tightening a wheel bearing

[Kestas]

Originally Posted By: Kiwi_ME
I'm sure I was taught to leave them slightly loose in high school auto shop but after engineering school and 30 years life experience I now know that is wrong. A tapered roller pair must have a preload to track correctly. A running clearance and/or too much grease is more likely to damage them than slightly too much preload.

They'll track correctly, and this is important for applications such as machine spindles. But preloading automotive wheel hub tapers, which are heavily loaded during cornering, will shorten the bearing life.

 

[The_Eric]

Originally Posted By: Kestas
But preloading automotive wheel hub tapers, which are heavily loaded during cornering, will shorten the bearing life.


I was hoping you'd chime in... While I certainly can't compete with your knowledge and experience, would you be able to hazard a guess as to why the repair manual on my Suburban specs the bearings to be so tight?

 

[berniedd]

VW says this about tightening the axle nut (on Beetles): Tighten the nut till very snug, then back off in small increments until you can just barely move the washer under the nut very slightly with a flat blade screwdriver. This is the correct setting. Then insert the cotter pin.

 

[mechtech2]

Old VWs has ROLLER bearings, not tapered.

One more thing:
Old GM vehicles [and other makes] often had POSITIVE camber [tops of the wheels out more than the bottoms]
Why? It was not good for handling.
It was so that the larger inner wheel bearing would be loaded.

 

[SteveSRT8]

Originally Posted By: Kestas
I was told that automotive double tapers achieve maximum bearing life when the axial clearance is set to 0.003".

In the absence of a dial indicator, I follow the following procedure:

- Tighten axle nut while turning the hub (with or without wheel) until the hub is hard to turn. This is important to properly seat the races should there be any burr or debris.
- Loosen axle nut until slack. Don't turn the hub.
- Tighten axle nut finger tight, or until resistance is felt.
- Lock it as close to that position as possible.



This is the same way I have set them for over 40 years in all manner of vehicles. Just shooting for minimal drag, but there has to be some preload.

Imagine how it increases when you put the vehicle weight on it.

 

[Donald]

In this case its a trailer with no steering abilities.

 

[1 FMF]

in this application if you picture it it becomes clear.
with z axis being from the road surface upward, x-axis being vehicle forward direction of travel, y-axis being to the left/right of vehicle and parallel with the shaft the bearing rotates around. If you have end play then the tapered roller bearing, because it is tapered, does not seat completely and evenly against its race. and a force in the z direction, the weight of the vehicle or boat on trailer, is applied the bearing cocks in the race. The top of the bearing falls inward and down and the bottom of the bearing falls outward and down, and your wheel rotates at an angle- like a ghetto ride where the springs are cut and the car has massive amounts of negative camber.
just loosen your nut and have end play where you can wobble the top and bottom of the wheel by hand about half and inch. whether you have massive amounts of end play and wheel wobble, or just 0.001" of end play the result is the same. with loads placed on a tapered bearing like that it causes failure and wheels to separate from trailers. when you go over uneven road and the boat on the trailer is bouncing up and down, all that load is transmitted through the wheel bearing and it gets shocked cocking back and forth in its race, scoring the rollers and the race. The reality is, you can have some end play and be fine. The main thing is not overloading the trailer or running it under conditions that exceed the strength of the bearings and axles. but from the engineering point of view, you want the correct preload to stabilize the bearing and evenly distribute forces along the entire bearing race and roller bearing elements. When you have the correct preload, the bearing assembly can handle maximum load. When you don't and have end play, the load the bearing assembly can handle is reduced because of the bearing being cocked in the race and the roller element riding on it's ends.

 

[Donald]

well lets hope I get it right then. In this case I am only doing 2 of 4 hubs as the other hubs stay the same because of no brakes.

 

[Kestas]

Originally Posted By: The_Eric
Originally Posted By: Kestas
But preloading automotive wheel hub tapers, which are heavily loaded during cornering, will shorten the bearing life.

I was hoping you'd chime in... While I certainly can't compete with your knowledge and experience, would you be able to hazard a guess as to why the repair manual on my Suburban specs the bearings to be so tight?

Different application. The wheel bearing application needs the axial play to compensate for the thermal expansion experienced during braking. This is not experienced with a Dana 44.

Otherwise, tapered bearings have the longest life with just a bit of preload. I've seen the graphs plotting bearing life versus preload and the bearing life drops dramatically if preload is set beyond the "little bit" of preload. It would be next to impossible to expect the DIYer to set this preload properly.

 

[The_Eric]

I'm probably suffering from a bit of reading comprehension failure, but wouldn't it be the same? You mention "The wheel bearing application needs the axial play to compensate for the thermal expansion experienced during braking." then go to say "This is not experienced with a Dana 44", but my 'Burb had the 44 up from and obviously, disk brakes. What did I miss?

 

[doitmyself]

I'm confused also. You have bearing manufacturer Timken recommending .001-.007 end play and others here recommending pre load. Which is it?

 

[Kiwi_ME]

Perhaps he might have been referring to the tapered roller pair supporting the pinion shaft in a Dana 44, which no-doubt is always installed with a preload

Or the tapered pair holding the ring gear...

SKF's site has lots on information on the subject.
What I got out of it is that a wheel application probably has sufficient continuous radial load (due to the weight of the vehicle or trailer,) as a proportion of the rated bearing capacity, to keep the rollers uniformly loaded even if a small amount of axial clearance is present.
If I had no better instructions from the manufacturer, I would adjust the retainer nut with a few lbs-ins of torque, then back it off to the nearest slot. Then re-adjust after running it up to normal operating temperature.

 

[yonyon]

Originally Posted By: The_Eric

I'm probably suffering from a bit of reading comprehension failure, but wouldn't it be the same? You mention "The wheel bearing application needs the axial play to compensate for the thermal expansion experienced during braking." then go to say "This is not experienced with a Dana 44", but my 'Burb had the 44 up from and obviously, disk brakes. What did I miss?


Just a guess here but I think the icy cold brakes on the Suburban might be a factor. The long wheelbase vehicles don't dump all their weight on the front wheels during braking the way some SWB cars do. With heavy cargo demanding more of the brakes most of that extra weight will be in the back.

 

[Kestas]

When Eric mentioned Dana 44, I thought he was talking about the differential bearings, which is why I made the comment "different application."