input shaft nut torque

[doffo25]

Hi All

Im currently rebuilding the irs on my '72 roadster and am now attempting to replace the pinion oil seal which is leaking. Reading the various posts the job appeared fairly straight forward. I found the largest spanner I could to undo the input shaft nut (having marked its position), jammed the output shaft and braced myself.......
...and the nut undid with such little effort that I cannot believe it is correct. (In fact I dont think the torque would register on a torquewrench.)
The car has 56K on the clock which appears to be correct and the irs doesnt appear to have be messed with. Unfortunately I have never driven the car on the road so dont know whether there was an existing problem.
To complicate matters I then mistakenly tightened the nut 1/8th turn past its marked position.
So my question is could it be correct that the nut was so un-torqued and secondly will my overtightening have compressed the collapsible spacer (or is it elastic?) and so the diff will need to be set up?

Any crumbs of help would be much appreciated please!

David Offord

[mgcjag]

Hi David....you have 2 options.....take the diff out and have it striped and checked......or just go with it.....there isnt a torque figure the nut is tightened to compress the crush spacer......so undoing it could be easy or not,...do you have a castlated nut and split pin.....doing the nut back up is just turning it till you reach the mark you made.....did the extra bit of turn take much doing if it did then you probably squashed the spacer a bit more.....they do not spring back......if it were mine i would run with it.......worse that can happen is you will get some wear and need a diff rebuild......all the best .. Steve

[mgcjag]

Should have added, if the diff is out of the car then get it checked over......if its still installed tben go with it

[doffo25]

Thanks Steve for your reply. No, there is no castellated nut...tho the nut looks like some sort of anti vibration (?) nut. The nut appeared to continue to tighten at a constant torque. So presumably the collapslible spacer compresses to provide the required bearing preload? Just seem odd that it takes so little torque to compress and yet some threads on earlier cars talk of doing the nut up to mega lbsft.

regards

David

[mgcjag]

Hi David....no the spacer does not colapse to provide the required preload.....it does take some force to crush it, so undoing it should be much easier........doing it back up should again be easy until you meet the spacer.....then it becomes harder as you crush the spacer more...but you dont want to get to that stage, thats why i asked did it do back up easy to the mark and was the last 1/8 turn easy or hard if easy you probably didnt crush anymore....problem is its done now and you need to make the decision.....if the diff is out then check if.....if its not?

Ps.....are you sure you done enough revolutions of the nut...your mark line would line up several times till in the right place

[doffo25]

Yes Steve pretty sure I had tightened nut correct no of revs. The extra 1/8th turn was as easy to turn as tightening up to the mark. Is there any harm in tightening nut up until resistance is felt ie up to the spacer?

[mgcjag]

Hi David......the problem is how easy to do up is easy.........i have changed the input shaft seal befor but only tightened upto the marks......how about a bit of background.....you say your rebuilding the irs....so the diff is out....was the car a driver when you hot it.....how easy does the nut turn, by hand or 1 finger on a spanner, if so you could revolve it a bit more till you feel a stop.......but if this loose why is it going past your marks......its all a bit of a gamble.....if you want to be sure get it checked out......its a lot of work rebuilding the irs and haveing to take it out again.....have you also changed the output seals, if you have then definatly take it and have it checked,... i replaced the o/p seals twice and they still leaked till i upgraded the shafts....

[angelw]

Hello David,
Steve is correct in that the collapsible spacer doesn't crush to provide preload, but incorrect that there isn't a specific torque range to apply to the spacer; its 120 to 140 lb. ft.

It states in the repair manual that if the Input Shaft nut is inadvertently tightened beyond the aforementioned torque, the system should be dismantled and the collapsible spacer replaced with a new one.

Overall, this system using a collapsible spacer is flawed and was introduced for the purpose of speedy assembly, not because its the best method. Please read on.

