Differential Failure

[rgh0]

Diff back in the car and ran it for 20 miles. The old diff had been hard to remove from the car but the new casing must be fractionally larger than the original as there was no way it was going back in until if jacked the bottom chassis down about 1 mm.

No noise at all and no leaks so it looks like I have assembled it correctly . There have been comments about whine from the Burton replacement crown wheels and pinions but I certainly don't have any.

Lotus 2015 national meeting and track day next weekend, this should run it in completely. I will change the oil after that and it will be ready for my next race meeting early November.

cheers
Rohan

[promotor]

Nice work Rohan, nice approach and you must be very gratified that it all works. Well done!

[rgh0]

Nice work Rohan, nice approach and you must be very gratified that it all works. Well done!

The most interesting bit was that there is no way you can get the right pinion bearing pre-load with the Burton crush spacer just by torquing the pinion drive flange nut. I had 150+ fl-lb on the nut and it would not "crush". Putting the assembly in a press got the needed crush to pre-load the pinon bearings but it took a lot of careful increments of press load to get it right. I wonder if this is the source of the diff whine people have as they have not pre-loaded the pinion bearing right?

cheers
Rohan

[elanner]

Hmm. Mine is still whining slightly. Only at around 30 mph on a light throttle. I have about 600 miles on it. I'm desperately trying to convince myself that it's slowly getting better. It certainly has improved, but whether it will completely go away remains to be heard.

I didn't rebuild it myself because I don't have the tools or skills. So I have to hope that the crush washer was set correctly (it too was from Burton). I hope to God that I don't have to remove the diff again, for the 4th time.

The only ray of sunshine is that I really like the 3.55 ratio. I think it's a great improvement for modern roads.

Nick

[prezoom]

Got tired of using crush sleeves on my H series racing differentials. Wound up taking an old one, torqued it up on the new pinion and measured depth. Then took one from a R series diff, much heavier and essentially a solid sleeve, put it in a lathe and trimmed it to the correct length to obtain rotational torque. On occasion we used one of the R series, pre trimmed, and used shims to get the correct measurement. Never had a failure and all diffs remained quiet. IMHO crush sleeves are just that, an easy way to assemble. A solid sleeve will never move once assembled.

Will try a Burton sleeve when I do the 3.55 treatment, but if I have problems I will make another solid sleeve and use shims/trimming to get what is needed.

[promotor]

Nice work Rohan, nice approach and you must be very gratified that it all works. Well done!

The most interesting bit was that there is no way you can get the right pinion bearing pre-load with the Burton crush spacer just by torquing the pinion drive flange nut. I had 150+ fl-lb on the nut and it would not "crush". Putting the assembly in a press got the needed crush to pre-load the pinon bearings but it took a lot of careful increments of press load to get it right. I wonder if this is the source of the diff whine people have as they have not pre-loaded the pinion bearing right?

cheers
Rohan

I overhauled a diff recently that had been supposedly overhauled just before it was sent to me and on inspection it hadn't had the crush washer crushed (among almost every other possible thing wrong with the whole overhaul!!) - the pinion nut was tight and I guess the builder presumed this meant the bearings were loaded (which they weren't) - and therefore the bearings weren't seated (side play resulted) and therefore weren't running with any pre-load.
They do take a lot of crushing when I do them and it feels like either the long bar I use is going to break, the pinion thread is going to strip or the shaft break off, or the cartilage in my elbow is going to pop! Some crush washers are harder than others to crush depending on supplier. Despite this I certainly prefer them to the original Ford ones that were made of a much soft material - there was hardly any point in them being fitted!

How did you end up checking the final bearing drag/running torque of the pinion bearings, and also how did you measure cap spread of the side bearing housings? (which is the method of obtaining the side bearing running torque since there is no other accurate way of doing it)

[Chancer]

When I did them years ago I would "redress" the crush tube by gently hitting the bulge radially with a hammer while it was on the pinions shaft to increase its length, its true that they didnt put up much resistance.

In my experience even on the rare occasions that people achieved the correct tooth engagement incorrect preload of the pinion or crownwheel bearings usually was the cause of premature failure.

I once had a Suzuki Jeep from new, it developed an intermittent groaning screeching sound from the rear diff which would grind the vehicle to a halt, people would be wincing from 100 yards away it sounded so terminal yet after backing up a few feet I could drive away again in silence waiting for the next occurrence.

I pulled the diff and there was no visible wear, no metallic pi?ces or even paste, the bearings were all in good condition but had no preload, I re-assembled it and preloaded the bearings and it was silent for the next 40K miles until I sold it.

