Guys
I have recently returned from a trip to England to see friends and nieces. 1800 miles, average motorway speed 70 mph, overall average speed 49 mph, overall consumption 19 mpg, motorway consumption 21 mpg.
Car perfect in every way so VV pleased that all my care and work paid off; EXCEPT that 60 miles from home a guastly noise from the rear axle occured, luckily I was able to get off the motorway (tol road) as an exit was only a couple of Km away and call the towtruck.
A very nice guy arrived, who turned out to have done his apprenticeship on the V12, and loved them!

A couple of days later the car turned up at home and it was clear that a hub bearing had somehow failed. In fact the outer bearing inner race was turning on the hub tube as the LHS bearing had locked solid. No idea why. Odd handling problems from the rear had been evident for about 100 miles before the full failure, so I was aware something was happening. Seee this guastly mess:

Wrecked hub tube

New parts ordered from Manners, spare hub cleaned up as the old one was properly ruined. I decided to rebuild both hubs, just in case when I rebuilt the axle I had made some enndfloat error; but whether insufficient endfloat was the cause I do not know. The RHS hub bearing was perfect.

Now to the point of this thread. I happen to have in stock a huge hoard of all the necessary spacer shims to set the hub bearing endfloat, these I picked up as a job lot some years ago. This is a typical one:

Oil seal track (large silver item) and endfloat spacer shim the brass-coloured item in the driveshaft, though it is not made of brass

So the new outer races of the new bearings were knocked into the hub carrier and the outer bearing's inner race was drifted onto the hub tube:

New outer cones of the bearings in place, plus oil seal on the outer end of the hub carrier


New outer bearing inner race drifted into position

Now the next step is to push the hub into position in the hub carrier and the drift the inner bearing's inner race onto the hub tube. This is a very tight fit indeed. After which this is what things look like:

New inner race of the inner hub bearing partially drifted into place on the hub tube.

Bearing fully drifted into place. It is the space between the bearing (pointer on it) and the hub tube (inside the bearing below it) that the spacer is required for

Now comes the hard part, in principle this is the idea:

• with the bearings fully drifted together there is no endfloat, while the spec is to have between 2 and 4 thou endfloat.
• you measure the gap between the bearing and the hub tube, and you require a spacer that is that amount plus (say) 3 thou.
  
  Measure gap with caliper
• 
  Gap reading
• Using a suitably sized drift, with the hub carrier held in a Workmate's wooden jaws, tap the hub tube down a touch to release the bearing from its "no enbdfloat" condition
• Place in the gap a spacer as measured plus 3 thou, assemble hub onto splined driveshaft and job done
• BUT:

The problem comes in verifying the setup having done all this. The Workshop manual calls for the endfloat to be measured to ensure it is correct. In practice, this is VERY difficult to do. I have a driveshaft whose universal joint I have welded solid. This I placed in a vice so I could install the hub and carrier assembly on the driveshaft splines, and then tighten down the driveshaft castellated nut, just as if the assembly was on the car's driveshaft. Then I set up my dial guage and measured the endfloat.
The measured endfloat proved VERY difficult to verify, as I found it very hard get consistent, repeatable readings. Remember that the entire tube and bearings are greased up with new grease, and you are trying to measure 3 thou by lifting the hub carrier. NOT easy.

After a day I was at LAST able to get repeatable readings, but only because I had a load of Jaguar endfloat setting shims and I gradually closed in on a repeatable reading, as each time I thought I had one, the endfloat seemed to get bigger and a smaller shim was needed. Anyway, job done and I am certain the endfloat is to spec. I shall, though, test it after a few hundred miles by rocking the rear wheels to ensure it is OK, and if too loose or tight it is easy to chnage to a smaller or larger shim. You can feel endfloat at the tyre as a very slight movement on rocking it in and out.

But what about those unlucky souls among us who do not have a bagfull of various sized shims, a spare driveshaft to weld up and no dial guage? I got to thinking that there must be an easier way, and this is my idea (at last, I hear readers cry!).