If the Collapsible Spacer were to be omitted, the Inner and Outer Pinion Shaft bearings would be drawn together when the nut securing the Companion Flange to the Pinion Shaft is tightened. There would come a point where the rollers of the two taper bearing are in close contact with their respective outer bearing race and no end-play would exist. Any further tightening of the nut would result in less than no clearance, which is preload. In fact, when determining the correct position of the Pinion Gear in relation to the Crown Wheel, the Pinion Shaft is assembled without the Collapsible Spacer.

You will note from the attached drawing that, when assembled, the Collapsible Spacer is resisted by the shoulder of a fixed component of the Pinion Shaft and that the Collapsible Spacer is located between the Inner and Outer Pinion Shaft bearings. Accordingly, rather than the Collapsible Spacer applying a preload to the two bearings, it attempts (and does a good job) to keep them apart. Correspondingly, in a Perfect World situation, the Collapsible Spacer should collapse to a length, when the nut is tightened to 120lb.ft. to 140lb.ft., that brings the Inner and Outer Pinion Shaft bearing together until there is no end play. How likely is that? Not very.



There are so many Stars that have to be aligned for that to happen:

1. The ductility of the material in the spacer being the same as that when this flawed system was thought up.
2. The wall thickness of the spacer being the same as the design size.
3. The amount of initial collapse (bulge in the centre of spacer) being exactly as the design meant it to be.

In my shop we use a solid, non-collapsible Master Spacer that is of a length that will ensure measurable end-play when the companion flange nut is tightened. The end-play is measured using a dial-indicator. Given the known length of the Master Spacer and the end-play, its simple to determine the correct length of a spacer to give preload.

Earlier Jaguars had the preload set by using spacers of varying length until a specified torque was reached to rotate the pinion shaft. That is how I initially determined the amount of preload to apply to the Input Shaft bearings.


Best regards,

Bill

[doffo25]

Thanks Bill for the detailed info and great sectional diagram. How much torque roughly does it take to compress the spacer? When I inadvertantly went past mark by 1/8th turn I had single finger on spanner. If I put 120lb.ft on it I would go way way past the mark. Could this "loose nut" be simply due to the 56Kmiles of bearing wear ? Do you know what type of nut the pinion nut is and what prevents it coming loose? And no Steve Ive not changed any of the seals yet. Tho Im a risk taker (car was bought from New York unseen via ebay) Im feeling out of my depth on this one and dont fancy a diff rebuild in a years time......maybe I should get it checked over......

best

David

[Hugo]

I'm no expert on these diffs but I would not feel happy having the nut as loose as you describe. Marks or no marks, that doesn't sound right. If it were mine, what I would do is feel the free play in the pinion. Presumably it will rotate freely for a fraction of a turn of free play before the teeth hit? I would then tighten the nut up properly and see what it feels like. If it has now got stiff, you have a problem & will have to strip it & start over. I would take that risk rather than the bigger risk of leaving the nut loose & hope for the best.

[angelw]

Hello David,
I don't agree with many of the procedures and comments made in the Jaguar Repair Manual and clearly, there are many mistakes printed therein. For example and on the subject that you're most interested in, it states that "As the collapsible spacer preloading the pinion shaft taper roller bearings must be replaced each time the pinion flange securing nut is released", their reference to the collapsible spacer preloading the bearings is totally wrong. This spacer resists the two bearings coming together; accordingly, it applies no load whatsoever to preload the bearings. Rather than preload the bearings (for this the collapsible spacer would have to be on the outward side of the bearings, not between them), the collapsible spacer resists the two bearings coming together and would only effect the tension applied to the pinion shaft, therefore aiding in keeping the pinion shaft nut from undoing.

The nut is a self-locking nut, with an area at the outer end of the nut that locks it. Accordingly, this nut is a one time use item.

I'm not sure what torque is required to start collapsing the spacer. I've not tested that, as in my opinion its a flawed system for the reasons given in my previous post and therefore, not interested in employing it. Subsequent to training as a coach builder, I trained as a mechanical engineer, working for the most part in component design and CNC application and methodology.

