To Other TR6 Pages

November 22, 2014

Differential

[Click the pics for a better view]

I've always been terrified of differentials.  I was intimidated by all the ominous warnings about the specialized tools needed, and the very precise clearance and preload specs.  For this project, I decided to suck it up and at least dig into what I'd need to tackle the job.  The manuals I had didn't really ease my trepidation much at first.  Some of them started off by advising against attempting the rebuild, and then launched into a bewildering series of complicated steps.

I persevered, though, and finally deciphered and understood the logic behind the processes.  With better understanding, I was pretty sure I could devise substitutes for the factory tools, and have a better than even chance of being successful.  So here is the story of this journey.

The TR6 differential is a 70 pound lump of cast iron crammed with beefy gears and bearings.  It is mounted in a fixed position on the frame, and delivers power to the rear drive wheels through articulated half axles.




Disassembly starts with removing the rear cover and the inner axle shaft assemblies.  Inside, we find a 37-tooth crown wheel driven by a 10-tooth pinion still hidden deep inside.  These two gears provide a 3.70 gear ratio, proving my Haynes manual, which claims the ratio is 3.45, wrong.




The inner differential carrier assembly is what allows the two rear wheels to turn at different speeds.  It is held in place with two bearing caps.  The manuals say to heed the markings on the caps and matching ones stamped onto the case flange to ensure they go back in their original positions.  I assume this is because bearing pockets in the housing were machined with the caps on place.  On my unit, which I'm pretty certain has never been apart before, the caps were reversed, judging by the markings.  The markings on the other side are the reverse of that in the pic ("4" on the cap and "7" on the housing flange).




After the caps are removed, the differential carrier still won't come out because it is "pinched" in the case with an interference fit.  This effectively preloads the differential carrier bearings so there is no play.  To relieve the preload and unpinch the differential assembly, the case has to be stretched.  There is a factory tool for this, but many rebuilders fabricate their own.  Mine consists of a bar screwed to each end of the case, and some threaded rods arranged so that turning nuts on them forces the bars apart.  The tool only has to distort the case by a few thousandths of an inch.  When it does this, the differential caririer can be pulled out.




The next major part in the case is the pinion gear.  It runs in two good sized tapered roller bearings, and can be pressed or tapped out after the drive flange is removed.




With everything taken apart, I ordered new bearings and seals, and started working on the case.  I masked the openings and machined surfaces and blasted the rust and grime off of it.




Looking over the clean case, I noticed that there was a boss where a drain plug would logically go, but there was no drain.  This seemed odd, so I decided to finish the job that Triumph saved a few pennies by not doing.




I painted the case with a high operformance urethane paint (POR15), and while waiting for it to dry, I blasted the aluminum rear cover.




There is a potential trouble spot on the rear cover in the vent arrangement.  The vent is important because without it, higher temperatures will increase the pressure within the sealed case, which can force oil past seals or gaskets.  The stock vent is just a disk with a small hole with a cotter pin loosely in it.  Apparently, the cotter pin's job is to rattle around and keep the hole open.  This doesn't always work, and the vent can become clogged with road grime.  This is the vent with the cotter pin removed:




This seemed like another opportunity for a simple upgrade.  There are purpose made vents designed for differentials and transmissions.  They are much less prone to clogging and even come with in internal filter.




I removed the vent disk and opened up the existing vent hole and threaded it for the new vent.




Then powder coated the cover and installed the vent.  I'm not sure if the original cover was natural or painted black.




By this time, the replacement bearings had arrived, and I started with the pinion assembly.  The two critical adjustments for the pinion gear are its fore/aft position, and the preload on its two tapered roller bearings.  There is an interesting thing about the fore/aft adjustment.  The factory set the pinion position with shims directly adjacent to the pinion head--between the head and the rear bearing cone.  The ring on the left in the left pic is a 0.060" factory shim.  However, these shims are apparently no longer supplied by most of the regular parts suppliers.  For rebuild or repair work, shims are supplied to go under the rear bearing's outer race.  While either method moves the pinion position, the factory method does it without affecting the preload, since it just moves the pinion, not the rear bearing.  I'd be interested in hearing the reasoning behind this situation.  I decided to install the pinion with just the factory shim, then check pinion position.




The pinion preload is set by adjusting the distance between the two pinion bearing cones.  Most of this distance is provided by the spacer cylinder, while fine adjustment is done by adding shims.  To install the pinion assembly, the two bearing outer races must first be installed into the case. The pics below show my version of the race install tool.  It keeps the races square and pulls them both into place at once.





