Hie;
We have two ID fans in our plant that run parallel to each other. Recently we had a breakdown on one of them and inspected both bearings. On the DE, the bearing was still alright but on the NDE, the inner race was pitted. We then decided to replace bothe of them. The following is some data with regards to the bearings and housings.

DE
Bearing size: 22228C3
Bearing outer diameter: 250.00mm
Bearing housing bore: 250.04mm
Shaft diameter: 140.06mm
bearing manufacturer: FAG

DE
Bearing size: 22220C3
Bearing outer diameter: 180.00mm
Bearing housing bore: 180.06mm
Shaft diameter: 140.03mm
bearing manufacturer: FAG

It was proposed to use shims so as to avoid bearing turning inside the housings. Now my quetions are;
1.Do we need to put some shims on these bearings?
2.If yes, what are the recommended sizes of shims to be used?
3. If no, why and what are the consequences of shimming rolling element bearings?

Are these pillow block bearings? Are you asking if you shim the bearings internally or externally?

Yes they are pillow block bearings. I am talking about externally.

quote:

1.Do we need to put some shims on these bearings?

Check your OEM data, but I don't think so.

What housing fit classification are you shooting for? For split housings, SKF Bearing Maintenance Handbook gives choices J7, H7, H8, G7 depending on conditions.

The tightest of these is J7.
For 250mm OD bearing, J7 is 0.016mm tight to 0.060mm loose. You are in spec for J7 as well as the others. I cannot imagine any possible reason to shim this or sleeve the housings.

For 180 OD beairng J7 is 0.014mm tight to 0.051mm loose. You are just a tad too loose at 0.060 mm loose. But I doubt you really need J7. J7 is for direction of load indeterminate. For the normal case of stationary load, you can go looser to H7 which would be 0-065mm loose for 180mm OD bearing (you meet it). Or if there is heat conduction from shaft through bearing, you might even go looser to G7 (and you will still meet it).

quote:

what are the consequences of shimming rolling element bearings?

I don't think you need shims (see above), but if you use them I don't think they are a reliable long-term fix. The bearing demands a smooth firm seat. I don't think you can get that with a shim. You will likely have bumps at the edges or overlaps or seams. That flexes the outer ring as the elements roll over it and will lead to premature failure. That's why sleeving the housing is the preferred standard fix IF you decide you need a tigher fit (and I'm not convinced you need a tighter fit)

By the way, was there anything to suggest the failure was related to a loose housing? The link between inner ring pits and loose housing is not clear in my mind. Was there severe fretting of the outer ring?

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

In the attached pdf, you see SKF recommendations for housing fits (J7, H7, H8, G7)

The tables defining these fits are shown here:
http://www.skf.com/portal/skf/home/products?maincatalog...g=en&newlink=1_0_180

SKF_SplitOrSolidHousingFit.pdf (29 Kb, 18 downloads)

quote:

if you use them [SHIMS]I don't think they are a reliable long-term fix. The bearing demands a smooth firm seat.

I should temper this comment by mentioning that I work with motors that only have full cylindrical seat housings for rolling element bearings. I am not very familiar with pillow blocks. Shimming may be more acceptable for pillow blocks, but I'm not sure.

Was there evidence that the outer race had been turning in the housing? If not, then why shim anyway. Also, one of these bearings should be allowed to float in the housing. If you shim the floating bearing, you in a sense turn it into a held bearing. If excess clearance is suspected, you are better off replacing the housing instead of shimming. I agree that shimming a bearing is only a temporary get by.
Also, eneven shimming can result in a "pinched" bearing which will egg the outer race of the bearing, thus causing premature failure.

(Yes they are pillow block bearings. I am talking about externally.)

What was the race pitting from?
Contamination, Weak bearing support, Bearing base not flat or level.

We have shimmed pillow block bearings on emission control fans with no problems. Over 12 years of continuous operation. Shimming the entire base (not just the bolted ends) is necessary in order to support the block properly. The base should be level and flat so shimming will be somewhat equal all the way around. If your shimming because the base is significantly lower on one side than the other, then the base needs corrected.

Something to think about.

The term "external" seems a little ambigous to me in this case.

I was under the impression that the shimming to be considered is between the bearing and housing, and not under the entire pillowblock (otherwise why did he give us bearing and housing measurmenets and mention shimming as alternative to sleeving)

Maybe jingas can clarify

Good question electricpete! This statement would make me think that it was internal shimming too.

( It was proposed to use shims so as to avoid bearing turning inside the housings.)

That is why I asked the question of internal or external shimming.

I wonder if "external" meant between the ourter ring and the housing, and "internal" meant between the shaft and the bore of the bearing.

I agree that there is no reason at all to put shims in these bearings.

I have a fan with pillow block bearings and shimmed on the bearing outer surface.

Its in smooth operation for 2 years.

quote:

Originally posted by yahoo:
I have a fan with pillow block bearings and shimmed on the bearing outer surface.

Its in smooth operation for 2 years.

Just because it worked this time doesn't make it good mechanical practice...just expedience. Show us the vibration signature in two more years...

If you're referring to the bearing housing's foot or feet then shimming is common and a correct procedure when base problems are encountered.

Only shim pillow block bearings under the feet, if you are shimming the bore of the pillow block, only do it temporary until the pillow block housing can be replaced, you are asking for trouble by shimming the bore, misalignment, outer race will not be supported well, cocking of the bearing. All these are reasons not to shim.

Thanks guys for the contributions.
By external I meant between the outer race and the housing. Another issue that I can also mention is that on the previous bearing, what was observed as a failure was pitting on the outer race of the bearing. Again, it was shimmed so can we say that the failure was due to shimming?

The pitting of the outer race OD is a symptom of looseness, replace the housing, it is too loose.

quote:

Originally posted by Jim F:
The pitting of the outer race OD is a symptom of looseness, replace the housing, it is too loose.

Jingas
Instead of changing the bearing housing you can scrap the parting planes faces split bearing housing . The best way to decide , how much to scrap the parting planes is to place shims of 0.5mm thickness (for example)on parting planes. Then place lead wire over outer race of bearing. Now put top half bearing housing on the bottom half.Tighten the bolts.Now
Calculate x=0.5(i.e. shim thickness)-residual lead wire thickness. It should be around 0.010 to 0.020 mm(i.e. interference between bearing outer race and housing).If it turns out to be negative, this means , the bearing is loose in housing.Scrap the parting plane by x+0.01. to 0.020 mm to achieve desired interference. This is a common practice followed in industry and is called nip adjustment.

quote:

Originally posted by AKHTAR:

quote:

Originally posted by Jim F:
The pitting of the outer race OD is a symptom of looseness, replace the housing, it is too loose.

Jingas
Instead of changing the bearing housing you can scrap the parting planes faces split bearing housing . The best way to decide , how much to scrap the parting planes is to place shims of 0.5mm thickness (for example)on parting planes. Then place lead wire over outer race of bearing. Now put top half bearing housing on the bottom half.Tighten the bolts.Now
Calculate x=0.5(i.e. shim thickness)-residual lead wire thickness. It should be around 0.010 to 0.020 mm(i.e. interference between bearing outer race and housing).If it turns out to be negative, this means , the bearing is loose in housing.Scrap the parting plane by x+0.01. to 0.020 mm to achieve desired interference. This is a common practice followed in industry and is called nip adjustment.

This would cause the housing to become out of round and I would only use this as a temporary fix until the housing can be replaced with a new one! If the parting line is reduced the point of bearing contact in the housing become 90 degrees away from the parting line.