MK3 rear axle (2017)

mdt-son said:

It seems we need to source a supplier in the UK for this.
L.A.B., do you know of someone who could undertake making a small batch?

Click to expand...

No, sorry, I don't.

Hi all,

I have had an email exchange with Andover Norton on the subject. Here is a section of my letter with them:

"As for the failure rate of the Mk3 rear axle, this is clearly a high cycle fatigue failure due to inherent design stress risers. Failure will depend on loading, road conditions, mileage, riding style, and possibly manufacturing and material defects. Several of the reported breaks have been on race bikes. The lean around a course will increase axle loading towards one side (mostly the left side), thus increasing bending and subsequent crack growth.

In order to dismiss the fatigue failure problem as irelevant, one really needs to examine the service record and rider´s notes, especially of those bikes which have covered substantial mileage.

At first glance the pre-Mk3 models should be even more affected due to their thinner axle - maybe they are not used as much as the Mk3 - and/or the bearing sleeve/spacer 06.2069 acts as a reinforcing member to the axle. This sleeve/spacer is absent in the Mk3 design.

Rather than perform a re-design of the Mk3 rear axle, as a least resort I suggest the workshop manual be amended by prescribing the rear axle / dummy axle to be replaced say every 150,000 miles for safety reasons. "


As noted by Andover-Norton, examining service records also requires traceability of the fitted parts. This requirement can´t be fulfilled due to the various sources of parts.

Regards,

Knut Sonsteby

mdt-son said:

"As for the failure rate of the Mk3 rear axle, this is clearly a high cycle fatigue failure due to inherent design stress risers. Failure will depend on loading, road conditions, mileage, riding style, and possibly manufacturing and material defects. Several of the reported breaks have been on race bikes. The lean around a course will increase axle loading towards one side (mostly the left side), thus increasing bending and subsequent crack growth.

In order to dismiss the fatigue failure problem as irelevant, one really needs to examine the service record and rider´s notes, especially of those bikes which have covered substantial mileage.

Click to expand...

"Examine the service record and the rider's notes...." indeed. Is this serious!

mdt-son said:

At first glance the pre-Mk3 models should be even more affected due to their thinner axle

Click to expand...

At first glance it should be noted that although the Mk3 axle is thicker than pre-Mk3, the threaded section where it breaks is thinner (1/2" dia.) than pre-Mk3, not only that, but the Mk3 dummy axle isn't as substantial because it passes through the sprocket bearing and speedo drive gearbox, therefore the dummy axle isn't likely to be as rigid as pre-Mk3.

mdt-son said:

- maybe they are not used as much as the Mk3

Click to expand...

I doubt that. :wink:

mdt-son said:

- and/or the bearing sleeve/spacer 06.2069 acts as a reinforcing member to the axle. This sleeve/spacer is absent in the Mk3
design.

Click to expand...

While it's true the Mk3 assembly doesn't have the 06-2069 spacer, it does have the 06-5550 spacer but the Mk3 axle doesn't break in that area.

mdt-son said:

At first glance the pre-Mk3 models should be even more affected due to their thinner axle

Click to expand...

L.A.B. said:

At first glance it should be noted that although the Mk3 axle is thicker than pre-Mk3, the threaded section where it breaks is thinner (1/2" dia.) than pre-Mk3, not only that, but the Mk3 dummy axle isn't as substantial because it passes through the sprocket bearing and speedo drive gearbox, therefore the dummy axle isn't likely to be as rigid as pre-Mk3.

Click to expand...

As for the dimensions, both Mk3 and Mk1/2 spindles feature a 9/16 UNF thread (18 tpi) which has a basic minor diameter of 0.4943", thus no change there. Agreed, the Mk3 dummy axle is longer and therefore not as rigid as the pre-Mk3 version.

mdt-son said:

- and/or the bearing sleeve/spacer 06.2069 acts as a reinforcing member to the axle. This sleeve/spacer is absent in the Mk3
design.

Click to expand...

L.A.B. said:

While it's true the Mk3 assembly doesn't have the 06-2069 spacer, it does have the 06-5550 spacer but the Mk3 axle doesn't break in that area.

Click to expand...