In the attempt to install the pinion and set its preload, I hit the first real snag.  With the pinion shaft installed with the new bearings, I wasn't able to get the necessary preload.  With the original preload shim pack and the pinion nut tightened to 100 ft-lb, the pinion still had significant end play.  With the shim pack removed, I could get the end play down to essentially zero, but could still not get anywhere near the preload spec.  

After a lot of head scratching and consultations on some TR6 forums, it seemed obvious that something was different with the bearings.  On close inspection, I found that the new front bearing had a much thicker internal race than the original bearing.  




The original bearing was marked with part number 15100SR, while the replacement bearing, supplied from one of the major British car parts suppliers, was marked 15100.  A search of bearing manufacturer's web sites confirmed that the two bearings are not just variants of each other--they have one important dimension that is different.  My supplier tried to be helpful, and did acknowledge the difference, but still insisted that the bearing they supplied was the correct one.  It baffles me that after 40 years supplying parts for these cars, this is apparently the first time they have heard of this problem.  In hindsight, I believe that the supplied bearing would probably work in the later differentials where the solid spacer is replaced by a crush tube.  Also, if a differential had a preload shim pack larger than about 0.036", the supplied bearing might work if most or all of the shims were removed.

It was at this point that I was preparing to shorten the spacer to make up for the extra bearing thickness, but in the end, I was able to get a correct bearing from a kind member of one of the popular TR6 forums.  





Preloading bearings not only removes any free play, but also causes the bearings to stiffen their rotation.  The preload for the pinion is correct when it takes 15 to 18 inch-pounds of torque to rotate the shaft.  Torque can be easily determined by measuring the force required to turn the shaft when applied  at the end of an arm of known length.  The product of the force and the arm length is the torque.  It took me three iterations of adjusting the preload shim pack to get the preload within spec.




The final check on the pinion is its fore/aft position.  There is a fancy factory tool for this which probably can't be found, but a little tinkering produced a way to accurately determine pinion position.  The block and bar in the pic are precision machinist's tools, but aren't really expensive.  The bar is 0.500" high, and the block is 2.000" inches high.  Measuring with a caliper from the top of the bar to the top of the block, and doing a little arithmetic told me that the pinion position was correct.  This method is based loosely on one described in one of the very helpful "Six Tech Manuals", available here: http://www.74tr6.com/pdfs/Differential.pdf 




With the pinion finally in good shape, I installed the new oil seal, added the powder coated mounting plate, and did the final torque of the pinion nut.




Next up was the differential carrier.  It carries the crown gear which meshes with the pinion, and also two sun and two planet gears inside, which allow the two rear wheels to turn at different speeds.  




The goal for the sun/planet gear arrangement is zero backlash among all the gears.  Backlash is adjusted by varying the thickness of the spherical thrust washers behind the planet gears.  I could easily detect backlash in my gears.  It only measured about 0.002", but was easy to feel.




I first tried installing a pair of new 0.060" washers--the same as I found in the unit--on the theory that most of the wear would have been in the washers.  This didn't seem to change the backlash very much, so I then tried a pair of 0.062" washers.  This resulted in no detectable backlash.  The gears were slightly stiff, but with good lubrication and a few miles, I expect they will loosen up.




So here is the differential carrier all back together (except for the crown wheel, which has to come later).




Two more critical specs in the differential are the preload of the differential carrier bearings, and the side-to-side position of the crown wheel.  There is a shim pack under each differential carrier bearing cone.  The total thickness of these shims determines the preload, while the distribution of them (some on the right, the rest on the left) changes the crown wheel position.  The process for determining the size and distribution of the shims seems complex at first, but it is entirely straightforward and logical.

First, the differential carrier bearing cones are mounted without any shims.




The carrier is then set into the case, and the end float measured.  This measurement is how much wider the case is than the carrier with bearings.  To this measurement, we add 0.003" for preload.  The total is the thickness of shims that will increase the width of the carrier (including bearings) to 0.003" wider than the case.

I actually found it a little easier and more repeatable to measure the float with loose feeler guages behind one of the carrier outer races (second pic).  In my case, the end float was 0.045", so the total preload shim pack would need to be 0.048".




Then it was time to mount the crown wheel back in its previously marked position on the carrier.  The crown wheel fasteners are grade "T" (close to SAE Grade 8), so it's not a place to use hardware store screws.  I used new grade 8 lock washers, and Locktite, per the manual.