One test I did do with a scrap diff housing and pinion shaft, was to assemble a shaft and with new bearings to determine at what point the collapsible bearing crushed to give no end-play; 240lb ft (twice that of the low end torque specified in the repair manual) before there was no end play measured using a dial indicator). At 140 lb ft, there was still approximately 0.4mm of end play. Another thing that I've found, is that any collapsible spacers I've purchased that have been current manufacture, resist collapse more than NOS units I've tested (using a press).

I still put to you that its improbable that the collapsible spacer system will crush to the extent that the bearings will be under optimal preload on a consistent basis. You may get lucky and then again, you may not.

Regards,

Bill

[doffo25]

Thanks all for your much appreciated help.....if the design is flawed and I know I am as well ,,,,I will take it to a pro to check over and let you know the outcome...

best

David

[angelw]

Hello David,
Most will simply follow the procedure described in the repair manual and if all you're doing is replacing the front oil seal, you could do the same yourself by using a new collapsible spacer and torquing the companion flange to 120 to 140lb ft (use a new lock nut). However, if you do it this way, I would advise that you check if there is any end-play in the pinion shaft using a dial indicator. The preferred setup of tapered roller bearings is with them in pre-load and definitely not clearance.

In manufacture, all components are made within a tolerance range to allow random selection of mating parts to be assembled to form a working unit. The majority of parts will be made to mid tolerance, with a lessor number being just inside the lower and upper limits of the tolerance range. Accordingly, most assemblies result in a design fit of parts. However, using a large sample, its possible that an assembly, such as a motor vehicle, could be put together using parts randomly selected that are all the largest shaft and smallest bore components within a tolerance range, or smallest shaft and largest bore components within a tolerance range, resulting in assembles that are verging on seizing (or long lasting if they survive without seizing), or worn out from the day they are assembled. This is how the really great new cars and lemons respectively are created.

So, following the thread that parts are made within a tolerance range, its easy to see that its highly likely that the assembly fit of the taper bearings of the pinion shaft can be unreliable when the crush of a spacer at a particular load is used to control the fit of parts.

Regards,

Bill

[Series1 Stu]

So, following the thread that parts are made within a tolerance range, its easy to see that its highly likely that the assembly fit of the taper bearings of the pinion shaft can be unreliable when the crush of a spacer at a particular load is used to control the fit of parts.

Regards,

Bill

Sorry Bill but I completely disagree with this. When engineers design components they (we) do so with a view to them always fitting together and working wherever each component lies within it's tolerance band. There should be no problem with the fit of parts as long as they all fall within their tolerance band. This is the essence of engineering.

I'm sorry for this admonishment but it is important that people understand that things don't just happen to fit together and work by sheer luck. There are highly skilled, competent and dedicated people behind just about everything that you come into contact with in your daily life.

Regards

[angelw]

Hello Stuart,
You can wish in one hand and shit in the other and see which one fills up the quickest.

I had been a consultant engineer to more than one car manufacturer in Australia for many years and I fully appreciate and understand the implementation of tolerances in the allowance for fit in the random selection of components that go together to make an assembly. The fact is, that if though the randomness of putting the biggest shaft in a tolerance range with the smallest bore within its tolerance range, there will be slight different in the operation of the assembly compared to when parts at the opposite ends of the tolerance range are put together.

If in the case of the Pinion Shaft assembly in the Diff Housing using a collapsible spacer, the distance between the shoulders in the Diff Housing that the outer race of the taper bearing locate is at the bottom end of the tolerance range and the shoulder of the Pinion Shaft that the collapsible spacer locates against is a max metal condition, then the pre-load, or clearance between the bearings is affected. Couple with that the tolerance of the collapsible spacer (initial length, the wall thickness of the spacer - (they are not machined diameters), initial collapse and ductility of the material) you have a system that is rather random in its resulting assembly.

In the Jaguar repair manual, its stated that the collapsible spacer is used to pre-load the bearings (I assume that one of Jaguar's engineers proof read the document before it went to press), how can pre-load be applied by a device that is between the two bearing and where its final length is affected by more than one characteristic? Rather than apply a consistent pre-load, this system is attempting to effectively diminish pre-load by keeping the opposing components apart.

