Wheel Bearing Preload - How to check Hub Temp? (2 Viewers)

[agaisin]

There are a couple really good wheel bearing threads that are closed, otherwise I would have posted to them. In short, I want to make sure I have attained proper preload without overtightening the adjusting nut on the wheel bearings. I am concerned b/c at the time I was working with the premise that more preload is better than loose wheel bearings, and I got to the 15lbs (on 9.5-15lb range) for the fish scale. That took more torque than I used last time and I'm sure more torque than most would be comfortable with.

For reference, here are the two threads I'm referring to that deal with proper preload on 100's and that are both closed:

https://forum.ih8mud.com/100-series-cruisers/420496-wheel-bearing-fsm-torque-light-just-right.html

https://forum.ih8mud.com/100-series-cruisers/293907-front-wheel-bearing-help-preload-question.html

The topic of proper preload is extremely important - too little or too much are both bad.

From what I'd seen on the board it seems more common for people to have loose wheel bearings so I was always more afraid of too loose than too tight. But further reading makes me realize how dangerous too much preload can be.

Problem I have is that it seems that anyone using the FSM steps end up tightening the adjusting nut a lot in order to attain the FSM spec of 9.5-15lbs using a spring tension guage (also fish scale) as described in the FSM. Alternatively they ignore FSM and use either experience or other repeatable steps to get to the proper preload (either sticking with a set torque value for the adjusting nut, or using ability to move/shift thrust washer as a guage, etc).

Checking for loose wheel bearings is easy as you can just check for play. But, how do you know if you have overtightened your bearings? The only practical way I've seen talked about is checking the temperature of the hubs?

My question is where do you take your temperature measurements from?

In the image above, I've pointed out 6 spots that I toyed with b/c they are all easily accessible without removing the wheel and they each had different temperature readings when I tested:
1) Rotor/Disc
2) Rotor where it meets the hat
3) Rotor hat
4) Edge of Hub
5) Face of Hub Flange
6) Dust Cap

I assume 1,2,3 are irrelevant b/c they are the rotor not the hub.
I figure #4 is irrelevant b/c while it's technically part of the hub, it's further away from the portion of hub that makes contact with the grease.

I think #5 and #6 are the right places to measure but wanted to know what others think and use as a guage? Also, what temperatures do others use as a guide for "too hot"?

 

[12guns]

Agreed, 5 or 6 is your best bet. Brake parts will be smokin' hot regardless.
I'm glad you posted, I've noticed my front wheels seem to get hotter than on other vehicles I've owned in the past. May be nothing, but may also mean my bearings are overtight from the previous owner.
I'll be curious if we can come up with an "acceptable" temp for hubs/wheels.
Do your alloy wheels get hot to the touch? Some is normal since it's all connected to the rotor/brakes, but I wonder how much heat is ok??

 

[Layonnn]

Going by the FSM of tightening back to 4-6lbs, only gave me about 3-4lbs on the fish scale. I had to torque my adjusting nut closer to 30-35ft/lbs, this gave me roughly 10-12lbs on the fish scale.

 

[Trunk Monkey]

There are a couple really good wheel bearing threads that are closed, otherwise I would have posted to them.

Really old threads get auto-closed, software and Woody's choice. If there's one you'd like opened, just PM me.

As for preload, I would go use the FSM method of preload rather than temps. There's too many other factors that can contribute to temperature differentials.

 

[super90]

Going by the FSM of tightening back to 4-6lbs, only gave me about 3-4lbs on the fish scale. I had to torque my adjusting nut closer to 30-35ft/lbs, this gave me roughly 10-12lbs on the fish scale.

I just did this yesterday and had the same results. It took about 30 ft lbs on the nut to end up with 10 lbs on the scale. Nice to see consistent data across the fleet !


Also, to address the OP, if you have the bearing preload too tight, you will feel it with your hand on the hub. If they are right, you can hold on to the hub after being on the road at 70 mph for a while. Also, there will be no play when you put it on the jack and try and rock the wheel while holding at 12 and 6 o'clock. If they are too tight, the hub will be hot to the touch after driving at speed. Not super scientific, but that's the old school way to give a quick test and it is pretty obvious.

