Bearing Cap Oil Leak 2

[todcaesar]

I have a nuisance oil leak that is getting a lot of attention from management since it has been deemed a safety issue.

The oil is mostly coming up the four bolt holes on the bearing cap. We usually go out and retorque while online (nuclear power plant) and then replace the gasket during outages- but the leaks keep coming back. We have tried sealants with a limited amount of success.

Any thoughts? Perhaps a different gasket material?

Pump:
Ingersoll Rand Boiler Feed Pump
14X17CN-1

Oil: DTE 32 Turboil

Gasket Material:
GARLOCK STYLE 3000 SYNTHETIC FIBERS, FILLER AND ELASTOMERIC BINDER, 1/64 inch

Thanks, Tod Welsh 803 345-4584

 

[GPRnD]

I assume it is not in a high radiation area, hence elastomer disintegration is not a factor ?

You didn't say what type of bearings were on the IR pump. This might allow me to zero in a little better on the design and possible causes.

On intital review, the gasket material choice seems good. That grade 3000 is well proven and commonly used. I would think about the following:

Using a thicker gasket (1/32)to improve the relative spring rates and sealing as the bolted joint loads and unloads

Check the overall flatness of the housing and cap.

Consider using Spiralock fasteners or Loctite threadlock compound if loosening of the bolts under vibration is suspected.

Consider using a gasket material with a lower creep relaxation such as Garlock Gylon 3510


If you can provide some more background I can probably be more specific.

 

[electricpete]

"getting a lot of attention from management"
LOL.

If it seems to be coming around the bolts, how about a thread sealant?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[todcaesar]

This plant is a PWR which means the secondary is a non-rad area.

The bearings are split sleeve bearings.

Any other info that you need?

Thanks, Tod

 

[GPRnD]

Split sleeve bearings are harder to seal at the end caps because you have a 2 way joint in that area.

What are you using the seal the split halves of the bearing housing to each other ?

 

[BobM3]

You say you are retorquing the screws? Are they coming loose (and then losing preload on the gasket)?

Do you think the new gasket stops the leak for awhile only because it is acting like a sponge and absorbing the oil?

I've had luck replacing gaskets entirely with Loctite Gasket Eliminator 515. Make sure the thickness of the gasket isn't needed for proper fitting though.

 

[JJPellin]

I am still not sure I understand which joint is giving you trouble. I would refer to the bearing cap as the top cover on the housing. On the thrust end, there could also be an end cover which may have originally had a gasket. I assume that the sleeve bearing is pressure lubricated with a circulating oil system. If this is the configuration, as noted above, the point where those two gaskets cross at a right angle is really tough to seal.

First, I have very rarely ever seen a bearing cap that used a gasket like this. All of ours that I can think of have a metal-to-metal joint that is sealed with some sort of sealer (RTV, Permatex, etc.). The end covers usually have gaskets. Sometimes these gaskets also serve as the shim to adjust the end float in the Kingsbury-type thrust bearing. We generally eliminate this gasket by machining the appropriate amount off of the nose of the cover.

We have a few compressors that tend to have chronic leaks coming along the bolts on the bearing housing top cap. In all cases, it is because the split-line is pressurized by the lubrication system. The oil is fed into the lower half housing but pressurizes an annulus that passes all the way around the radial bearing. Where that pressurized annulus crosses the split line, it is very difficult to seal. The oil eventually works its way into the split line and the first path it finds is often a bolt hole. But, our solution does not work with a gasketed joint. We have a very meticulous procedure for sealing the joint. We clean it and prime it with Acetone primer. Then we seal it with two-part catalyst activated RTV. A single oily thumb print on either surface will provide a leak path and the leak will come back.

I have seen one other instances of a chronic leak along a bolt hole that was a result of machining error / casting defect. The bottom of one bolt hole was breaking through by way of a casting inclusion into a pressurized oil galley that feeds the radial and thrust bearings. This would be believable for a single bolt. But it is unlikely if all four bolts are leaking.

If this is the case of two gaskets crossing at right angles, then you may have fewer options. You probably can’t eliminate the gasket since that sets the crush on the bearing. This also makes it difficult to change the gasket thickness. The only suggestion I would have then is to use a very precise procedure for cutting the horizontal gasket. Leave the gasket long and only make the final trim after the top cover is installed and torqued. Trim it off with a razor blade and apply a sealant (two-part RTVwould be a good choice) to the gasket on the end cover. Even if your procedures calls for these gaskets to be installed dry, a thin film of non-hardening sealant may give you the edge you need to avoid the leak.


Johnny Pellin

 

[todcaesar]

Brad, we will sometimes(usually not) use permatex, but not as a rule.

BobM3, Maint. reports that during the retorquing there is some movement noted. I think there is some loosening of the bolts goin on here.

I am not sure how to explain the temporary nature of the fix.

It seems that since the oil is making it to the bolts but not making it to the outside splitline, that it might be more flatness related. The shop has used a stone on the caps in the past. But this would not guarantee that both sides of the cap are in the same plane.

Perhaps shaft pressure is causing the inside of the bearing cap splitline to separate.If this is the case would a thicker gasket help?

 

[JJPellin]

You make an excellent point. If the thickness of the gasket determines the crush on the bearings, then the gasket thickness needs to be very nearly perfect. If the gasket were too thin, the crush on the bearing would be excessive and this could allow the split line to be held open near the bore, but closed tight at the outside. That is why we don't normally use a gasket in this joint. It is much easier to achieve a precise crush if this is a metal to metal joint. If the particular bearing design is not sensitive to crush, then you may be able to make a machining cut to reduce the crush, allowing the split line to stay closed at the bore. Some bearings require a minimum crush to achieve proper clearance. Other bearings are less dependant. If the bearing halves are bolted together, then your bearing design may not require crush. Check the crush and make sure it is not excessive relative to the manufacturer’s specification.

If your mechanics have been stoning these surfaces, they may have stoned off the cross-hatch pattern that helps secure and seal the gasket. You don’t want these surfaces to be smooth. A stone should only be used very lightly and very sparingly on a gasketed joint.


Johnny Pellin

 

[GPRnD]

In general a thickner gasket helps because it gives better spring rates in the bolted joint assembly and it is more tolerant of imperfect surfaces or flatness errors.

I'd certainly suspect the 2 way joint as being a problem.

It is common that people will fit the main split line gasket so the end of the gasket is flush with the machined surface where the bearing cap mates. This is not good practice because the gasket is never exactly flush after tightening.

The correct way is to leave excess gasket sticking out, tighten the main split line joint bolting, then using a razor blade or fine tooth file, trim the protruding gasket flush with the machined surface. At the same time you need to verify that the two halves don't have an axial mismatch with each other that forms a step.

Steps are obviously v bad when it comes to achieving a goos seal. The only way to rectify this is to remachine the surface. However you need to verify that the alignment dowels (or whatever other device is used to align the two halves), are not worn and sloppy.


If the end cap gasket is not used to set bearing float, I would go with a 1/32 gasket and for good measure use the RTV suggested by JJPellin

 

[GPRnD]

JJPellin, good point about the stoning perhaps making the surface too smooth.

todcaesar, you need to aim for surface finish between preferrably 125 to 250 micro inches for the gasket surface to give it something to bite into.