Fitting an Ashcroft ATB to my Salisbury axle

[Peaklander]

I haven’t measured it yet. I just put it in. That will be in the morning after I fit the pinion. First I have maybe to use threadlock and torque the CW bolts. Please see my question about that above.

@hurbie yes I understand that. I’m not reusing the bearings but what you say makes sense. 

Edited February 27 by Peaklander

[Snagger]

I’d tighten the ring bolts sequentially to full torque for the runout check, but would defer to Nige’s view on that.  Any difference it makes compared to tightening the bolts quickly with just the ratchet might be very small, but it doesn’t take much to be out of spec on a job like this.  Given the effort you went to on the pinion and in buying the puller, it’d be a shame to cut corners on something relatively simple.

[Snagger]

6 hours ago, Peaklander said:

I haven’t measured it yet. I just put it in. That will be in the morning after I fit the pinion. First I have maybe to use threadlock and torque the CW bolts. Please see my question about that above.

@hurbie yes I understand that. I’m not reusing the bearings but what you say makes sense. 

Did the puller get the bearings off without any apparent damage?  It’s not ideal that it pulls the cage rather than the race, and while that wouldn’t matter for worn out bearings, it could if it damages good bearings that had been intended for reuse.

[Peaklander]

10 hours ago, Peaklander said:

Do I torque-up fully without thread lock, check run-out and if OK then remove one bolt in turn, apply the red stuff and re-torque that?

Or do I have to do the full sequence which might affect run-out if it is fully released?

That was the question; I'm not intending to miss-out the torque, just wondering whether it would be OK to release one at a time to apply thread-lock. Maybe the best would be to assume the run-out will be ok, apply it and if they need undoing, the thread-lock won't have set.

It seems to have been very kind to the bearings; there is no obvious strain. I will post some pics when I get outside in a bit.

[Snagger]

5 hours ago, Peaklander said:

That was the question; I'm not intending to miss-out the torque, just wondering whether it would be OK to release one at a time to apply thread-lock.

That’s what I’d do.

[blz6662]

Hi All,

I'd just like to say what a great post this. I will be undertaking the very same task on my Salisbury shortly so this is all extremely helpful.

Anyone got a spreader for sale?

Cheers

Baden

[Snagger]

7 minutes ago, blz6662 said:

Hi All,

I'd just like to say what a great post this. I will be undertaking the very same task on my Salisbury shortly so this is all extremely helpful.

Anyone got a spreader for sale?

Cheers

Baden

You can do it without the spreader.  Takes a bit more effort, and you need two levers getting the diff out as it needs to be kept square without the spreader.  I have done it both with and without.

[Peaklander]

I inserted the ATB again using the dummy bearings, with the crown wheel fitted, following some work with the oil stone and brake cleaner, having torqued the bolts to spec., a mere 58Nm. Then I measured the run-out. It is variable, as much as 0.11mm over one revolution in one direction yet only 0.05mm in the other and the spec. max is 0.05mm. I can only assume that the dummy bearings are interfering with this. Although it does feel a bit strange to do this by the book method, where it says measure without the bearing caps fitted. So I'm not sure about that.

The new Timken pinion inner bearing is on the shaft. Freezer / warm stove-top and a press that is not so good but it went on OK. I needed a bit of pipe but the steel I have has a seam on the inside (scaffold pole) and it was that much undersize. With no way of grinding it away on that length I resorted to aluminium. It worked.

Yes my press is rubbish, the jack leaks in spite of my efforts to reseal it and that downward pressing bit onto the workpiece is cock-eyed. I need to improve it

 

The undersize pinion inner cup used for set-up is still in the case with the previously installed shim and the pinion now in again for another test fit, with some pre-load, just enough to make the flange slightly stiff to move by hand.

I'm now at the point where I will fit the shims to the ATB. The endfloat this morning measured 98 thou and with 5 pre-load to add, that gives me 103. I now need to use the spreader, hopefully it will help although I know I need to be extra cautious. The axle is on stands on the bench and pulling and prying isn't that easy with many Kgs trying to fall over. It is extremely heavy. Just the ATB alone seems heavier than the Salisbury it is replacing. Edit it is 22Kg without the bearings.