1. Buy a 3 thou bearing shim of the same I/D and O/D dimension as the OEM shims. These are available from bearing shim makers and I will post details and suppliers when things are open next week. Or even a selection of 2 three and four thou ones.
2. Assemble the hub and bearings and tap the inner bearing solidly into place against its cone so no endfloat.
3. Place the assembly onto the driveshaft with NO spacer shim, just the silver oil seal track.
4. Do up the castellated nut firmly but NOT hard, say 15FT Llbs MAX. This ensures that both your bearings are completely seated and are not held apart by grease etc etc.
5. Undo the castellated nut, carefullly withdraw the hub assembly and place it in the wooden jaws of your workmate, inner side upwards.
6. Now very carefully measure the gap with your calipers as shown in the pics above.
7. Order from Jaguar or your favouite supplier the shim of that exact dimension. These are idiotically expensive.
8. Install onto the hub tube and verify that the shim and the bearing are level. To check if level, take a short staight edge and lay it across the bearing/shim. If not quite level, slip a feeler gauge under the part with a gap to verify the amount, then see NOTE below.
9. Tap the hub tube inside the bearing downwards a touch with a suitable drift so as to produce some endfloat.
10. First add the 3 thou shim, the add the OEM one.
11. Now tap the oil seal into place with a suitable bearing drift
12. Install the assembly onto the driveshaft and tighten the castellated nut to spec. This is allowing the oil seal track to squeeze the inner bearing inner race down the hub tube UNTIL the combined hub tube/spacer/shim stack prevents the oil seal track pushing the bearing any further down the tube. Thus there has to be 3 thou endfloat, because the oil seal track is held 3 thou apart from touching the inner bearing inner race.
13. You are now guaranteed to have the correct endfloat and no need to use a dial guage or worry about it being incorrect.

NOTE, the OEM spacer shims come in increments of about 3 thou, so if you have an OEM shim requirement of (say) 126 thou, and there is available only a 125 thou OEM shim, then just use a bearing shim of 4 thou instead of three. Hence the idea to buy a selection.

Comments please!

 

Read it a few times, having NEVER done this, as s/hand GOOD hubs were very plentiful, and cheap ($20) when I needed the 2 over the years.

Sounds a very simple, and accurate, method, as the Manual was complicated, that I do remember.

 

One comment Greg. Your pic titled "Bearing fully drifted into place."

How did you drift it this inner race into place whilst ensuring that you didn't drift it -too far- & close up the clearance too much/give too much preload?
Because if you do do that, if you then fit a spacer, unless the spacer is smaller than the gap between the inner sleeve and the face of the bearing, when you torque up the axle nut you will just be torquing up the assembly without affecting the bearing at all.
Basically you will be sandwiching the axle and hub together, but the bearing will already be pressed beyond this & have too tight a clearance, because that inner race is an interference fit on the hub so it won't 'relax' back to whatever clearance is set by the spacer.

All of that said, I don't even understand why Jaguar specify free float for this - if you read any general guidance on opposite tapered roller bearings, it always says to setup with a preload for max bearing life.
I guess the theory is that if the assembly all heats up, the aluminium of the hub will expand more than the steel hub, which will effectively increase the distance between the outer races of the bearings, so closing the clearance up to 0? But with inboard brakes it doesn't really heat up much at all.

 

Asdrew, thanks for the comment. My ideas annotated in red below
Your pic titled "Bearing fully drifted into place."

How did you drift it this inner race into place whilst ensuring that you didn't drift it -too far- & close up the clearance too much/give too much preload? At that stage I didn't ensure it was not drifted too far in. There is a subsequent step (no. 9 above) which states:
Tap the hub tube inside the bearing downwards a touch with a suitable drift so as to produce some endfloat.".

Because if you do do that, if you then fit a spacer, unless the spacer is smaller than the gap between the inner sleeve and the face of the bearing, when you torque up the axle nut you will just be torquing up the assembly without affecting the bearing at all. I would have been, but read on:
Basically you will be sandwiching the axle and hub together, but the bearing will already be pressed beyond this & have too tight a clearance, because that inner race is an interference fit on the hub so it won't 'relax' back to whatever clearance is set by the spacer. By tapping the hub tube back down you are now producing endfloat, so when you add the shims and tighten up on the driveshaft, they will be squeezing the bearing back down onto the hub tube, and the shims will stop the "squeezing" 3 thou from the point where the bearing is tight against its outer cone.

All of that said, I don't even understand why Jaguar specify free float for this - if you read any general guidance on opposite tapered roller bearings, it always says to setup with a preload for max bearing life. Good point, and on the XJ40 equivalent, I am told that Jaguar specify some preload. Having had this failure though, I wanted to be 100% sure the bearings were precisely to spec, even if the reason for the failure is a mystery.