The carrier with crown wheel is then placed back in the case without preload shims, and the end float measured again.  This time, since the pinion gear will now limit the movement of the carrier in one direction, the measurement represents the difference between zero and maximum clearance between the crown wheel and the pinion gear.  Subtracting 0.005" from this measurement gives the shim thickness  that would give a crown wheel-pinion clearance of 0.005".  This is the shim pack that goes inder the bearing cone on the crown wheel side.  The remainder of the total preload shim thickness then goes on the other side.  In my case, of the 0.048" preload shim pack thickness, 0.030" went on the crown wheel side, and 0.018" went on the other side.

The bearing cones had to come back off to install the shims.  One of the carrier bearings got fairly loose on its journal, so I helped it out by adding some bearing mount compound.




Then the case spreader was mounted again to open the case enough to alow the carrier to drop in.  The bearing caps were installed and torqued, and the final backlash measured.  I saw with a sigh of relief that my backlash was 0.004" and compliant with the spec.


 

The last major part of the differential unit was the inner axles.  These axles deliver power to the external flanges that the rear half axles attach to.  The inner axles consist of short splined shafts that each carry a ball bearing and an external flange.  The flange is mounted on a taper on the outer end of the shaft, and it has to be removed to withdraw the bearing and oil seal.  This is where the trouble started.



Now, I'm not a novice at dislodging things that are pressed on, or wedged on, such as with a taper.  I have a decent hydraulic press and the wherewithall to create a wide range of implements of seperation.  I tried the press first, and got it to the point where it was creaking ominously.  I then added heat to the mix.  The axle mocked me.

Not to be bested by an intert lump of steel, I made a custom puller whose grade 8 hardware would be able to apply as much pressure as my press, but with the added pursuading power of an impact driver.



After several intense cycles of pressure, heat, impact, and swearing, the axle remained aloof.  I was pushing the system to the point where I was concerned about damaging the axle flange.





As I paused to catch my breath and assess the situation, my more analytical self began to consider whether there may be a smarter, less brutish way to solve the problem.   The bearing on both axles was mounted up against the hub, but the machined 1-1/8" bearing journal extended about 1/8" beyond the bearing to a shallow shoulder.  The shoulder was only about 0.006" high and 1/8" wide, and was the only thing standing between that bearing and sweet liberation.



Now I don't have much doubt that, given enough acetylene and violence, I would have eventually triumphed over those hubs, but on the other hand, the prospect of pulling the bearings off from the other direction began to look much more civilized.  It was just that little wisp of metal  to deal with, and removing it would be trivial.  I considered the shoulder and whether it was there intentionally for some purpose.  It certainly wasn't a locating shoulder, since the bearing was an eighth of an inch away from it.  Possibly it was a retaining shoulder, to prevent the axle from creeping out of the bearing.  Neither bearing showed any evidence of creeping, and further, it seems that if that were its purpose, it would have been closer to the bearing and a little more prominent than 0.006".

In the end, I decided to do away with the shoulder and think later about replacing its function, if any.




Ten minutes to turn off the shoulder, ten minutes to set up the press, and the bearing was off.  I now consider the axle and hub to be one piece.




Then a pleasant degrease and derust, and a nice powder coat.



Now, about the late shoulder.  The only earthly purpose I can imagine for it is that it would keep the shaft from pulling out of the bearing.  Even though it appeared unlikely that this would happen, it seemed prudent to restore that function, even if it wasn't intended in the original design.  I decided to fashion a collar to go next to the bearing to stand in for the missing shoulder.

I started out with a hunk of really good steel...




...and turned a couple of rings with the same ID as the bearings.




Then drilled and tapped three holes in each one for set screws.  The set screws are the kind that have a little nose on the end that goes into matching holes in the shaft.




Pressed the seals into their carriers.




Here are all the parts for the inner axles.




Pushed on the bearings with that little pipe thingy, and installed the collars with a little Locktite on the setscrews.  




Shoved in the axles and tightened down the seal flange.  The flanges don't have gaskets, but some sealer seemed like a good idea, so I broke out my trusty Permatex 3D.  I considered making some gaskets, but the flange also captures the bearing's outer race, which is otherwise a fairly loose fit in the case.  A gasket might give the race enough room to shift a little, or even turn.  Also installed stainless fill and drain plugs.





Rounded up all the replated fasteners for the rear cover, using new lockwashers.  A couple of dowels locate the cover.




Buttoned up the cover, and added new mounts to it.  This puppy goes on the shelf to make some bench space for something else.




So was the creativity with the axle shafts a good idea?  I don't know yet, but for the time being, it keeps me from admitting that the axle shafts beat me.


Send comments to:  elhollin1@yahoo.com


To Other TR6 Pages