YMMV

 

[agaisin]

Also, to address the OP, if you have the bear preload too tight, you will feel it with your hand on the hub. If they are right, you can hold on to the hub after being on the road at 70 mph for a while. Also, there will be no play when you put it on the jack and try and rock the wheel while holding at 12 and 6 o'clock. If they are too tight, they will be very hot to the touch after driving at speed. Not super scientific, but that's the old school way to give a quick test and it is pretty obvious.

YMMV

I feel silly asking, you say to touch the hub but my question is what do you touch? Our hub is not really exposed with the wheel on - only the dust cap and flange would be exposed.

So what part do you touch? See the picture I posted above...

Thanks!!

 

[super90]

Yep, the dust cap and flange. When things are wrong, that should get noticeably hot.

After doing mine, I was checking this very thing for the first few miles, and then again as I made a 200 mi. trip today. Stopped after the first 30 minutes to check the temp, by hand, and it was all fine.

 

[Porter]

When doing the bearing repack two months ago, I intended to go with a torque wrench, and follow specs on the mentioned threads here at mud. Well, I'm impressed with anyone that can get a solid torque value on those nuts- impossible for this shade tree mechanic. Of course I had no scale on hand.

I made an educated guess, and hand checked temps, and no issues. I believe I'll pick up a laser thermo for good measure as well. I'm guessing I'm not the only one winging it so speak.

 

[super90]

I hope you greased the axle shaft bushings while you had it all apart. If those get "dry" they will create heat too....

https://forum.ih8mud.com/100-series-cruisers/448945-how-lubricate-front-axle-shaft-bushings.html

 

[agaisin]

I hope you greased the axle shaft bushings while you had it all apart. If those get "dry" they will create heat too....

https://forum.ih8mud.com/100-series-cruisers/448945-how-lubricate-front-axle-shaft-bushings.html

Yep - used the slee tool!

 

[agaisin]

So - after yesterday's commute to and from work (25 mile/1:15 hour each way) with part high speed (55-75) and part slower speeds (25-45) and good amount of braking, I got pretty similar numbers on both trips:
- Front Dust caps: ~110F
- Front Flange face: 130F
- Tire Walls: 110F
- Wheel Spokes: 110F-130F depending on the wheel and the trip
- Rear Hub Faces: 130F
- Discs: 150-170F
- Where rotor disc meets rotor hat (heat sink?): 210
- Rotor Hat: 180F
- Hub Edge (what contacts rotor): 120-150

Also fyi - the Non-Contact IR Thermometers can be misleading with the small laser dots since they all have a distance to spot ratio. For example a D:S ratio of 6:1 means at 6 inches away from the dot it's measuring a 1 inch radius circle. So when you hold it 2 feet away and you see a small dot and think you're accurately measuring that spot, you're not - in the 6:1 case you'd be measuring 4 inch circular area... As my numbers above show, a 4 inch area includes a wide range of temperatures so if you use IR hold it close enough to the spot you're measuring...

Don't know what others have gotten, don't know for a fact what's good or bad other than the claim that it (cap or flange) should be cool to touch or not too hot to hold for x seconds). I've seen Timken documentation saying tapered roller bearings that reach 250F need to be replaced and both Timken and M1 Synthetic wheel bearing grease is good to ~300F.

As far as my cap and flange temps, I know I was on the high end of spring tension range at 15lbs but technically it was within spec (9.5-15lb with tension scale). So I assume the numbers I have on the cap and flange are higher than what others get but can't say for sure. Also, fact that rear hub faces were at 130F and those bearings are factory sealed, I think it's reasonable to say 130F is still safe. Again curious if anyone else has good info - I searched all of Timken's tech tips etc and couldn't find a range of acceptable operating temps for their bearings anywhere, just the following (in the context of heating up during installation/removal):
"Take extreme care that bearings are
never heated above 250°F (129°C).
If bearings are heated above this
temperature their metallurgical
structure may soften, rendering
them unsuitable for use."

Other than that it seems clear that higher operating temps would decrease life expectancy so the cooler the better (as long as the grease gets warm enough to flow around inside the hub and bearings).