I used to set the pre-load by feel, the works Ford manual showed you how to measure the pinion one with a pully, rope and a spring balance, as you say a spreader jig and a DTI are needed to set the crownwheel bearings, I first got my "feel" by turning new Ford differentials which were abundant back then and it became more refined the more diffs that I rebuilt, of course that feel is completely lost now, nowadays I snap rusty bolts that I would never have done in the past.

How do you set the pre-loads Promotor?

[rgh0]

How did you end up checking the final bearing drag/running torque of the pinion bearings, and also how did you measure cap spread of the side bearing housings? (which is the method of obtaining the side bearing running torque since there is no other accurate way of doing it)

I slowly pressed up the bearings until all the end float was taken up and the crush tube close to its needed length. I could then tighten it the final amount using the pinion flange nut. I hung a 5 litre oil bottle on a string on the back of the pinon flange which gives the correct torque and tightened the nut until it would just not turn the pinion.

I measured the spread using both the torque specified for the adjusters and also using a large micrometer which could span the bearing supports so I could measure the spread. The two different types of measurements lined up pretty well for me.

cheers
Rohan

[promotor]

I used to set the pre-load by feel, the works Ford manual showed you how to measure the pinion one with a pully, rope and a spring balance, as you say a spreader jig and a DTI are needed to set the crownwheel bearings, I first got my "feel" by turning new Ford differentials which were abundant back then and it became more refined the more diffs that I rebuilt, of course that feel is completely lost now, nowadays I snap rusty bolts that I would never have done in the past.

How do you set the pre-loads Promotor?

Feel is the way I try to do it but how I feel on the day doesn't always equate to it being right even if it feels that way! Some bearings have that initial stiction that diffs used to have but generally modern bearings are smooth on takeup even if running at the high end of the running torque. That affects my feel too.

I've got the proper original workshop equipment cap spread gauge and pinion pre load gauge (plus dummy pinion and gauge etc) and prefer to work with them as I know it's correct when I've used them.

I think you are right about getting the bearings loaded - loose pinion bearings = rumble. Some of the diffs I've seen have the side bearings as loose as wheel bearings which is way off - I have to use a fairly long pin-ended tool to turn the side adjusters.

How may diffs do you estimate you've built?

I measured the spread using both the torque specified for the adjusters and also using a large micrometer which could span the bearing supports so I could measure the spread. The two different types of measurements lined up pretty well for me.

cheers
Rohan

Good idea on the micrometer method!
Where is this specified torque for the adjusters found? I have numerous genuine workshop manuals (Lotus Cortina Mk1, Cortina mk2, Mk1/Mk2 escort etc, plus a genuine Lotus Elan manual) and have never seen that anywhere! I don't think I'll use it though as the cap spread gauge is a really easy method.

[rgh0]

Where is this specified torque for the adjusters found? I have numerous genuine workshop manuals (Lotus Cortina Mk1, Cortina mk2, Mk1/Mk2 escort etc, plus a genuine Lotus Elan manual) and have never seen that anywhere! I don't think I'll use it though as the cap spread gauge is a really easy method.

I referenced the "differential bearing adjusting nuts" torque of 12 -15 ft-lbs that is specified in the Final drive torque settings section on page 26 of the technical data section of my Elan manual. I ended up using about 16 - 18 ft/lbs to get the correct 5 to 7 thou cap spread I measured with my micrometer. Getting torque and load to match that closely is pretty good or pretty lucky as torque versus load can vary significantly with surface finish and lubrication of the nuts and the surface they bear on. This is why using stretch gauges on critical fasteners such as con rod bolts or measuring spread in this case is much better than just torquing them up.

Many people including trained mechanics and fitters ( millwrights in USA terms) don't understand the purpose of pre-loading back to back taper roller bearings and correctly loading fasteners and tend to skip over those details (quote - "I cant be bothered reading the manual I know how to bolt these things together, been doing it without problems for year"). The easiest way I judge a maintenance workshop when I go to manage a new process plant is to ask to see their measuring equipment and torque wrenches and the specifications for fits and clearances and adjustments and bolt torques on the machinery they rebuild. If I get blank looks and their equipment amounts to an air wrench and a steel rule I know I have a challenge to improve their rebuild standards

cheers
Rohan

[promotor]

Where is this specified torque for the adjusters found? I have numerous genuine workshop manuals (Lotus Cortina Mk1, Cortina mk2, Mk1/Mk2 escort etc, plus a genuine Lotus Elan manual) and have never seen that anywhere! I don't think I'll use it though as the cap spread gauge is a really easy method.

I referenced the "differential bearing adjusting nuts" torque of 12 -15 ft-lbs that is specified in the Final drive torque settings section on page 26 of the technical data section of my Elan manual. I ended up using about 16 - 18 ft/lbs to get the correct 5 to 7 thou cap spread I measured with my micrometer. Getting torque and load to match that closely is pretty good or pretty lucky as torque versus load can vary significantly with surface finish and lubrication of the nuts and the surface they bear on. This is why using stretch gauges on critical fasteners such as con rod bolts or measuring spread in this case is much better than just torquing them up.

