MG-Cars.net

Welcome to our resource for MG Car Information.

Recommendations

Parts

MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG Midget and Sprite Technical - Rear Axle knowledge?

In working on my axle over the last few weeks, I've researched the archives here (TONS of great information), studied the Bentley and Haynes manuals, and read postings on several other midget websites (and MGB, for contrast), and ran across a great number of bits of data that I wished was all gathered together in one place for easy reference. Mostly for myself, but then I thought, maybe someone else could benefit?

So, here is a draft. I plan to add photos where "==>" indicates.

Do please let me know if you see any inaccuracy, or missing information!

Norm "learning the knowledge" Kerr


Rear Axle

Oil leak:
- plugged breather (plastic "mushroom" located at the top of the axle housing, RH of the differential). Pull off the top "cap" and then unscrew the "stem" to clean.
- bad wheel bearing (causing seal to not touch properly)
- worn out seal (or dried by age, especially on a car not run for years)
- worn housing (if you can feel a groove, have a machine shop install speedy sleeves to restore the sealing surface)
==> photos before/after
- torn or re-used gasket (trim the inner diameter of the new gasket so it doesn't touch the O ring. Max gasket thickness 0.010", any thicker will result in a loose wheel bearing)

Make sure the leak is actually from the axle seal. Rear brake cylinders develop leaks (look for fluid around the dust seals), and are cheap to replace ($20). If the liquid is on the cylinder and under its dust seals, it is leaking. If it is around the hub (not easy for the brake fluid to get there), then the axle seal is leaking. If there is liquid all over everything, then both are probably leaking.

When replacing the bearing, use a 207-FF (double sealed bearing, which then acts as a supplemental seal against axle oil leak). You can purchase from any foreign auto parts store (also used in Subarus).

Hub seal dimensions OD 2.5, ID 1.75, seal lap 0.375".

If you have a persistent leak between the axle and its mating surface on the hub, your axle may be bent, or you've swapped them R x L (RH is longer and will jam against the diff pin, preventing seal). The correct O ring is 0.115", though a larger dimension could be tried if struggling with leaks.
==> is one axle really longer? "Left side has a green dot" RH is 1/32" longer on wire wheeled car.

Hub nut torque (not shown in the manuals): 40 ~ 45 ft-lb. No need for much torque, as this only needs to hold the bearing race securely against the axle housing, and rotation is prevented by the lock tab.
LH hub nut has reverse thread (same reason as the wire wheel nut, but opposite side of the car, because of the rotation direction effect)


Axle break:
- "spin and snatch" a rear wheel on the shoulder under power can snap an axle
- bad wheel bearing (excess play = overloading the axle while cornering)
- bent axle housing (have a machine shop confirm true)
- axle spline worn by the differential gear's mating spline (groove width wear on the axle of more than +0.010" and the axle should be replaced. Also, the presence of this wear creates a stress riser at the most highly loaded portion of the axle/spline)
==> photo before/after
Growlerize to minimize both the original stress riser where the spline cut ends, and also from the wear caused by the edge of the differential gear mentioned above.
- weak material (pre '68 axles only EN7, from '68 BMC adopted EN17). Identify later axles by a groove machined at the end of the differential splines.
==> photo before / after
==> is there an "EN17" marked on the outside, domed face?
- axles swapped R x L (reversing the loading reverses the twist in the material and can lead to failure). Also, they should be different lengths (RH is 1/32" longer?).
Double bearings have been used in an attempt to increase axle life, especially for racing, but some have had sealing issues with them, and some racers say they still use stock bearings, without ever breaking an axle. Growler's treatment should be effective to minimize stress in the area where most axles fail (near the diff).
- Peter May makes EN24 shafts
- Frontline makes EN40B nitrided shafts. Peter May says EN40B is too hard/brittle, "EN24 provides the best balance of strength and toughness".
- both are about double the cost of new, stock axles.

Growlerize the shafts (named for the screen name of the machinist on the MG Enthusiast BBS in the UK who thought of it): cut down the diameter of the shaft to the root diameter of the splines. Start the cut 1" from the end of the shaft (so that the very ends of the differential splines are over air, not against the face of the axle splines), and provide a generous radius at the end of the splines. Continue the root diameter for 4", then gradually taper over 6" back to the original diameter. De-burr and then shot peen (like is done with a connecting rod). Why does this work? The axle itself is plenty strong enough to handle the engine torque, so reducing its diameter near the diff, although it gives up torsional stiffness, is offset by the benefit gained for strength by avoiding cracks at those eliminated stress risers, and removing the sharp corners at the end of the spline cuts, and eliminating the stress riser where the diff splines end, eliminates two key weaknesses in its design that lead to cracks/axle failure. Ideally, the reduced torsional stiffness caused by this diameter reduction will allow the shaft to twist a bit under severe load (like a torsion bar in a Minor's or Jaguar's front suspension) instead of snap.