 

[super90]

Temps sound higher than mine. I set for about 9 to 10 lbs on the scale and I feel sure my temps at the flange are 20-degrees cooler. I'm out of town right now and my infrared thermometer is at the house, so I can't get you exact numbers today. But, heck, sometimes too much data is a bad thing.

It's all about what you are comfortable with, but I think I would back off on the preload.

 

[TrekboxX]

Truly, no offense, but it's not about what "we" are comfortable with. It's about what it should be, what is right. I've been second guessing my preload ever since I replaced the front bearings. I've read all those other threads as well, and nobody seems to really know. For me, with brand new races and bearings, adding torque above a certain point hardly changed the resistance in the bearings. When I finally got to 53lbs (go ahead), I barely had 10lbs of resistance. I wanted to get to the higher end of the scale because it seemed the consensus was that with bigger tires, etc, the FSM specs were way too low. They aren't even in the ball park! 4-6lbs of torque is nowhere near enough to get anywhere close to 10-15 lbs, which was my target. It just didn't "feel" right to torque them that high, but I kept reminding myself that a 3 ton SUV was flying around on these bearings. Hate to tear into it again, but do you guys think I should back it off? Or maybe try the heat test? Thanks, all.

 

[agaisin]

(I have to put in a disclaimer - I have repacked my bearings a number of times but I am not a certified mechanic and have relatively limited overall experience. Wheel bearings are serious business and if you're unsure have it checked out by a pro. I feel comfortable writing everything below b/c I have obsessed over the tech aspect and read a good amount of official manufacturer and expert guidance as well as diy advice on the topic and I think it's valid and useful information condensed from a broad range of sources).

1407driver - you seem concerned that you may have overtightened. I suggest you check the temp on the dust caps after an extended drive - at least by touch. If they are so hot you can't even touch them, you should inspect bearings and races and either repack or replace the bearings if they show signs of damage. If on the other hand they are not too hot to touch you should be fine. You can also use a dial indicator to check for between .001 - .005" play. The heat method sounds unscientific but the basic point is that when it's overtightened it will generate enough heat to damage the bearings. Exactly how hot is too hot is not documented officially by anyone anywhere that I've seen so it's possible the too hot to touch heat test is overly conservative but even more important is that when it's adjusted correctly it shouldn't be too hot to touch. Also - aside from the amount of torque being so high, does everything feel ok while driving?

Conceptually there are two approaches for wheel bearing adjustment - "preload" or "end play". By definition setting preload means there is no end play. It's easy to verify end play accurately with a dial indicator but it's harder to accurately measure preload. Which is why bearing manufacturers and the trucking industry have fixed on using end play approach to bearing adjustment (see RP 618). Technically speaking preload is ideal but it's hard to attain reliably with the tools at our disposal and there's also a very narrow window whereas end play has a wide range (relatively speaking) so with end play there's more tolerance for error than with preload.

There's lots of instructions out there and approaches for setting bearing preload but technically I think our FSM approach with a spring scale, and the approach of checking thrust washer's (lack of) play are the only ones that are used for the preload approach. Any time you're backing off or using a dial indicator for measurements we're talking end play. Again - FSM specs preload, but end play is what many are used to and is what's documented by Timken and most repeated and frankly it's probably less error prone than the preload approach documented in our manuals.

Here's more on preload vs. end play: http://www.vehicleservicepros.com/article/10712957/preload-instead-of-setting-endplay-gaining-favor

The FSM's intent with the spring scale is to be able to measure preload.
This approach is also described on P.6 here:
http://www.timken.com/en-in/Documents/Metal Industry 22-11-11.pdf

Interestingly the instructions for bearing adjustment in 80 series iirc tell you to first measure the amount of drag from the grease (using spring scale) when the hub is on spindle but nuts are loose and the target measurement on spring scale is that + 6lbs iirc. That would seem to account a little bit better for the differences in viscosity and temperature which can vary greatly. Our 100 series manual just has one range for spring scale measurement and doesn't account for drag from the grease...

Anyhow - the Timken and other manufacturers just say to adjust so you get between .001"-.005" end play per RP618 guidelines.