There is some concern about the relative size of the ATB 'stubs' and the Timken bearing IDs. The bearing ID is 2-1/4" or 57.15mm.The difference is 0.8mm / 3thou." I feel a bit better as I have compared the ATB OD with the Salisbury, now that the bearings are off and the ATB is about 57.22mm and the old one 57.25mm. That's with a vernier that just fits across. So I think I will be ok. Maybe the ATB will need to spend some hours in the freezer, just to make things a bit easier.

 

 

Edited February 28 by Peaklander

[Snagger]

Your press may be imperfect, but at least you have one!  I had to remove and refit the bearings on the Rover diff I built up two years ago using a hammer and drift - I couldn’t even find a length of pipe.  Trying to phone engineering workshops here is a fool’s errand.

[Green Bob]

This is a very informative thread and I am very glad I paid someone else to fit mine!

[Peaklander]

I have got stuck. Not literally but stuck at the point where I am trying to locate the ATB with the shims and dummy bearings.

Hopefully I'm not repeating words I've written earlier - just to say that I removed shims on CW side 30+30 and opposite 30+9+4+2. (I had thought those last ones were both 3  but they are definitely not the same. So that was 105.

Before the pinion went in this morning I measured the end-float again at 98 and so aimed for 98 + 5 = 103. I know the 'reset' instructions are to put it all at the CW but I couldn't see the point in that as I have a reasonable idea of the split. So 30+30 went in at the CW as it was before and 30+9+4 on the other side.

It won't go in at that. Now I am using the spreader but I am being 'super' careful. I marked the flats so I can count. The Book says a max of three flats after the nuts are finger tight but it won't go. It gets a long way across the axle tube, isn't getting stuck on the angle (I can pull it back out) but it reaches a stop and it isn't the pinion. It will go in if I remove 30 but then of course it's loose!

I am doing what I need to do and calming down in the house. Perhaps I should stretch a bit more... Any ideas???

[Hybrid_From_Hell]

13 hours ago, Snagger said:

I’d tighten the ring bolts sequentially to full torque for the runout check, but would defer to Nige’s view on that.  Any difference it makes compared to tightening the bolts quickly with just the ratchet might be very small, but it doesn’t take much to be out of spec on a job like this.  Given the effort you went to on the pinion and in buying the puller, it’d be a shame to cut corners on something relatively simple.

100% agree

 

check crown whe has no burrs on inside of bits of old loctite poking about clean clean clean wipe ATB as well with acetone or brake cleaner

 

assemble and bolt up in stages and do it like a cylinder head not go around the circle 3 stages of going up is fine 

 

then go around and double check each bolt is up to torque 

and loctite it as it’s done and there isn’t much else you can do there so go for it clean clean clean

[Peaklander]

Does anyone know how much the case should spread? I assume it has to at least move by the pre-load amount of 5 thou. I don't know how much grunt to put on my spreader. Maybe I should measure the opening as I do it and find out.

Edit: yes Nige, I have done exactly that on the crown wheel assembly to the ATB.

Edited February 28 by Peaklander

[Peaklander]

Right well I am very confused and need a break.

If I spread it by 0.005" using the dti across to one side, it feels very hard work on the spreader I have to reach that. I read on the interweb that a Dana 60 case can be spread up to 0.015" which would be way more than needed and a tough ask on my spreader - which I thought I'd managed to create as heavy duty! Can a Salisbury really be spread that much? So much stuff says just knock it back in with a hammer. There's no way mine will go that way, it doesn't sound as though it is moving once it is in about half-way.

I was very careful with the CW mounting, oilstone, break cleaner etc. Surely I don't need to re-visit that. It must be the lack of spread. 

[hurbie]

the Land rover manual state's this : 

[Snagger]

I would spread it until the diff will drop in without needing a hammer, but only just, and with the diff wiggled very slightly to make sure that it gets to align square and slot in.  The first few flats on the adjuster under load are going to be the flex in the spreader, not the casing.However, the manual comment about three flats on the nut are specific to the LR special tool, and would depend entirely on the pitch of the threads and the distance from axle centreline to threaded rod centreline.  Any other spreader would have different geometry.

[Peaklander]

That’s useful detail. I’d remembered the three flats and thought it would be spreader specific but I didn’t remember that it says 0.012”.

If I can spread just a little more it will be ok, I’m sure and now I have that leeway.

