We have a few horizontal centrifugal pumps running at 3585 RPM and have had more bearing failures than we'd like. The pumps have been acquired over the last couple of years and a few have had bearings and shafts replaced several times.

I suspect the oil rings are not providing enough oil for cooling and lubricating the back-to-back ball thrust bearings (7312 BEGAM). We were getting high copper in the oil analysis but switched to phonelic rings and the copper went away. The rings still appear to be wearing unevenly. We also have heat (190F+ isn't unusual) on the thrust bearing end.

I have a couple of questions.
1) Does anyone else have experience with oil rings not functioning? What were the symptoms and what changes were made?
2) If a ring is acting more like a hoola-hoop on the shaft, has anyone seen vibration FFTs to indicate this? We can't really get a visual on it to say one way or another, but the oil is being churned quite heavily, and it is at the proper level.

Thanks for any assistance.

This message has been edited. Last edited by: Bould173,

I'm very familiar with oil rings on sleeve bearings but I've never heard of using rings with ball bearings. Any chance you have a picture?

You may be able to remove a vent port (mushroom cap) on top of the pump oil reservoir to look in with flashlight and see what the oil ring is doing. It may be that the oil ring is jumping out of the shaft groove that helps keeps it in place axially. We had a horizontal overhung pump (deep groove bearing and back-to-back angle contact bearings) that did this. It caused blackened oil. Our plant worked with the OEM on the problem and I think the resolution included different oil ring dimension and maybe a change to the shaft groove dimension.

I don't remember what we saw on vibration. Will ask on Monday.

Joe,
Check out http://www.pump-zone.com/article.php?articleid=235

There's a diagram down the page a little way.

A cutaway is at http://www.gouldspumps.com/download_files/3700/3700_poster.jpg and you can see them hanging off the shaft (brass rings).

electricpete,
We tried looking through the port, and while oil is indeed flinging everywhere, the rings themselves are not clearly visible. I also expected them to be seen through the ports which is why I suspect they are hoola-hooping.

I can see how changing ring/groove dimensions might help. Which oil ring dimensions did you change? width? diameter? thickness? Same questions for the groove.

Thanks...

From your last post, it would seem that these are Goulds pumps. The first place I'd go is back to the manufacturers. I can't believe that they would put hundreds of pumps into the market with a design defect.
Have you cantacted them? What have they said?

Interesting: the bearings act as a pump by their natural action and can move lubricant from one side's reservoir to the other. High failure rates must be addressed and possible design change if cost justifable. Damming on one side from flow or natural pumping action moving toward the oiler ring could be a possible problem. Opening one side to the others ensuring that the bearing isn't being starved may be a solution. An attached taper slinging/pumping ring could be another to funnel oil to the bearing.

As suggested; contacting the OEM, who may already have addressed this with another could be a direct path to the problem solved.

Please keep us posted.

Heinz Bloch published a good article on lubrication at the TAMU Turbo symposium (attached).

The suitable lubrication method depends in part on D*N number (bearing bore diameter times speed in RPM).

For very low D*N, flooded oil lubrication is OK (somewhere below 1/4 or 1/2 the lowest rolling element).

For higher D*N, flooded lubrication this would cause overheating. Bearings are used above the oil level and we rely on auxiliary device such as oil ring to bring the oil to the bearing.

Heinz Bloch is not particularly fond of oil rings, especially as the D*N number increases.

quote:

"Four factors—shaft horizontality, oil viscosity, depth of immersion in the lubricant, and ring concentricity—have an effect on the operating stability of oil rings. These four factors inevitably vary from pump to pump and day to day.....
"It explains that, unless laboratory conditions are maintained, oil rings tend to become unstable whenever DN exceeds
the region from 6000 to perhaps 8000."

Your D*N number is 2.36" bore (for 7212) times 3600 = 8500. In the range at/above where Heinz says oil rings become unstable.

Heinz' recommendation is to use an alternate type of flinger disk as opposed to oil ring. We haven't done that here.

Whether Heinz is really on the mark about the unsuitability of oil rings in this D*N range, I don't know for sure. I do know we have had a moderate number of problems (particularly darkening oil) on some of our overhung 3600 pumps with back-to-back angle contact bearings plus a single deep groove bearing in an oil reservoir lubricated by rings (I think it's a pretty common design for several OEMs). I know there can certainly be other tricky issues on these machines such as how do you get the right preload on those back-to-back bearings. But I also know that our plant and several others often coordinate with the OEM's for tweaks of the oil ring/groove design, which tells me that (surprise) the original design is not always perfect.

WHOOPS... I had the attachment
"CENTRIFUGAL PUMP COOLING AND LUBRICANT APPLICATION— A “BEST TECHNOLOGY” UPDATE" by Bloch from "PROCEEDINGS OF THE TWENTY-SECOND INTERNATIONAL 93 AL PUMP USERS SYMPOSIUM • 2005"
It is 350kb and the system says I'm limited to 300k. You should be able to find a link to download this for free using google.

Here is the article I was referring to:
http://turbolab.tamu.edu/pubs/Pump22/P22pg092.pdf

Note also the excerpt above did mention just a few items you might check for proper oil ring operation: "horizontality, oil viscosity, depth of immersion in the lubricant, and ring concentricity"

One should do an indepth oil analyses first. Type of failure: skidding, dry, spalling, etc...

As I said earlier; an attached taper flinger acts as a pump and volume can be controled via angle of taper.

In addition: one may want to tie one side to the other depending somewhat or largely on oil analysis. Where do you want to take oil and do you need to incorporate a cooler into the scheme of things?

Just to clarify, the bearings are above the oil level and rely on proper oil ring function for lubrication.

I've read a fair bit of Dr. Heinz Bloch and it actually started me down this path. We are currently pursuing a jet-oil type system to ensure we are getting sufficient oil through the thrust bearings. At the very least this will add filtration and additional bearing cooling since the second thrust bearing currently gets quite hot.

We've done oil analysis extensively. There was a lot of copper before changing to phonelic rings, but if not much oil is going through the bearing, there won't necessarily be wear metals in the oil either. In other words, the oil rings wore but that's all we could tell.

We have the last failed bearings away for analysis but it's taking a painfully long time.

In response to the OEM sending a product with a defect; it happens. I'm not saying this exact product won't work somewhere else under different conditions. I'm saying it's not working for us and there's not a lot of support from the OEM.

We have checked viscosity, horizontality, depth of immersion and ring concentricity and to the best of our measuring they are all within spec. The crew doing the work is top notch and goes above and beyond to make sure things are done to spec and properly the first time. So if everything is correct procedure and spec wise, the pump should achieve its full life once in a while. This leads me to believe there are improvements to the design that should assist and starting with the lube system seems logical.

I'll post back when we get the bearing failure report and see if that confirms the lubrication problem.

Thanks for all the posts and suggestions.

I see that you have checked the viscosity. What is the grade of oil, 32, 68? Some flinger rings we use do wear but not very fast. The rings tend to smooth out with the inside dimension opening up. Oil pick up becomes less at that point. I have seen minor hoola-hooping but is more like rocking back and forth during operation.