What's new
What's new

Tapered roller bearing questions

Hbjj

Cast Iron
Joined
Mar 18, 2013
Location
Washington st
I'm working on a gearbox that has two pairs of timken bearings #563 with #567
It runs at 90 rpm grease and oil bath lube 30 hp input

Caps with shims to set end play
The question is how much?
I've sent inquiries to timken and the OEM wont answer the phone
20190115_055438.jpg20190117_102938.jpg20190117_103042.jpg20190117_102948.jpg
 
I just rebuilt trans axle with a lot of roller taper bearings. No bearing had end play. When possible a drag associated with the preload was specified, generally in the range of 5 to 15 inch pounds. I got tired of swapping shims. Only the axle bearings did not have a drag spec, just a tighten nut snug + 1/6 turn. Took a 2 foot wrench to turn that last bit.

When you get a little drag, 1 or 2 thousandths change in shim thickness can take you from one end of the specs to the other.

Bearing life charts show that these bearings last longer when tight than when loose. However, too tight and the life span falls off a cliff.

I think that you need the OEM info as I am sure the correct loading is a factor of design and application. The mfg knows how much the case flexes, how the loads are applied, operating temperatures and other things I don't know about. I think even Timken would be guessing but they would have a better guess than I would.
 
I dont think the OEM exists any more
Still waiting on timken...

I dont expect a hard number on the preload/end play
I'm looking more for what other experanced people have used on similar stuff
 
The Timken catalog gives engineering data including recommended clearances. I believe it is available on line. Unfortunately, you can't be sure the OEM followed the Timken recommendations.
 
I've sent inquiries to timken and the OEM wont answer the phone
In the absence of any help, I'd stick a butane camp stove underneath, heat it up to 100* or so then adjust until the bearings were snug but not tight, if you know what I mean. That ought to get you in the ballpark ?

btw, a well-designed pair of gears should be about 98% efficient, so if you are transmitting 30 hp through one pair of gears, you will be dissipating 3/5 of a hp just through the mesh and some through the bearings, so the box will be putting 1000 watts of heat into the casing at least ...
 
I jus did a dynex gear reduction drive little more horsepower going in to it that rpm out, I looked everywhere for specs some just are not available. finally talked to a mechanic that had worked on them before he said .001 of shaft movement or less no drag. but one thing I noticed taking gear boxes and rear ends apart it seems like a lot of assemblers will throw it together to fit shims without taking the time to torque the bolts then when its torqued to a even foot pounds everything is binding tight. lost a rearend and in a 10 wheeler in 5000 miles because of that.
I sent one of the shims out of the dynex box and had this place Custom Metal Shims, Spacers, Brackets, Plates, Laser Cutting | The Shim Shack make me a up a new shim stack I was impressed with their work it was a 40 year old box and manufacture did not have the specs or shims anymore and he drew them up and had them back to me in less than a week.
 
If I were to rebuild the gear case, which I have rebuilt a bunch of them, I would set the bearings with a slight preload. Don’t get it too tight or it will not last.
 
You could look at front wheel bearing tighten procedure for a vehicle with similar size bearings. Tighten to a spec, back off and tighten to a smaller spec. New or used = different specs.
 
It's been a few years, but .001-.003 preload would be my guess on a new bearing.

Agreed. You don't want the shafts to be wobbling around at all, with all the bolts torqued tight. It should feel free turning when the bolts are only snugged up, and should get a bit of drag when you do tighten the bolts up. If it were to require a wrench to turn the shafts, then I'd say it is too tight. Your thinnest shim is likely only .001", so when you get that close, you'll have to decide if you like the feel with it in or out. No shaft shake and a bit of drag is the ideal.
 
Agreed. You don't want the shafts to be wobbling around at all, with all the bolts torqued tight. It should feel free turning when the bolts are only snugged up, and should get a bit of drag when you do tighten the bolts up. If it were to require a wrench to turn the shafts, then I'd say it is too tight. Your thinnest shim is likely only .001", so when you get that close, you'll have to decide if you like the feel with it in or out. No shaft shake and a bit of drag is the ideal.

Typical procedure was to depth mic the bearing depth and the shoulder on the cap. Subtract bearing depth from shoulder depth to calculate shim pack. Add or remove shims as desired to achieve preload or play as desired. An alternative is to shim it loose and measure the endplay with a dial indicator, then remove shims to end up at your final number. One thing to be aware of is the nature of tapered rollers when they are lowered upside down into a race. The rollers will fall to the bottom of the cage and won't properly seat upwards against their thrust shoulder on the inner race until it is rotated several times allowing the shaft to settle downwards.
 
It's been a few years, but .001-.003 preload would be my guess on a new bearing.

