Hello all,
Can anyone give me the formula for calculating the shaft-to bearing clearance for journal bearings? I know it is related to motor speed and shaft diameter, but not much else. Thanks!

For motors, see EASA AR100 table 2-7. EASA recommends these values for use when OEM data is not available.

The 2001 version is available here:
http://www.easa.com/indus/ar100-2001.pdf
The table 2.7 mentioned is on page 12 of that document.

If you don't want to look everything up in the table, you can use thumbrule on mils diametrical clearance per inch of shaft diameter. The number decreases as shaft size increases roughly as follows:
~2.5 mils/inch for 2" shaft
~ 2 mils/inch for 4" shaft
~ 1.5 mils/inch for 6" shaft
~ 1 mil per inch for 16" shaft

Thanks, Pete.

Exactly what I needed. I had the question posed to me the other day whether I thought that .013" was too much on a 1500 hp TECO/Westinghouse motor. I think it had a 6" shaft. Using either the EASA table or the rule of thumb, it is too much.

ElPete's reference also states that the bearings should be circular - not all bearings for motors have a circular geometry. For a good sized motor (relative to ones frame of reference) like this I would think/hope that the bearing had been 'designed' for the application rather than use a rkule of thumb.

Bill is correct in seeking clearance data. There are a lot of considerations, not just shaft diameter, but run, speed and temperature so a bearings may work out to 1.23 X shaft diameter.

So 13 mils may be on the outside tolerance???

Picky, picky, picky ;-)

I did say "for use when OEM data is not available..." That should indicate to the reader that the best possible data to be 100% sure comes from the OEM. But if you don't have OEM data, then I have provided an ANSI-approved standard document which is used by motor repair shops for determining required clearances when they have no OEM or customer or other spec.

I don't think that there is nearly as much variation in mils per inch of shaft diameter for motors as there is across other types of equipment. EPRI TR1000897 (medium voltage motor repair spec) has a similar set of limits, and it depends only on shaft size, not on speed or any other factor.

I assumed Richard was talking about plain circular journal bearings and would have been more specific if he were tilting pad or some other type. But we know about assumptions ...

EPete,
Right you are about the plain circular journal bearings. I was asked about the clearance by a maintenance manager when I called looseness in the outboard bearing of a large fan motor(see below). He had it checked out after I had called it and I needed to know at least what was hand grenade close.

Mill.doc (48 KB, 54 downloads)

Some of those tolerences on the large shafts are less than 1mm/m - seems tight.

I don't get the looseness idea with a fluid film bearing. The clearance may be somewhat open (They checked horizontal and vertical assembled clearance with plastigage or wire?), but this isn't way out of bounds.

What about bearing to housing fit? This would be what I understand as looseness with your symptoms. Any seal rubs (oil deflector or air)?

Bill,
I guess I really meant excessive clearance, not looseness. Not much experience with sleeve bearings-or mechanical maintenance or engineering for that matter. Here's what I'll do: I will tell you how I got to this conclusion and maybe you can see where I am off.

This motor was reworked by a motor shop several months ago-I had the data that was taken prior to removal and took data on it shortly after it was placed in service. The plot (top) is before and the bottom is after. There are several clear harmonics of turning speed and the 2x ts peak is the most predominant.
From what I have read about plain sleeve bearings, excessive internal clearance will cause 1. Elevated harmonics of turning speed, and 2. any exitation force, such as imbalance, misalignment, etc. to have a much more pronounced effect.
I suspected from the elevated 2x that perhaps the motor wasn't aligned properly to the gearbox and that the excessive clearance was amplifying the effect.
How do seal rubs show up in the data? Do they show up as 1/2 harmonics of ts or as some other fraction?

Mill.doc (78 KB, 31 downloads)

You could easily get multiples on a rub. Setting the seals on a fluid bearing may require consideration of the lift in the bearing.

I don't see the clearance as being so grossly out as to cause this. If the lubrication is by means of a slinger ring, perhaps one could have a starved bearing (not fully lubricated). There were harmonics at a lower level from the start. An open bearing usually starts with increased 1X (or a stability issue, usually not an issue on 1200 rpm horizontal motors).

Did the bearing just have a large clearance, or had it failed also (damaged bearing)?

He spoke of babbitt metal in the thread title.

I have some practical experience with this original, patent anti-friction metal.

In the Model T, with its crankshaft of 1.25" diameter, it is found that by setting the bearing to rub the journal lightly, and then to run the journal with plenty of light, clean oil, the shaft bearings will decide for themselves the ideal running fit, inside of about three hundred driving miles.

It was found to be the most precise way to fit a bearing. In those old days, and today, among savvy T mechanics, we knew and know how to scrape a split bearing to an approximate fit.
These bearings are hydrodyamic, of course, and support very large, dymanic loads of about 3,000 PSI at full throttle. And they last for as long as clean oil is supplied, without further wear to speak about. The old, thick, poured babbbitt bearings eventually fatigue and crack and fail. It was for this reason that IC engines evolved to use steel shell bearings of finest precision, tin plated with just a couple of mils of anti-gall tin.

The closer the fit, the thinner the oil may be.
The higher the load the bearing can support.
The expansion of the shaft with heat, and the thermal growth of a babbitt bearing, are both in the same direction: toward closing the fit.
Therefore, the minimal clearance for finest running is determined by thermal conditions.
And in a machine with thin shell bearings, it would not do to employ a rub-to-fit clearance technique. That would cause ruination for sure.