Many people including trained mechanics and fitters ( millwrights in USA terms) don't understand the purpose of pre-loading back to back taper roller bearings and correctly loading fasteners and tend to skip over those details (quote - "I cant be bothered reading the manual I know how to bolt these things together, been doing it without problems for year"). The easiest way I judge a maintenance workshop when I go to manage a new process plant is to ask to see their measuring equipment and torque wrenches and the specifications for fits and clearances and adjustments and bolt torques on the machinery they rebuild. If I get blank looks and their equipment amounts to an air wrench and a steel rule I know I have a challenge to improve their rebuild standards

cheers
Rohan

It looks like you found a lucky coincidence regarding a torque figure for the side adjusters - in the Ford manuals the differential bearing locking plate bolts have the same torque range of 12-15lbs ft.
There is no figure for a torque on the side adjuster nuts and there is no mention in the Ford manual on using a torque wrench for this step - that is what the cap-spread gauge is for. However, it does take a fair amount of torque with a long pin wrench to spread the caps.
I asked my old man (a Ford Master Technician) about this and he's never seen or heard of it and he was known as "The Font (of all knowledge)" wherever he worked.
Similarly there is no figure in the Lotus manual for bearing locking plate bolts - I believe it's safe to assume that Lotus have confused their description as the step by step process of overhauling the diff doesn't make mention of a torque for the side adjusters either!

It's not to say I won't be seeing what figure is obtained with a torque wrench to see where it's at but it's not the proscribed method!

Fortunately, there is usually more than one way to skin a cat!

[rgh0]

Yes Promotor I was somewhat doubtful of the quoted torque for the same reasons you mention as its a little unusual to quote a torque for a pin drive adjuster used for pre-loading bearings. Lotus has been known to make more than a few mistakes in their workshop manuals

But the description for the torque setting for the bearing adjusters is the same term as used for the bearing adjusters in the assembly description and it did align reasonably with the spread i measured. Maybe it is something Lotus discovered and added at some stage ?? or like you say maybe its a mistake and just a coincidence the torque and spread aligned reasonably at least for my assembly.

Whatever --- the diff has now done around 100 miles includng the last 70 at racing speeds and full load and no noises or leaks -- time for a oil change to redline in time for my next race in a few weeks.

The third gear upchange in my Elan gearbox is getting a little slow so looks like time for a gearbox rebuild next as my christmas project.

cheers
Rohan

[l10tus]

Rohan,

Thanks for sharing that insight with us, great article, complete from start to finish.

Excellent photo record to boot!

Regards,

Phil.

[seniorchristo]

I am building a 3.55 differential from scratch. I'm starting with an alloy housing and 3.55 ring and pinion from Burtons. Their housing is marked with the required pinion shim. In this case .1405". There was no correction marked on the pinion, so I installed the shim and bearing and installed it in the housing. I measured pinion depth and came up with 3.243" which is .047" short of the standard (3.290"). I have a good used 3.77 setup to compare with and I got very similar measurements. The head pinion bearing was only resting in its race, I have not installed the tail bearing or set the preload.

My question is: Will installing both bearings and setting the preload move the pinion away from the crown wheel getting me closer to the desired measurement? Since the bearings were supplied with the races already installed, I assume the supplier was able to check the required pinion height and that the shim dimension is correct.

Thanks
Chris

[2cams70]

The key thing to remember is that it's all in the final tooth contact pattern you get at the end. If that's correct you know you have the pinion in correct relationship with the ring gear. It doesn't matter so much how you get there whether it's by measurement, trial and error or a mix of both.

If that's correct, the pinion bearing preload is correct and the cap spread is correct so as to correctly preload the ring gear carrier bearings all is good.

Tip: Buy yourself another set of pinion bearings of the same make you already have. Use a brake cylinder hone to grind out the bores of the bearings so that they become a slip fit rather than a press fit on the pinion. This then becomes your set of bearings for trial assembly purposes only (much easier to remove and replace). Use these to do a few trial builds to see where things sit in terms of tooth contact pattern and adjust shims according to what you observe. Yes it's true that things may change slightly on final build with correct preloads, cap spread, and press fit rather than slip fit pinion bearings but it likely won't be significant. You'll of course be doing a final check of tooth contact pattern in any case.

As a matter of interest do Burton's quote figures to use for cap spread and pinion preload? I'm always a bit skeptical of these aluminium replacement components. Yes aluminium is lighter but it does have significantly different mechanical properties compared to cast iron. You can't just substitute one material for the other unless it's a ground up redesign.