Timken on Bearing adjustment:
http://www.timken.com/ja-jp/solutio...rket/heavyduty/techtips/Documents/Vol1No7.pdf
http://www.timken.com/en-in/solutio...1_Proper_Tapered_Bearing_Settings_English.pdf

Here's from Timken on damage analysis: http://connect.timken.com/uploads/tapered-roller-bearing-damage-analysis-part-1.pdf
and
http://connect.timken.com/uploads/tapered-roller-bearing-damage-analysis-part-2.pdf

 

[TrekboxX]

Thanks for the info. The temp check is next on my list, good suggestion. And yes, besides the original noise that is still there (see below), the truck drives fine. The bearings/hubs turned quite smoothly even at those seemingly high tq settings.

I'm not a pro either, thank goodness. The last "pro" that did my bearings used a pair of channel locks to tighten the nuts and marred them up pretty good. Also, this "pro" didn't bend a tab on the star washer outboard, so the passenger side nuts were not even finger tight. I was hearing a noise from the passenger side and thought it was a bearing, so I took it to another pro who said the bearings were fine. Got it home, jacked it up, and clunk clunk by hand. Decided to do it myself. So, all went well until taking the advice of a major "pro" (toyota fsm) who "says" to tighten preload to 6-15 lbs. Total bs. You'd need a long cheater pipe to get enough torque to get to 15lbs, and it would still be in spec? If the manuals/general knowledge regarding helicopter maintenance was this poor, I'd have been dead long ago! So, I'd rather do it myself...!

In the article regarding end play vs preload, it states pretty clearly that the author(s) think preload is better. But, it also speaks of these magical newfangled preload tools (???) of which the world has not previously seen. Wtf is wrong with an accurate torque wrench? If that won't do it, then a bearing assembly redesign is in order. Or, with the thousands upon thousands of 100's (and many others in the same boat) on the road, being worked on by pro and amatuer alike, maybe it's not an exact science at all. How many catastrophic failures have happened as a rest of bearings? I don't know, but I just wish there was a "right" way.

Not vent at you, just venting! Thanks for the good insight. I will check temps asap...

 

[agaisin]

Wtf is wrong with an accurate torque wrench? If that won't do it, then a bearing assembly redesign is in order. Or, with the thousands upon thousands of 100's (and many others in the same boat) on the road, being worked on by pro and amatuer alike, maybe it's not an exact science at all. How many catastrophic failures have happened as a rest of bearings? I don't know, but I just wish there was a "right" way.

1) Why not a torque wrench? Good question - not saying I have a definite answer but I have a few points to make there. First - "A properly tightened bolt is one that is stretched such that it acts like a very ridged spring pulling mating surfaces together." That's what we're going for when we torque bolts or nuts to a specified value. Here with preload, we're going for a different goal - removing play. We're not trying to tighten it down so much that the metal stretches and pulls the mating surfaces together. We're just trying to tighten down to the point where we've eliminated play. Think of it as .000" of play. That takes some amount of torque to do, but I guess it can vary by circumstances and torque is not the factor, it's how much play do you have - that's the key factor. Once you start applying torque beyond preload (zero endplay) then you're overtightening. I can't explain scientifically why but it seems clear that there isn't one torque value to tighten the nuts at to get the proper preload.
Also - b/c there's grease on the threads torque values are all off - despite the best attempts to clean it out there's gonna be some residue on the spindle threads and nuts.
I guess unlike bolts/nuts and steel on steel when you talk bearings and grease maybe the variables make it impossible to determine a specific torque value.

2) Also regarding your comment about redesign - I think there are redesigns out there that address the issue of proper preload. For example Sealed bearing units like the rears on our vehicle. On older cruisers neither the fronts nor the rears were sealed. I believe on newer cruisers both the rears and the fronts are sealed. So you could argue they did redesign. The downside of sealed is that you can't inspect them or repack them. inspecting and repacking every 30k means you are more likely to know about impending issues ahead of catastrophic failure, and you can extend their life significantly. seems like the sky's the limit on longevity for well taken care of bearings on our vehicles. if you do water crossings with sealed units once they're contaminated they need to be replaced. With adjustable bearings you just clean/repack 'em.