On a related topic, I just fitted some Timken front wheel bearings, there was a Timken video on Youtube detailing the process of preloading and measuring end float to get a figure of between 0.001" and 0.005" maximum. It's been a few days and I managed to take the car for a long drive without braking and the bearings were running cool.

 
These are probably not matched high precision bearings such as a spindle has, just standard off the shelf to fit the application. A Timken bearing set on all my vehicle front wheels are set with play, tighten until zero and back the castle nut until the cotter pin slides in and done. Got a couple hundred thousand miles on those trucks and they're still going.
Harley wheels with Timkens come with one spacer between them (there are only 3 different lengths to choose from) and all have a few thousands end play.
Harley crank shafts have a matched set of Timkens on the left side for end play and to carry the drive sprocket. Those are speced to have .001" end play after everything is torqued.
One thing thats been proven to me over the years is over graeasing or over tightening will kill a bearing faster than anything other than river/lake water and whatever it has in it.
If I had that job on front of me I hope I would be thinking clearly enough to assemble it with some clearance if only .001" just to be safe.
Dan
 
is this a speed reducer or increaser or? Parallel shaft? The 90 RPM is input or output?

Horizontal mount gear box. Oil lubricated splash bath with sump. Operating temperature 120F to 140F Input rotation is 90 RPM Assembly temperature 72F to 80F All shims and mating surfaces are clean and smooth.

Lock nut for cone is torqued and locked. Assemble with free play. Rotate shaft several rotations and load back and forth several times. Measure end play with dial indicator. End play is 0.XXX

Remove shims to remove endplay plus 0.001 to 0.003 shim. assemble and torque retainer cap screws. Confirm smooth rotation of shaft by hand and zero end play.

This will yield near zero end play/preload at operating temperature. Document the procedure and store for future use and modification.

If this unit is vertical shaft mount and the grease is introduced through the seal retainer, the grease is to prevent dry start ups of the top bearings and lubricate the lip seal to shaft contact.

Grease lubricated applications such as vehicles, are different.
 
I dont think the OEM exists any more
Still waiting on timken...

I dont expect a hard number on the preload/end play
I'm looking more for what other experanced people have used on similar stuff
Maybe find a manufacturer still out there making a similar unit, and follow there idea of the preload if you get no firm answer.
 
is this a speed reducer or increaser or? Parallel shaft? The 90 RPM is input or output?

Horizontal mount gear box. Oil lubricated splash bath with sump. Operating temperature 120F to 140F Input rotation is 90 RPM Assembly temperature 72F to 80F All shims and mating surfaces are clean

No speed change it's just a pair of gears that provide a second counter rotating output

Assembly temp is more like 45°

Its vertically mounted and I see no provisions for oil to get to the top bearings but there are grease fittings on the top caps

Operating temp? I'm not sure its inside a plant but I doubt it's heated

Yes I have things fairly clean I removed all of the previous hammer marks ect
No hot or solvent tank here so I do what I can with what I have

Btw this thing ate a bunch of metal from the input seals/top cap so the gears are not so nice anymore

20190118_054125.jpg20190118_054221.jpg20190118_054244.jpg20190118_054131.jpg
 
On a gearbox like that if they're straight cut gears I'd set a thou or two end play. With helical or angled teeth you might want a little tighter. Don't consider your job done after setting the cold preload/end play. Get the gearbox up to operating temp and check the end play again. Use that operating temp end play as your relevant number. I used to rebuild very similar gearboxes at one of the larger shops I worked in.

BTW, since you sound like you don't understand what's getting discussed here regarding operating temp - when guys are talking about operating temperature they mean after the unit has been in operation for a while. The friction from the gears and bearings rotating will heat the whole unit up a bit. That will change clearances and such when the steel expands and everything moves around a little due to thermal expansion. They're not taking about the place where the gearbox is installed being heated.

If this is a continuously operating gearbox that's important. If it's intermittently operated with a low duty cycle then room temp or whatever the normal temperature where it's installed at is good enough.
 
We used to fit large diameter ( 20" dia ) " Timken " bearings into the facing slides of the " Kearns-Richards " SF-125 Hor bores. Believe it or not they had an interference of 0.020". They were well lubricated with oil and the max speed was about 200 rpm.

Regards Tyrone.
 
We used to fit large diameter ( 20" dia ) " Timken " bearings into the facing slides of the " Kearns-Richards " SF-125 Hor bores. Believe it or not they had an interference of 0.020". They were well lubricated with oil and the max speed was about 200 rpm.

Regards Tyrone.

They also were precision bearings and most likely driven with angled gear teeth. In a precision machine tool and not a simple power transmitting gearbox. Like another fellow said above, little different situation. You can get away with a tighter fit when the bearings are considerably higher tolerance on roundness and size and especially running in an oil bath. .020" does seem high though, wow. Are you talking about interference fit on a race or preload on the assembled bearing, Tyrone?
 








 
Back
Top