Axle noise:
- loud whine/howl at speed = ring and pinion are shot (gear tooth surface worn/damaged). Replace diff with a used one (typically, they last 150k miles with no maintenance, as long as they are never run dry, or been exposed to broken axle fragments).
Clunk noise:
- worn U-Joint (confirm no play)
- loose brake back plate (confirm bolts tight and holes not oval)
- axle worn at diff (max 0.010" spline wear). The axles seem to be softer than diff gear, so they wear first.
- wire wheel spline wear (new splines are squared off on the top, worn splines are pointed). Clean off the grease from both the hub and the rim and then check the play without the nut. Maximum 1/8" play at tire. Replace wheels and hubs at the same time (one new one will quickly wear the other). They are NOT cheap (new hubs and wheels), so don't only do it half way. Think of how nice wire wheels look, while you are writing the check. Original hub spline outside diameter 2.450". Note, though, that a fair amount of spline wear can occur before this diameter decreases by much, hence the tire rotation test above, which directly measures the wear as a rotational free play.
==> photo before/after
- worn splines encourage extremely tight wheel nut (pounding with a hammer to try and hold the wheel still by friction), which leads to spline bottoming out on the inner taper of the hub. So, spline marks on that taper are another indicator of worn splines.
==> photo


Axle shaft history ("or, so I've heard.." ):
- Frogeye = 2A7085 (? material)
- 1098 era = BTA 501 (EN7 material)
- 1968 ~ = BTA 806 (disc wheel) / -807 (wire wheel), (EN17 material, with Identifying groove on the inboard end)


Differential:
- pinion play should be 0.0 fore-aft, max 1/4" rotational play (if 1/2" or more, diff is toast)
- '78 ~ '80 Midgets came with 3.77 ratio (better for highway use than the 3.9 used from '61 ~ '77, but rare, so more expensive). "58 ~ '61 (997cc) used 4.22. The ratio is stamped on the outer edge of the diff housing.
==> confirm first year of 3.9
==> photo of diff ratio stamp
- to replace pinion seal, mark the nut location/flat position and re-install it with the exact same amount of rotation. Not doing this will mess up the pre-load and require replacement of the crush spacer.
- planet and sun gears have thrust washers that wear away over time, and can be replaced, but the diff must be completely disassembled to access them.
- trying to judge pinion free play on the car can be misleading because holding the tire and rotating the pinion will show you several free-plays combined: wire wheel play, axle shaft spline wear/play and pinion/crown wheel play. Measuring the play of a diff on the bench is the accurate way, with a dial indicator. The spec is engraved on the back of the gear (each matched set will have a matching set number, and the crown wheel play for that set, and the amount of shim offset to the right or left, all engraved on the back of the crown wheel.
==> photo
- after an axle break, or any time the rear end has been apart, it is wise to assure that the insides are surgically clean for maximum diff life: refill the axle with kerosene, with the car up on jack stands, and run the engine in gear for a minute (i.e. no load) and then drain through a strainer. Repeat until the kerosene drains clean (may take several cycles). Fill with 80/90 gear oil (2.1 US pints / 0.99l).
- the drain/fill plugs have a tapered square hole that can be fitted with the square end of a ratchet wrench, but the taper of the hole is problematic. Hardware stores sell pipe nipples with a normal, hex drive shaped hole to ease future servicing, or you can buy a ratchet extension and grind it down to match the taper of the original plug for maximum fit.


Norm Kerr

Norm

Growler is a Kiwi - he did indeed come up with the concept of reducing half shaft diameter to take out the weak spot. (I seem to remember his real name is Grant Bowyer) I presume he finally got rid of his Sprite as I haven't seen any posts from him for years...

Regards

JB
James Bilsland

I've always wondered what happened to him.
Brad (Sprite IV 1380)

Great work Norm, I'll print it out and clip it in my workshop manual.

For Brits, most of Norm's references to Axle read Half-shaft. Can anyone confirm the 1/32" longer point, it's new to me. Any other little disputes or corrections?

Great to see the photos if possible.

Gary L - a good tech article, if Norm's ok with it!

Andy (just going to check my backplate bolts) Cross
Andy Pie-Crust

Norm when completed with pics how about loading it to the Spridget wiki please?
http://taratec.org/mediawiki/index.php?title=Main_Page
David Smith

Norm

The 3.9 diff was first introduced in December 1968.
Dave O'Neill 2

kewl, thanks guys! I was wondering how to put this on the web as an article. I will look into the wiki and do so!

I'll make those corrections.

