thanks,

Aubrey

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

[QUOTE]
BTW, the motors are only 250 horse. To the PdM team we dont understand why they would use two small motors on one pump.


And to think that all this time reading this tread I was thinking 4-8K HP. Who in their right mind would ever say a 250 hp motor needed two VFD's? This is really starting to look like shaft currents. If you have access to the bearings, can you please post some pictures?

(Shaft brushes for a fix?) Thanks I needed a good laugh this morning. I have personally installed these on motors and witnessed a bearing failure in 36 hours.

Here is the transient data for the fan pump.

This data was collected on the inboard bearing of the east motor. This point has the highest readings.

I still haven't found the bearings.

Aubrey

fan_pump_tran.doc (36 Kb, 59 downloads) Pump Transient

Aubrey,

The TWF does not look to me as beating of two close frequencies. In the case of true beating amplitude transition is is more smooth.

I'd recommend taking transient electric current data.

Also, generate spectra from the existing TWF in its various regions.

Dave

Aubrey,

When a high vibration event occurs, it looks to have uniformly spaced waveform spikes. What do the the waveforms and spectra look like during these high vibration events? Simultaneous current analysis of each motor and shaft current measurements may help to indicate bearing fault cause.

Walt

"We have a fan pump that is driven by two motors." (How long have you monitored this machine, been trouble free until bearing change out?)

"About two months ago a vibration showed up at the inner race frequency on the pump."
(Did you take a look at the bearing to see if it was inner race fault)

"We changed the bearings and put back in service. Five weeks later the bearings failed again." (Did you catch the bearing with vibe data before it failed or did it truly fail? Any machine work, balancing, alignment, etc?)

"Now anytime the pump is run above 500 rpm there seems to be a beat pattern every 45 seconds" (I am assuming this beat started only after the bearings were changed and it has never been there before?)

It's still a machine that rotates. The data can be analyzed and narrowed down to possible problems.

"If we drop below 500 rpm the vibration levels drop dramatically." (What drops, all axis of measurement? On which component fan, pump, motor 1, motor 2?)

"Initially the west bearing on the pump failed, then it failed again. Inner and outer race looked eat up."(Can you explain? If the inner race or the outer race DO NOT look like a new bearing out of the box, you should be measuring both surfaces with a micrometer to verify proper fits)

Todays reading seems to indicate the east inboard motor bearing may be failing. (Any update on this?)

A beat frequency, in my experience, is typically created by mechanical looseness. I have yet to walk up to a machine that is making a beating noise and not find mechanical loosenes in vibration spectrum. Do you have any vibration or phase data to share with us?

Ron Brook,

Does the same approach to grounding apply to dc motors?

Danny,
No. DC machines, once they are of a large frame size, may have low frequency circulating currents from one end to the other and insulating one bearing will solve this issue. Be careful. If a non-insulated tach is installed, or non-insulated bearing RTD's, the ground path will be re-estabished.

I just looked at the time waveforms!!! If that is truly 35 to 40 seconds between the excessive vibration spikes, this has got to be process related, or an issue with the load sharing on the drives. I still don't understand this setup. Why would anyone design a pump with only 250 hp and require two drives?