Further to my first post, I have had the following thought. Setting up endfloat with new greased-up bearings is inherently difficult to get dead accurate, as my own attempts showed, as I gradually closed in on obtaining a repeatable dial gauge reading. What about:

• When installing new bearings you were to buy an extra inner one, and using a flapwheel slightly ease the inner diameter so it became a sliding fit onto the hub tube,
• Install the outer races normally and the outer bearing inner race, and push the hub into the hub carrier (all as described above),
• Then (all bearings greaseless) using the sliding fit inner bearing inner race to set up the endfloat and order up the required shim as described above in post 1,
• Then remove the sliding fit "dummy" inner bearing inner race, remove the hub and grease everything up, replace the hub and this time drive on the inner bearing inner race. Drive it on partially, or if fully by mistake, tap the hub tube back down to produce some endfloat.
• then assemble with the shims and torque down the driveshaft. Job done and no worries with the grease affecting matters.

More comments please!

 

My .02c
I eventually quit reading the how-to's on this and just did it on the bench with a 50 year old dial indicator set in the most simple way possible, pry the thing up and down and measure how far it moves.
There is not imo enough precision in measuring such with a sliding digital caliper on the surfaces one has to work with, I want to see exactly what the movement is.
It was pretty simple and repeatable if a little fussy to set up.
Set it with no grease, it's easier to tell when it moves and it returns to the other position more reliably, once you have your ship grease it up and do the final assembly.
It's worth buying a small upright press for this job, you'll use it for others and there is almost always one cheap second hand to be had.
With Timken USA bearings replacing Timken England, one side was alright with the factory spacer and one was not.
I bought some generic shim stock and cut out what was needed for the one side.
Measure several points around the spacer, they aren't as precision (least not at this point) as one might think.
If I had plenty of shim (and patience) I'd have set it tighter than .002, more like zero or maybe .001 preload if I trusted my measurements.
The outer fulcrum shaft was if anything more trouble to measure and set but the principle is the same, I made shims for those.
There's way too many shims in these cars.

I've had a chance to examine several of these with known mileages now and contamination seems to be way more of a thing than wear, or what causes
wear rather than use. Much as I dislike the idea it is probably a good plan to pull these apart and clean/grease/check play every 30-50K or so in addition
to giving the wheel a shake now and then to check for play. Maybe more often if you drive in the wet a lot. Bout the same for the fronts.

 

Greg,

I'll read your procedure in more detail tomorrow and comment after I've thought about it, but this comment about bearings caught my eye.

I have recently rebuilt the differential in a 1968 E Type and some previous "genius" went through the entire diff and sanded the bearing races. Why, I don't know. Just about everything was overtightened and out of spec in one way or another, so I had to start from scratch in setting up the differential. It had the original Made in England Timkin bearings, so I replaced them with the same number, which are now made in USA. Unlike the later differentials with crush sleeves, the pinion bearing preload is set with shims. So I assemble with a known shim pack, torque the pinion nut to spec and measure the bearing drag to determine if the preload is sufficient. If it's incorrect, everything is pressed apart and the shim pack adjusted.

The USA made bearings were very tight to press onto the pinion. I have done other Jaguar differentials, I know roughly what setting on my press should make the bearings move, and this was nearly triple the usual. In order to disassemble, the pinion is pressed out of the case, and the pressure goes through the rollers - there is no other way to do it. In this case, the bearings were so tight the force flatspotted the rollers and ended up rendering the bearings junk. I measured the inner bore of the bearing and it was nearly 0.002" undersize. In the world of bearings, that's a massive out of tolerance; an order of magnitude greater than what should be expected. I bought them from a local industrial bearing supplier, so it wasn't a dodgy counterfeit sold on Amazon.

I ordered a new set of bearings, this time I bought SKF brand. I opened the boxes and was a bit disappointed. One bearing was Japanese, the rest made in China. I used them and assembled everything. They pressed on very nicely, the pressure on the press was right where I would have expected it to be and I was able to adjust the shim pack without difficulty. The diff is now smooth and silent in the car, with backlash exactly on the etched clearance.

So we seem to have come a full circle, now Made in USA = junk and Made in China = quality, at least for name brand bearings.

 

Wolf
Thanks for your comments, and I do agree (with hindsight!) setting them up greaseless is better, which is why I posted the bit in purple at the end of post No. 4. I am not keen on drifting the inner bearing inner race on and off to grease it after setup, and am trying to see of there is a more straightforward easier method.

At bottom, for whatever reason, I found it a very hard indeed to measure the endfloat accurately and repeatably, so what I am trying to achieve is a method that is easier to do for the home mechanic and more certain with minimum trial and error. When I have my car back together I am going to buy a new set of bearings, plus one to ease as a setup bearing for the inner, and using another spare hub and hub carrier go through the suggested method and see how straightforward and accurate it is. I shall take multiple photos and post the routine ith illustrations - if it works!