Its a different story where it can be guaranteed that a solid spacer is of an exact length (within a very small tolerance), or in the case where shims are used to adjust the distance between two items. In such cases, a predictable pre-load, or clearance can be expected.

I've carried out many tests using off the shelf collapsible spacers being supplied by all the well known Jaguar parts suppliers and I've not had one where other than excessive end-play results.


Regards,

Bill

[Series1 Stu]

Bill

I'm not saying that the collapsible spacer is the best possible engineering solution but it is a compromise made in order to meet production demands. This compromise will have been properly designed and toleranced taking account of all the variables present in the system, including material properties.

I also think that pre-load will not be present during use, even if it is used in the initial set up. The reason for this being that there is always a tendency for bevel gears to try to push apart when running. If they can't do this then they will seize. These forces may cause further collapse of the spacer and provide a suitable amount of end float. After all, opposing taper roller bearings are usually set up with an amount of end float. See your wheel bearings.

I also think that the reproduction parts available from our usual suppliers are, in the main, complete rubbish. There is little evidence of engineering prowess in anything they produce but lots of evidence of corners being cut and a "near enough" attitude. In general, they are to be avoided.

So maybe we're at crossed purposes because you're trying to make a decent fist of something using crappy aftermarket parts.

Regards

[Hugo]

It's not a compromise - it's a cheat, to save on assembly times. I work with Morris Minors also and they have adopted a similar crush sleeve. There are numerous reports on the forums that the preload is reached long before the torque on the pinion nut is reached. You have to set the nut to a lower torque to make it work (120 against 140 I think). I just bought a reconditioned diff for my wife's Minor and you can tell just by turning the flange that it is too tight. I queried it with the seller and he assured me it was done by the book, which probably explains it.
You are also incorrect in your assertion that when in use the spacer will collapse further, causing more end-float. Once the nut is done up, the end float (or rather pre-load) is set. You could take the crush spacer out and throw it away at that point and it wouldn't make any difference. It is only there to provide a resistance as you tighten the nut, so that when you reach the pre-scribed torque, the pre-load will be correct. That's the theory anyway. Once it has done that job it is redundant. The only way to increase the end float is to loosen the nut.
Personally I would never use one of these wretched crush sleeves. I would take the trouble to shim it up and bolt it up solid. But then, I'm not on a production line.

[angelw]

Hello Stuart,
The most favorable application for a Tapered Roller bearing is in pre-load, not end-play.

Don't start me on wheel bearing application, particularly where front stub axles are concerned, or we may be here for a long time. The overwhelming number of front wheel bearings removed will reveal that the inner race of the inner bearing has been spinning on its journal of the stub axle. The allowance for fit between the Stub Axle journal and the bore of the inner race is to facilitate the ease of assembly and disassembly, not because it’s the best design fit. For my own vehicles and for clients I will service in the future, I fit stub axles were the bearings are a light interference fit on the journal.

Hugo is correct in what he has written and what I have stated in this Thread since day dot, and that is the collapsible spacer doesn't provide pre-load to the bearings, as is incorrectly stated in the repair manual. The collapsible spacer system is a cheap, nasty method derived to hasten assembly.

With regards to “I’m trying to make a decent fist of something using crappy aftermarket parts”, aftermarket components of this collapsible spacer is all that’s available. And no, I don’t try and make a decent fist of anything using crappy after market parts. If the concept is correct and only crap parts are available, we make high quality replacements in my own CNC equipped machine shop. If the concept is crap, as is the collapsible spacer concept of the E Type diff, then we improve on it and make high quality parts for the purpose.

Regards,

Bill

[Hugo]

Simple enough to make a proper spacer and shim it up. That's certainly what I would do if I ever had to overhaul a diff.
I am slightly amused by the semantic debate about whether the crush sleeve 'provides pre-load'. I knew exactly what they meant by it - I hadn't even spotted their mistake - and I suspect the authors never expected anybody to dissect their language to such an extent!
Still, accuracy of language is important. I guess Jaguar owners are more literate than most others!