That´s not my point. In the pre-Mk3 design, the sleeve/spacer 06-2069 provides additional rigidity to the spindle because it transmits a counteracting bending moment to the cantilevered spindle on account of the dual supported sleeve (proviso tight clearance between spindle and sleeve). In the Mk3 design, the 06-5550 sleeve is supported on one side only and can´t apply a counteracting bending moment. The ball bearing in vicinity of the spindle/dummy spindle joint is not able to restrain the spindle either. Therefore, the "cantilever" spindle (I am not sure it deserves to be considered a cantilever design) is quite flimsy and will deflect considerably more at the vulnerable spindle/dummy axle joint, mainly due to the poor support condition of the spindle.

One wonders who cocked up this design when they could have adopted the pre-Mk3 design quite easily?

-Knut

mdt-son said:

As for the dimensions, both Mk3 and Mk1/2 spindles feature a 9/16 UNF thread (18 tpi) which has a basic minor diameter of 0.4943", thus no change there.

Click to expand...

Only the Mk3 dummy axle thread is 9/16" UNF.

The Mk3 axle/spindle thread is 1/2" x 20 UNF (0.4387").

L.A.B. said:

The Mk3 axle/spindle thread is 1/2" x 20 UNF (0.4387").

Click to expand...

I stand corrected again! Thank you L.A.B.
The information is hard to come by and my source was obviously erroneous .... however, Old Britts list the thread size as well.

Sizing makes me wonder if the flexural behaviour is deliberate .... it`s frightening, in my view.
It also makes me wonder which material was selected and the fatigue life prescribed in design.

-Knut

I guess the two piece axle, especially in the case of the MKIII, should be considered a wear item and be replaced at some time. Has any data been compiled that shows the failure rate so that a reasonable replacement schedule can be established? According to a prior post AN seems to believe that 150,000 miles is the magic number. Would it be out of line to assume if a average rider were to change the stub and long axle at 50k miles he could sleep a little better at night? This seems to be the only solution at the moment for the MKIII riders. ( except for the inevitable " sell your MKIII and buy a real Norton" )
Pete

mdt-son said:

Sizing makes me wonder if the flexural behaviour is deliberate ....

Click to expand...

I don't know, but I think they did try to prevent it, or at least, reduce it. The axle thread can't be much larger than 1/2" because the shaft of the dummy axle it screws into is also 17mm diameter.

I'm not totally sure about this without taking the rear wheel assembly apart (and I've already had the axle out once, today :wink: ), but as the axle is screwed in, I think part of the unthreaded 17mm section is also drawn into the head of the dummy axle, and the tightness of the 'fit' is probably intended to increase the rigidity of the joint, however, repeated removal and replacement of the axle eventually slackens the fit and allows movement, so periodic replacement could reduce the chance of it failing.

I think part of the unthreaded 17mm section is also drawn into the head of the dummy axle,
If this is true, then there is obviously room to make the internal threads of the dummy 17mm as well, instead of necking it down to 1/2".
Enlarging the dummy overall wd then permit 17mm flats on the dummy to go with a 17 slot in the swingarm, no?

Mr. Rick said:

I think part of the unthreaded 17mm section is also drawn into the head of the dummy axle,
If this is true, then there is obviously room to make the internal threads of the dummy 17mm as well, instead of necking it down to 1/2".

Click to expand...

The head is only a comparatively short section and couldn't be made longer as it is the 'spacer' between the LH wheel bearing and the double-row sprocket bearing. The 17mm thread would have to be the same 20 tpi as the 1/2" thread and both thread pitches would have to be aligned or the two separate threads would jam as the assembly was tightened.

Mr. Rick said:

Enlarging the dummy overall wd then permit 17mm flats on the dummy to go with a 17 slot in the swingarm, no?

Click to expand...

It isn't possible to enlarge the diameter of the Mk3 dummy axle, at least not without increasing the size of the sprocket bearing which would probably involve changing the complete sprocket assembly (and modification of the speedo drive gearbox) because unlike the pre-Mk3 drum brake dummy axle, the Mk3 dummy axle passes through the 17mm double-row sprocket bearing.

Deets55 said:

I guess the two piece axle, especially in the case of the MKIII, should be considered a wear item and be replaced at some time. Has any data been compiled that shows the failure rate so that a reasonable replacement schedule can be established? According to a prior post AN seems to believe that 150,000 miles is the magic number. Would it be out of line to assume if a average rider were to change the stub and long axle at 50k miles he could sleep a little better at night? This seems to be the only solution at the moment for the MKIII riders.