I take your point about calipers and dial guages, but the key for me is passing a straight edge across the installed shim and bearing inner,as I describe in Post I point 8, and remember the bearings will have been tightended onto the driveshaft beforehand, as in step No. 4. It is extremely easy to see if one or other of the two surfaces is proud, and thus adjust the extra aftermarket shim to suit.

 

With a bearing separator and a hydraulic press it's really a matter of minutes to assemble or disassemble these things once they are all cleaned up and you figure out what you're using for bracing and pressing.
And the dial indicator set up something like I had it with no grease is super easy to read the play with and very repeatable. I found the key is to take the time to have the tools needed, otherwise it's an annoyance
full of self doubt and inconsistency and dropping things on the floor and cursing. Always more than one way to skin a cat but this seemed to work out pretty well.

 

--------------------------------------------------------------------------------------------------------------------------------------------------------------
I have a copy of the Timken UK article on tapered roller bearings. There is no doubt longest bearing life is achieved with slight preload.
A little too much preload degrades life significantly, the tolerance is tight on correct preload.
If the bearing is set with slight end float the life is not significantly degraded in comparison with preload.
Increasing the end float will degrade life but not as much as a similar increase in preload, meaning the tolerance is wider.
Perhaps for that reason a slight end float is recommended, a compromise between life and ease of fitting.
Also end float should accommodate the effects of heating more reliably.

Many years ago I replaced the hub bearings on my coupe.
I am no fan of interference fits for bearings. I linished the journal for the inner race to give a sliding fit.
I used a bearing grade of Loctite on the inner races and just nipped it all up so the bearings ran concentric and were on the edge of preload.
I have not done a great mileage on the car, but so far no problems.

 

Great idea to use the bearing Loctite with the eased bearing, which gets rid of the need to buy an extra bearing!

 

The Book had a few articles about using sleeve locking compound on the front bearings, they are apparently known to spin the race, and they aren't press fit.

My first failed attempt with I suspect excessive end float on the rear of mine, they spun both the inner bearings inner races(and never made a sound).
I don't think I'd want to encourage that though I'm sure that was install error on my part from trying to rush it.

I'd bought new second hand assemblies to rebuild and swap in since one of mine had the
usual crack anyway and I was suspicious of them. Turns out I was right. The used units I bought were from a 90K ish mile car that had an engine fire,
the original England stamped bearings would have been fine save for water contaminating the grease and them clearly sitting for a long time,
you could see where the rollers that were at either the top or bottom when it was parked corroded and started eating into the race
where they were touching. Drive the cars, it's good for em.

 

There are a couple ways to set end float. I always press the assembly together, I never use hammer around bearings unless I don't intend to reuse them. Once everything is seated (with greased bearings) I made up a collar that sits on top of the bearing. Doesn't matter how thick, as long as the thickness is known. It's simply a spacer to enable measuring using a depth micrometer without the micrometer being obstructed by the seal or the castings. Then you want to measure the distance from the top of the collar to the end of the hub.

I use a depth micrometer, sitting on top of the collar to measure down to the end of the hub. Subtract the collar thickness, add the required endfloat from that measurement and you have the shim thickness you need.

You could probably use a vernier caliper in depth mode, with an engineer square across the collar to give something to measure from, but vernier calipers are not very accurate. The general rule is they are no better than 0.002", no matter how finely graduated they are. A micrometer is inherently more accurate.

 

Thanks Jagboi.
One question: if you have the bearings together fully before you measure, are you setting them up with zero endfloat, or do you release the bearing a bit and allow extra in your shim measurment for the endfloat?

 

Once the hub is fully assembled in the car, the endfloat is the amount of free play the bearing has between the halfshaft spacer and the carrier. If you had the shim sized for zero endfloat, when the axle nut was torqued, the bearing would just touch that inner spacer that goes against the UJ and runs inside the oil seal, and the bearing would be unable to move axially along the axle shaft

Jaguar specified float of 2-6 thou, so when torqued the inner spacer touches the shim and then the hub shaft that goes through the bearings. There bearing then has a little free play between the rollers and races. When measuring for the shim size, you want all the free play to be taken out of the bearings and rollers tight against the race - not excessively so, but firmly. Then you can accurately know how much clearance will be in the system when the axle nut is torqued.