Another update I have to make: "If using a sealed bearing, be sure to lube the (original) axle seal real well before installing the hub to the axle, because it won't be getting any of the oil it normally would have seen."



Norm
Norm Kerr

Norm

When replacing the sun & planet gear thrust washers, its true that you do need to strip the diff, but not completely, and the good point you dont disturb any of the important shim settings in the pinion or the side bearings on the carrier, so its a reasonably straight forward job.
Just doing one at the moment.
Ian
Ian Webb '73 GAN5

Norm,

Im not sure who, but there are several guys converting the arcives into a wiki forum/page.

You need to get in touch with them and post your data for the rear axle assembly...NICE JOB

Prop
Prop

You might want to add the wire wheel housing is 1 inch narrower than the disc wheel housing. I have seen many references to switching the axles from disc to wire with out changing the housing. This puts the wire wheel .5 inch closer to the outer edge of the square body style body panel. This isn't really a problem if you use skinny tires or have a round wheel arch model car, but wider tires can rub the body. There was also another axle available, BTA 940 disc and BTA 943 wire listed in the special tuning section. Can anyone verify these axles were used on later 1500 models??? There is no mention of two part numbers for left and right sides, for the 1/32 inch difference. This info is from the official parts catalogue # AKM0036 1st edition.
The 3.7 ratio started in August,1977 GAN6-200001. Horler pg 132.
J Bubela

First I'd like to say: Fantastic!
second I noticed this;" - pinion play should be 0.0 fore-aft,"
0.0 means there's no play...
Alex G Matla

Never realised that Growler was a Kiwi, or which Kiwi for that matter. I've been doing that half shaft mod here for 30+ years, good on him for posting it when he did.
Paul Walbran

ooh!

I was hoping to find someone with experience with that shaft mod to ask some follow-up questions:

Paul, can you tell me, have you ever seen a shaft break after it has been modified like that? (for instance, they sometimes break farther outboard than the splines, presumably by over-torque, and I wonder how well, in the real world, the mod helps the shaft twist instead of snap).

Also, have you ever felt any kind of "whippiness" with the cut shafts? Removing diameter will naturally reduce a shaft's torsional stiffness (is it the square of the radius? gotta look that up), so I was curious if that difference was at all possible to feel/sense while driving, or if it was far too subtle for that.


thanks,
Norm
Norm Kerr

David gave Norm the link for the Wiki Prop

Keep up!

http://taratec.org/mediawiki/index.php?title=Main_Page

Great work too Norm

n
bill sdgpm

Yes, I have seen the odd half shaft like this fail - however, only after a lot more punishment than it would have otherwise taken. (Decades of autotesting is a good proving ground!) If they fail, it has been at the end of the spline - which is fortunate because if it was further out extracting the diff head would not be so easy - which is a reflection that the chunky square spline form has an inherent stress raiser which concentrate the stress into the centre of the halfshaft. Finer v-shaped splines spread it around a greater radius within the halfshaft, so the stress concentration is much reduced.

We have used EN30B and 4340 in the past made to this pattern, had had both types fail ... eventually, I must stress. They had done an awful lot of work - std halfshafts (unmodified) would be lucky to last a season, these went for 10 years.

When we built the K Midget, 185BHP from a VVC so torque from the word go, we investigated further. Same design principle of course, but we identified 2767 as theoretically a better material, and went for fine splines. When we bought the LSD fron Frontline, I asked Tim if he did a fine spline option - he didn't at that stage. We had a chat about the reasons why, and Tim followed up by developing his fine spline option.

The K has been going 8 years now and we have had no halfshafts issues despite using formula R's + lots of clutch dump while autotesting, not to mention the thump after getting briefly airborne when under power in unsealed events. Through all this there hasn't been any noticable effect on the feel of the car from the halfshaft winding up - though it obviously is as there is a bit more wear at outboard end of the splines than the inside end. Not much, but enough to leave a tell-tale print.

Final thought: if you are going to modify the halfshafts, get them crack tested first to make sure you are not wasting your time, and getting them stress relieved is worth considering.




Paul Walbran

Paul, great feedback, thank you!

One follow-up question so that I add your advice correctly to the wiki/article: should stress relieving be done before, or after machining?


Norm
Norm Kerr

Stress relieving is usually done prior to machining - sometimes (particularly in crankshafts) a small amount of distortion occurs in response to the stress relief, which can then be rectified if the machining process is the next step.
Paul Walbran

Norm I have just seen the entry in the WIKI

great stuff thanks

can be checked out here:

http://taratec.org/mediawiki/index.php?title=Axle
bill sdgpm

This thread was discussed between 14/02/2010 and 20/02/2010

MG Midget and Sprite Technical index

This thread is from the archive. The Live MG Midget and Sprite Technical BBS is active now.