Click to expand...

Just for the record, the mentioning of a replacement interval of 150,000 miles was my proposal for a service manual amendment. In view of L.A.B.`s recent post regarding details of the spindle / stub axle joint, I think a replacement interval of 75,000 miles is a better figure. These parts are cheap and replacing them is an easy task. As understood by the messages received from Andover Norton, no specific investigation has been undertaken to examine the failure rate of the two piece axle. I know A-N is observing this thread and I do hope they will issue a service bulletin or take a similar action.

All Mk3 owners should take notice and at least stay clear of non-approved parts.

-Knut

"All Mk3 owners should take notice and at least stay clear of non-approved parts."
approved by who?

mdt-son said:

Deets55 said:

I guess the two piece axle, especially in the case of the MKIII, should be considered a wear item and be replaced at some time. Has any data been compiled that shows the failure

Just for the record, the mentioning of a replacement interval of 150,000 miles was my proposal for a service manual amendment. In view of L.A.B.`s recent post regarding details of the spindle / stub axle joint, I think a replacement interval of 75,000 miles is a better figure. These parts are cheap and replacing them is an easy task. As understood by the messages received from Andover Norton, no specific investigation has been undertaken to examine the failure rate of the two piece axle. I know A-N is observing this thread and I do hope they will issue a service bulletin or take a similar action

-Knut

Click to expand...

Click to expand...

Knut,
I have personally never seen a snapped two piece axle, so what I am about to suggest is just speculation. The general consensus seems to be that the axle snaps due to metal fatigue as a result of flexing. Looking at the construction and the way the pieces mate, I wonder if overtightening the axle could be part of the problem.

When everything is fully assembled on the bike is it possible that:

a) the long axle can bottom out in the stub axle or

b) the shoulder (the area where the threaded section meets the unthreaded section) of the long axle contacts the corresponding area in the stub axle.

If overtightened, either one of these conditions would result in a high stressed area right where the axle seems to break. In my experience bolts that have snapped due to overtightening exhibit a different pattern on the broken section than those that have snapped due to flexing. As I said this is just speculation but I would be interested in your observations and opinion.

Thanks,
Pete

o0norton0o said:

I did exactly what you suggested. I bought a second swingarm and had a machinist mill the slots out to use a one piece 17mm axle. In my case, I modified the swingarm slots to use a yamaha cast wheel who's axle size was 17mm.

IF the only part of the MK3 rear wheel assembly which is not 17mm is the axle slot, then it might be easy to convert an MK3 to a one piece 17mm axle. The 17mm yamaha axle I used in my modified swingarm is available on ebay for about $15. I also bought the axle spacers on ebay too.

Click to expand...

Any details on the axle & spacers? Part#s? Donor bike Model/year? Sounds like the axle was used as is?

edit
o0norton0o said:

82 seca 550 in my case, but as I said there are even better choices that I discovered in later research.

74ok said:

Any details on the axle & spacers? Part#s? Donor bike Model/year? Sounds like the axle was used as is?

edit
o0norton0o said:

82 seca 550 in my case, but as I said there are even better choices that I discovered in later research.

CB750 rear wheel on 73 Commando ?

I'm unhappy with the cush drive, two piece axle, wobbly drum/sprocket etc. on my bike and rather than try to fix all that plan to swap in another wheel. Did see a reference here on people using CB750 wheels back in the day but there were no details...so if anyone knows more and/or has some...

www.accessnorton.com

Click to expand...

I used to have a few CB750's and their rear hub and brake drum are heavy!!

I have had 3 axles fracture over 40 years and very high mileages on my Mk3. They all fractured at the point where the shoulder reduces the diameter down to 1/2" and is then threaded. My personal view is this is caused by the manufacturing, design and wear on supporting components creating stress risers. if that is the correct term. I am not an engineer, but I smooth the transition point before the thread starts in the hope that I wont suffer another failure. It should be noted that I used to use my MK3 as a workhorse dispatch riding around London in the 80's and did very high mileage. Luckily non of the failures were at high speed...
If Don ever produces a mk3 version I will definitely put my hand up...