I use my press to assemble everything to ensure the races are fully seated in the carrier, and you also want the inner bearing in it's final position on the hub shaft. You don't want things to be moving after everything is assembled, or your shim measurement will be for naught. You need to be able to calculate an "in service" end float dimension.

Once everything has been assembled and put together, it seems to stay in place once the press is released. The main thing is getting the components seated and into position. Then I measure the shim thickness required. Long winded, but I hope that helps?

 

Jagboi
I think you misunderstood my question. You originally wrote:
"I always press the assembly together, I never use hammer around bearings unless I don't intend to reuse them. Once everything is seated ..."

And in your latest post you wrote:
"When measuring for the shim size, you want all the free play to be taken out of the bearings and rollers tight against the race - not excessively so, but firmly"

Under these circumstances, there would be no endfloat. Then you measure for the shim, so the measurement you make would, therefore, produce no endfloat unless the measurement had added to it (say) 3 thou and this 'measurement plus 3 thou' was used as the shim size.
Then even with the extra 3 thou added, the bearing inner would still be against the outer race with no play, unless the hub tube was tapped down from the bearing to produce some endfloat.
Then, as I understand fully, mounting it all on the splined driveshaft and tightening it down would squeeze the (released) bearing to within (say) 3 thou of the outer bearing cone and you have the endfloat needed.

So my question was and still remains:
"do you release the bearings a bit to achieve this, as if you do not, the endfloat "gap" would be between the oil seal track and the bearing inner race, rather than between the bearing inner race and the bearing outer cone?"

Or does it not matter and the bearing inner race works itself looser (ie towards the shim in spite of its interference fit on the hub?

This same point was made by Asdrew in post 3 above, and maybe his formulation is clearer!:
"How did you drift it this inner race into place whilst ensuring that you didn't drift it -too far- & close up the clearance too much/give too much preload?
Because if you do do that, if you then fit a spacer, unless the spacer is smaller than the gap between the inner sleeve and the face of the bearing, when you torque up the axle nut you will just be torquing up the assembly without affecting the bearing at all.
Basically you will be sandwiching the axle and hub together, but the bearing will already be pressed beyond this & have too tight a clearance, because that inner race is an interference fit on the hub so it won't 'relax' back to whatever clearance is set by the spacer."

 

When you're about to measure you don't press the assembly fully together, leave a little feel-able play. Tighten the nut to snug it the rest of the way down and measure.

I didn't find any measured difference in several checks of the nut at 50 vs 90ftlbs with everything all clean and polished up so no need to kill ones self trying to tighten to full spec on the bench.

I wouldn't go so far as to call this a trivial job but a small upright hydraulic press and a selection of sockets and race presses and such make this a pretty easy job once you see how it works.
Not having a selection of shims on hand is the biggest annoyance if you're in any kind of a rush. The outer fulcrum bearings were probably more annoying but you can usually buy bulk
generic shims that will or can be trimmed to fit them well enough at least.

 

I believe that is correct, yes. Perhaps we need to make a distinction between the cup and cones themselves, and the bearing system, which would include the hub carrier and the axle shaft and it's spacer/adapter. I press the bearings into the carrier, with the hub in place until everything is tight.

No. Bearing stays tight. What we are after is when the bearings are all seated, the shim sitting on the hub tube end is 3 thou proud of the bearing race.

It's impossible to drift in too far. I assemble the outboard bearing/hub carrier and hub assembly first, flip it over and out on the inner. It will only go as far as it can go, the rollers will contact the race and stop.

If you have a copy of the Jaguar World book on the 6 cylinder XK rebuild book, it also included rebuilding the IRS and goes through the procedure with photos.

 

Here was the description of the rebuild.

 

It is possible if you have a spacer that will land you on negative end float. If you can start with (or make) an oversize shim and go fully together, measure the float, do the math and you'd be done, great.
When you're re-using existing shims it's best to not press them all the way and run them snug with the nut while observing if you run out of float(XJS has the nut, earlier do not).
I wouldn't do it the way that article suggests unless I wasn't equipped to do otherwise, much too fiddly and prone to error. Press and a $50
dial indicator with a mounting arm is all that is needed.

 

Jagboi, Wolf
Thanks for the replies.
It seems to me we are all on the same page. If I grasp what Jagboi is saying correctly this is the position:

As long as upon assembly there is (say) 3 thou endfloat somewhere (either between the seal track and the inner bearing inner race, or between the inner bearing inner race and its outer cone) all will be well, as in the case of the former the inner race will adjust itself away from its cone in use to produce float between it and its cone.

Is this your understanding too?