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 MGA - Banjo Axle Hub Bearing Nut Torque

I am rebuilding my banjo rear axle for my Coupe and cannot seem to find a torque spec for the nut that holds the bearing carrier in place. Also, what are thoughts on whether or not to pack the bearing with grease prior to intallation? Factory manual says nothing, I assume since it is lubricated by the gear oil, but the Haynes says to grease before installation.
Thanks,
Scott
Scott Shirk

The rear axle bearings are lubricated by the gear oil in the differential, adding wheel bearing grease seems redundant at best. Having a suitable tool to turn the 8 sided nut seems to be half the battle. With a socket that wont slip tighten to 150-200 ft. lbs.
Here is Barney's discussion. http://www.mgaguru.com/mgtech/rearaxle/ra102.htm
R J Brown

Hi Guys, so what is the correct size socket for these nuts and is it similar to any other current make of car (so I can borrow from the workshop)?
regards
mark
Mark Mathiesen

It is 1-61/64 inches across flats. For socket sources and possible substitutes, see here:
http://www.mgaguru.com/mgtech/tools/hub_socket.htm

Give it all the torque you feel like applying, but not less than 140 lb-ft. It is hardly possible to over tighten it. The large thread could tolerate up to 900 lb-ft torque before stripping the threads, even with the thin nut.
Barney Gaylord

There is a socket for a current make of car that works great, but unfortunately it is for an American Ford 4x4. Here you can get one at just about any car parts store, but probably be tough to find down in Australia. Barney has a good page about all the options. The main thing is that the nut is octagonal.

Some modern dirt bikes(CR250?) use 50mm octagonal caps on their forks, and 50mm is extremely close to 1-61/64. Perhaps some motorcycle shop has a socket available at a reasonable price.



Jeff Schultz

Hi Mark,

If you are near Brisbane I have a hub spanner you can use.

Russell
R J Goebel

I don't see the sense in applying such high torque to this nut. All it does is stop the bearing carrier falling off. It just needs to be "snug" it can't back off because of the tab washer. There is no point in stressing the axle threads.
Art Pearse

Art the bearing needs to be tight enough so it does not move in relation the axle housing. If the inner race starts to move it will eat the axle housing and ruin it. The nut needs at least 150 preferably 200 or more. This is not just to hold the bearing from falling off the axle. The bearing to axle housing fit is too loose to leave on its own. It must be held tight enough to keep the inner race from moving in relation to the axle housing.
R J Brown

It's a big nut, so a big torque is needed!

Torque is force times distance from the centre so double the radius needs twice the torque.

It's a bit more complicated than that, but that requires complex calculations involving thread angles to calculate the axial force.

Assuming that the thread angles are the same (which is true for most thread series), the axial force is proportional to the tangential force on the surface of the thread.

40 ft lb on a head stud (approx diameter 1/2") is a 1920 lb force tangentially on the thread.

120 ft lb on 1 61/64 is only a 1474 lb force tangentially on the thread.

Dan Smithers

RJ, you are right about the need to clamp the bearing race, but that should not require a lot of force, since the only force turning it is the bearing friction, should be quite small.
Dan, I don't follow your "tangential force on the thread" argument. (I think you calculated it as 40ft.lb x 12ft/in /0.25 in bolt radius. This is irrelevant. The axial force depends on the torqu, the thread pitch and thread friction. For the nut in question, Assume 12 tpi. 140 ft.lb is 140 lb applied at 12" radius. The mechanical advantage of this screw (transferring tangential motion into axial) is 12" x pi x 12 = 452. (The ratio of distance travelled by force to distance travelled at the work) So, excluding friction, the 140 ft lb gives a force of 140 x 452 = 63,300 lb on the bearing ! Halve this for friction, say 30,000 lb. And this is hanging on about 4 threads! And Barney, 900 ft lbs ??? 40 ft lb is enough!
Art Pearse

40 ft lbs may be enough if the bearing is in good shape.
We see failed wheel bearings that spin the inner race and eat the hub all the time. By the time a customer notices the noise of a bad wheel bearing in a Corolla it has already destroyed the hub over 50% of the time.
The turning force on a bearing that is dying will be enough to over come 40 Ft lbs of torque. Be prudent don't risk destroying the axle housing torque it to a MINIMUM of 150 ft lbs.
R J Brown

Art,
I have to agree with the others. The problem is that you are grossly underestimating how much friction contributes to torque. Even with a highly lubricated nut, 140 ft/lbs of torque will generate less than 9,000 lbs of clamping force. The charts in the following links show allowable torque of over 1000 ft/lb for a well lubricated 2 inch grade 2 bolt.


http://www.portlandbolt.com/technicalinformation/bolt-torque-chart.html

http://www.portlandbolt.com/faqs/tension-vs-torque-explained-sort-of
Jeff Schultz

As I said,
"Assuming that the thread angles are the same ..."
which is a result of the thread pitch and bolt diameter.