Thanks both. Onwards…

[Peaklander]

It's gone in finally, with a struggle. I wound-up the spreader and took it to 6.5 thou. When it didn't drop in I did it some more. It would easily just drop most of the way but I finally resorted to a few copper hammer blows and it was in. That was much harder than I expected. This is still with the dummy bearings on of course. I know they are 'right' so it can't be that. Some EP80 on the sides definitely helped.

The backlash was over spec at 0.40mm / 16 thou. (spec. 6-11) so out it came, with a bar and spreader. I moved the 4 thou shim (which seems to measure 5 today) across to the CW side and in it went, same method but at least it was quicker.

The ATB by the way, with CW fitted, weighs 22Kg. It's quite a workout as my workbench is high and I need to stand on a step to be sure I'm at the correct angle. If you have ever 'tweaked' something that then takes six weeks or more to calm down, then you'll know why.

So now the backlash is between 11 and 5 as I measure in four quadrants. The runout is negligible. I will go and fasten the bearing caps on just to be sure nothing changes but it looks as though I need to move a little more shim to the CW.

Now I have 30+30+4 on the CW and 30+9 opposite, still maintaining 103 total. There's a selection of 3 / 5 / 10 new shims available to use and an original 2 which is slightly damaged but maybe would be ok.

I could go to 30+30+5+2 and 30+3+3 to keep 103 total.

Perhaps before I do, I could check the contact pattern. If it looks horrible then I will probably make a pinion height change and that would change the backlash anyway.

[Peaklander]

So, I have looked at the contact pattern and will post some photos.

The situation is that the pinion is in, now with a Timken bearing and the shim that was there is now under a Timken cup, ground undersize so that I can remove it if needed. That will be changed to a pressed cup at the end. The book suggests the pinion should be over 10 thou lower than it is, due to the +9 marking, so that is more out of mesh. However, give or take the Koyo - Timken change, it is in the same position as before.

The ATB is in with dummy bearings and shims as detailed above. There's now a total of 0.103" and there was 0.105" with the old final drive fitted. The bearing caps are fitted but not torqued.

Backlash is under 0.011"; it varies a little on the rotation.

There's no measurable run-out.

Here are the photos. I have posted several as the shadows and angles really affect what you can see:

Drive side (convex)

 

 

To my novice eye, that looks good. Central up and down and not on the edge or down in the root.

 

Coast (concave)

 

 

I'm not so sure about these coast markings, they might be too far to the toe (inside).

I believe that pinion depth moves the pattern between root (deeper) and tooth (shallower) but it also moves the drive and coast between toe to heel (inside outside) in opposite directions. So moving the pinion to get the coast mark towards the heel, would bring the drive mark towards the toe.

Backlash moves the marks in the same direction. Tighter mesh (higher pinion) moves to the toe and looser (lower pinion height) moves to the heel.

Trouble is I think my drive pattern is good as it is!

I need grown-ups. So I won't touch anything and wait and see what you all suggest...

 

Edit: I think if I do anything, that the drive is towards the Heel and the coast is towards the toe. Therefore I believe the pinion could come up a touch, more into mesh.

 

Edited February 29 by Peaklander

[Bowie69]

That looks pretty good to me, and your assumptions about how it all moves with adjustment are, I think, correct.

[fmmv]

So would you expect to adjust pinion to get drive and coast  more or less equally placed, then adjust the CW to get them in the right place? Thats just inferred from what you posted, not knowledge.

[Bowie69]

They are all inter-related, if you adjust CW you will push it more/less into the root, but will also move it out/in on the crownwheel due to the taper.

[Peaklander]

Yes. I think increasing the pinion height  (more mesh) should do that but where they then sit (together) I don't know. As it will reduce backlash anyway (and that is at the top of the range now), moving things towards the toe, I might be lucky.

[fmmv]

I guess you are aiming really to get things pretty close to where they were when you started.

[Snagger]

I think your shims are spot on.  I’d recommend you now just substitute the dummies for the final bearings and run another pattern with everything torqued up.  If it comes out similar, you’re done.

As I mentioned before, the pattern diagrams mention in the top “correct” line that the coast side is often shifted a bit towards the toe, but as long as the imprint is centred between the root and peak (ie not in the bottom of the valleys or just touching the tips of the teeth), then it’s good.  Used gears tend to shift the coast side towards the toe, I read somewhere, too.  I honestly think you have it nailed.