3) Regarding the number of catastrophic failures - I think that just goes to the point that most cases of bearing issues are too loose and too loose isn't usually catastrophic (as long as the nuts don't back off completely). Too loose is better than too tight. And also - there is a lot of wiggle room for loose but acceptable/to spec. I think it also goes to the point that RP 618 is the best way to go - i.e. even though preload is ideal for bearing performance, it's too hard to attain and becomes a safety issue. So adjusting for end play is the best/safest bet - and that's probably why so many people do it - tighten down, then back off by 1/4 or 1/8" etc... It's harder to go wrong than trying to adjust for preload.

 

[agaisin]

1407driver - I think the points in my previous post are valid but you got me thinking with your question about why not use a torque wrench and I think in the course of revising/rethinking my previous post I came to a real answer. It came from a more careful read of p. 6 in the Timken article I referenced above:
When bearing preload is required, the most convenient way
to determine the preload is to read the torque from the shaft’s
rotation. To measure torque, wrap a string around a gear or
other known diameter (Fig. 10). Attach a spring scale to the
string and pull the spring scale to rotate the shaft at a slow,
steady pace. Read the force indicated on the spring scale.
Multiply the force times the radius to determine the bearing
torque. You can obtain a direct torque reading by using a torque
wrench (Fig. 11 and 12). If necessary, you can convert torque
into preload force or dimensional preload.

1) the torque we are after is the "bearing torque" not the torque on the adjusting nut. I have to believe there's not enough of a direct correlation b/w the adjusting nut's torque and the 'bearing torque'.
2) even though they mention use of a torque wrench, they are talking about putting a torque wrench on the shaft to see the torque it takes to turn the shaft. In our automotive application the shaft doesn't turn inside a bearing. Yes the axle shaft (insie the spindle) turns inside the needle bearing but that's a different bearing than the ones we're talking about. The shaft in this case (spindle) is static and doesn't rotate - it's the wheel that moves around the spindle. So we can't take a torque reading with a torque wrench off the 'shaft' the way that's described in Fig. 11 in the Timken aritcle.
3) So if we're aiming for preload rather than end play, that only leaves us with the approach that is documented in our FSM and in the Timken article above which uses the spring scale attached to a point on the hub of a known diameter and multiply that value by the radius to get the bearing torque we're after. Toyota simplified it for us by leaving out the other math (they did it for us) and just telling us that if we get a spring scale measurement of 9.5-15lbs (when setup properly) we've gotten the right bearing torque....

Does this make more sense?

 

[TrekboxX]

Well written, and yes, it makes very good sense. I do agree that there is only a vague correlation between bearing torque and adjusting nut torque, as I indicated above when I spoke of my own experience recently. I theoretically understand the ideal of zero play, but I also understand with the forces acting upon a heavily leveraged bearing from a 3 ton fast moving vehicle, the amount of friction caused by slightly excessive torque values should be a drop in the bucket. Having said that, I do want to strive for the ideal, if possible. Regarding your revised post- it makes perfect sense. I know the variables are endless- density altitude, temperature, grease type, etc, but we put a man on the moon over 40 years ago. Am I to understand we cannot figure out a way to achieve safe, reliable bearing torque values on our autos? At least Toyota could have said "torque until bearing torque reads 6-15 lbs as indicated on a spring scale. Adjuster nut torque will vary, but should be between x and y. Do not exceed y."

Good talk!

 

[TrekboxX]

One more thing- since 53lbs on the nut gave me ~9.5 lbs on the lug, repeatable on each side, should I assume this is still ok? Seems very high after this discussion.

 

[agaisin]

One more thing- since 53lbs on the nut gave me ~9.5 lbs on the lug, repeatable on each side, should I assume this is still ok? Seems very high after this discussion.

I'm speculating and can't explain it but perhaps if you need 53 ft lbs of torque on the adjusting nut to get just 9.5 on the spring scale, that's an indicator of worn bearings or related parts?

At this point if I were you I'd use the empirical evidence of heat at the hub as an indicator of overtightening.

Sent from my iPhone using IH8MUD