I don't know the pitch on the hub nut, but looking at UNC values and using your 12"*pi*tpi formula for 14ft.lb:

diameter pitch 12pi.tpi "force"
1/2" 13 tpi 490 68612
1" 8 tpi 301 42223
2" 4.5 tpi 169 23750

As we have already noted, the value is not a force, but is proportional to the force.



Dan Smithers

All the math is totally beside the point. The point is if you don't use a torque setting of over 150 you run the risk of damaging the axle housing.
Remember the engineers that proved on paper that a bumble bee can't fly? In the real world bearings spin and damage axle houings if left too loose.
R J Brown

When I restored my car in 1988 I had to punch all around the end of the axle casing to get the bearing tight, as it was basically shot. ( I couldn't locate another casing then). It was always a pain and always leaked oil. Two years ago I gave up on the annual change of the oil seal etc, and bought a used axle case from Bob West. I have had no leaks at all since, as the bearing and hub are now tight rather than wandering around a bit.

I would strongly recommend getting that bolt very tight indeed to avoid any issues with the axle casing. Although you can sleeve and repair, it's a pain to do and much easier to do the nut up good and tight in the first place.


dominic clancy

Has anybody used Loctite. I would think that it would help seal and afix the bearing without having to rely on the torque of the nut.

Comments?

Chuck
C Schaefer

The standard bolt tables should be applied with reference to the bolt pitch, not diameter, in calculating clamping force. "A finer pitch creates more clamping force" Anyone disagree with that statement ? The standard 2" nut has a pitch of 4.5 tpi and thus requires more torque to get the clamping force. It can also withstand more because it has a lot more threads than the nut we are talking about!
I just measured the nut - it has 16 tpi. The clamping force from the Portland bolt Co tables for a 4.5 tpi lists 67500 lb at 1125 ft lb. Factor this to 140 ft lb and 16 tpi, gives 29866 lb. For a 16 tpi bolt at 7 ft b, it gives 2093 lb. Factoring that to 140 ft lb gives 41860 lb. So, somwhere between 15 and 21 tons force!! Such a force is unnecessary. This is not a standard nut. The axle case is not high tensile steel.
The bearings are ball bearings and cannot be accidentally over tightened, like tapered rollers, so should never create enough drag to spin the inner race around the axle even if the clamping force is only say 2 tons.
A Pearse

I try to stay out of religious wars and pissing contests, but I cannot stand by silently when safety is at stake.

Perhaps Art didn't rear RA-102 yet. That touches on some of the dynamic forces involved in wiggling the inner bearing race (while saying nothing about spinning it). The inner bearing race can be induced to wiggle and walk around on the axle housing with zero torque if the nut is not tight enough. A few of the forces involved are acceleration, braking, vibration, and harsh suspension bumps. 150 lb-ft torque is about one sixth of the normal assembly torque for a nut that size and that thin. It is only about twice the normal torque applied for a wheel lug nut. It is really not much, considering the application.

Coefficient of friction between two smooth steel surfaces can be as low a 0.2, so a radial bump of 1000 pounds force could move the bearing race against a clamping force of many thousands of pounds. Would anyone like to venture a guess what radial loads the wheel bearing encounters with a rough road and other harsh driving conditions?

Maybe Art would also like to tell us that 15 lb-ft is enough torque for a wheel lug nut, because the wheel can't move when installed. Some people refuse to believe spec's in a workshop manual, no matter what it says. Perhaps we should let Art believe whatever he likes, and let him replace the leaky hub seals along with with every engine oil change.
Barney Gaylord

This thread was discussed between 13/01/2008 and 20/01/2008

MG MGA index

This thread is from the archive. The Live MG MGA BBS is active now.