What's old is still good, and nothing exceeds the lifetime of a well fitted plain sleeve bearing. Lignum Vitae was also extraordinary; one of the bearing materials for low loads, low speeds, where only boundary lubrication occurs: they need never wear out at all.

The old tech of pianos: their bearings are unlubricated nickle-brass pins working in woven felt bushings. The loads relative to journal size are high. The wear is entirely absent, even after 20 million repetitions of a piano hammer, if its pinning was done correctly, the running fit has not altered by even one ten thousandth of an inch. I was a piano technician, and can-so attest by experience.

The machine which suffers a worn bearing was:
-either set up incorrectly at first
-or was not designed ideally, not sealed or scaled to the load
-or was neglected or run in otherwise adverse conditions.

Interesting stuff. My theory book says my T should have had about 1.25 mils of crank/bearing oil clearance. But in practice, using standard motor oils, it was found by Plastigage check, to have rubbed in a fit with a total clearance of .75 mil, only. Tighter than they "say" it should be. But it is practice which counts, and in a babbitt bearing, the machine can be it's own best judge of what is the ideal oil clearance. Fitting by rubbing requires, however, careful monitoring of bearing frictional heat, or seizure and babbitt melt will occur.

My closely-fit T bearings held that small clearance for tens of thousands of miles (the car got in a wreck years later). This proves to me that that clearance was entirely sufficient to provide for vital shaft eccentricity within the bearing space (for hydrodynamic oil pump action).

I doubt that professional engineers have time or the safety to experiment with old-school methods. The learning curve could be $$$, ouch!

But as soon as oil clearance exceeds the load bearing capacity of the oil used in that bearing, for that running speed--boundary effects obtain, wear occurs and the bearing eventually requires replacement, and the shaft is liable to have become worn to boot.

This message has been edited. Last edited by: Reid Welch,

Another important effect on a heavily loaded bearing such as the Model 'T' would be mechanical deformation under load; this may play an important part in the bearing operation.

One of the issue that happens on large equipment like generators is that the bearing on turning gear (if no lift oil) will wear to fit the shaft geometric curvature; this is like a negative preload factor on a bearing. If you have ever seen this happen, you know the bearing doesn't work well and may fail because of this.

Today on precision bearings (like for compressors) hand scraping to remove metal, which changes the geometry, is forbidden.

Oh yes, so right.

The T had no oil filtration, no air filter, as issued.

I put both onto my car--and this surely helped longevity.

The T's shafts are soft steel. It is often alleged that soft bearing metals are that way, so that fine grit and wear products may embed as a "protective" mechanism.

That, imo, is a Big Lie. One can make a dandy lap by pounding diamond dust into lead and use that to rub down hard steel.

Soft bearings are soft so that they make fittings easier and more forgiving of high spots, which may wear down.
Best bearing, highest load is hard steel on hard steel--but not always practical due to ruination in case of the least oil starvation.

Steel on cast iron is the next best load bearer, imo, and has less tendency to seize hard. It won't embed lapping agents.

Wear products and silica, etc, embed in softer bearings, and once there, are lapping agents attacking the shaft during start-up and during other times of non-hydrodynamic operation.

((you all know this; I only learned it by observations and by making errors in fitting things))

The plain bearing shafts in old autos most definitely would wear out of of round and with taper. Crank throws were most liable to show wear first. The heavy (hundred pound) flywheel of a T, concentrating weight on the third main bearing, was liable to wear that journal to a taper.

Then it would become impossible to "snug up" the crank bearings with any sort of lasting results.

Scraping by hand is a tedious process and it is simply not precise enough for a true fit.
I'd scrape to rough things well (we use bearing blue or a marking ink) and then do a careful rubbing-in by the friction and oil method.

These techniques are simply not amenable to modern machinery, so true.

Main point: the "recommended clearances" are already on the outside range of hydrodynamic sustainability, -if-

a) the shaft speed and loads are low
b) if either of the surfaces were not honed to size. Reaming is N/G. Lathe turning is N/G. Parts must be ground round and finely finished.

You guys know all of this already.
I only learned about by skinning knuckles.
But it was great fun. A machine put in good order is almost like a living thing.

This was true of that ubiquitous car, the Ford T.

In those awful, bad old days, of unpaved dirt roads, and plain, unfortified oils, and sulfuric acid in the fuels, and no air or water filtration, the mean time between engine overhauls was 10,000 miles. You'd rebore and fit larger pistons, and perhaps replace the crank and then repour the babbitt bearings.

But if it were run as we run our autos today, that 1908 design engine, holds up just fine.

Bearings are remarkable, basic devices that have interested me since childhood, their being wonders of longevity if done right,
and disasters of short life if misunderstood, and -just slightly not-right-.

The ones that don't last after overhaul were probably put in by hacks working for the motor shop, perhaps buffing oval shafts with emery cloth; hammer mechanics. That sort of 'deal'.

The GE sleeve bearing 1/3 HP fan motor venting my workshop attic runs 24/7 nearly, and has been powered since 1984. I've oiled it twice and serviced it never. And there's nothing remarkable about that at all.

If the machine is made right, and the service is not adverse (such as very heavy loads with very slow speed, insufficient to pump the oil film), well---bearings would not wear, hardly at all.

This message has been edited. Last edited by: Reid Welch,