Main bearing failure

I fitted this FAG NJ306E.M1.C3 main bearing (without India stamped on)about four years ago and have done about 8,000 miles on it. I've had the crankcases apart for a reason unrelated to this bearing, but just prior to what was going to be re-assembly time, I noticed this :shock: :shock: :shock:



This is the first main bearing failure I've ever had on a Commando, and I've run one for about 30 years now approximately.

I am slightly puzzled why this has occurred at such a low mileage (and at all). A friend of mine who is an engineer has told me that bearings usually fail due to lubrication issues. How are these bearings lubricated? I assume by splash from the crankcase and oil from the rocker feed that drains from the back of the head into the case. I have done the mod where the three holes are blocked off in to the timing case and a new drain hole put adjacent to the lower end of the oil pump. I assume that this would result in the oil level in the case being potentially below the main bearing level although nobody else seems to have had an issue after doing this mod??
I'm puzzled?

Martyn,

I'm not a bearing expert, but IMHO that is a manufacturing defect. Either insufficiently deep hardening has worn through and allowed the base metal to get chewed up or an over brittle hardening has allowed / caused a chunk to break away.

Yes the mains are fed by splash. I do not believe the timing cover oil level plays a big part in main bearing lubrication, if it did, the primary side bearings would be in serious trouble!

I assume you're the kinda guy to use decent oil, and if the rest of your internals are showing no oil related issues, I would take this as further evidence of a manufacturing defect.

I was going to post in the other thread to say that Indian made bearings are fine IMHO, what matters is that they are genuine and of the correct spec etc. whether made in India or England, both can suffer occasional quality failures.

Sadly, your find seems to prove this point.

i'm also no expert but that has got to be a manufacturing fault ,nothing to do with lubrication in my opinion and lucky you found it when you did ,,,,,,,baz

baz wrote;

lucky you found it when you did ,,,,,,,baz

Click to expand...

That's very true. I had not noticed any symptoms such as noise or rough running, and as I said, I'm so complacent about main bearings now after 30 years without problems that I only discovered this by giving them a cursory inspection. I'm riding the Norton to Italy on one of my main holidays next year so am very pleased to have found this now.

Nigel wrote;

I assume you're the kinda guy to use decent oil, and if the rest of your internals are showing no oil related issues, I would take this as further evidence of a manufacturing defect.

Click to expand...

Millers 20/50 semi synthetic and changed every year. The big ends look fine, the other main bearing to my none expert eye looks ok and the cam and cam followers all look good.

Interesting that you think it may be a manufacturing fault? I might see if the bearing shop that I bought them from are interested in investigating this.

Thanks.

Nigel wrote;

Yes the mains are fed by splash. I do not believe the timing cover oil level plays a big part in main bearing lubrication, if it did, the primary side bearings would be in serious trouble!

Click to expand...

Yes good point and very logical.

It certainly doesn't show any marks due to interaction with rolling elements and no discolouration associated with heat/lubrication failure.

Add another one to the manufacturing defect club - possibly a void in the material or pre-existing fracture?
If it was a stress fracture you'd expect to see an initiating location with fracture lines radiating from it, but doesn't look like it to me, unless they're emanating from the centre.

where were the big-ends in relation to the failure location?

Andy wrote;

where were the big-ends in relation to the failure location?

Click to expand...

This is the timing side main.

Never seen a main
bearing do that. Wonder where the bits ended all up ?

Torontonian wrote;

Wonder where the bits ended all up ?

Click to expand...

There was quite a lot of "iron filing" debris on the magnetic sump plug..........which should've been a clue.
There didn't appear to be any in the sludge trap, so I assume most was taken out by the filter. I have still yet to dismantle the oil pump.

Wow, that is a first for me with a roller main. I would also have to think poor hardening would be the culprit although you might want to check and make sure the bearing bore and or shaft is round before you fit a replacement.
The "superblends" in my bike are still the originals with well over 100K miles on them. Jim

This looks like a defect in the steel possibly from when it was cast or during the later rolling operations, a weak point is hidden just under the surface which then erupts under the strain of the rollers running over it. Impossible to detect unless a part of it was on the surface so the grinding operation could expose it at that point. Bearing failure can happen when Manganese inclusions are just under the surface in the same way and here is an example which looks very similar.

Andy wrote;

where were the big-ends in relation to the failure location?

Click to expand...

Reggie said:

This is the timing side main.

Click to expand...

Andy was referring to what tick of the clock the damage is, in reference to the big ends.

If that was at bdc or tdc say, or near to, it could be the combustion stresses extra load was too much for the steel at that point.
Or, it could just be random damage...

it could be that your magnetic sump plug saved the oil pump. Maybe run that bearing around to a few supply houses and get their opinion. It does look like a manufacturing defect though. You were extremely fortunate to find this the easy way.

I'm with Comnoz, verify no out of round condition exists in case or crankshaft.

Rohan wrote;

Andy was referring to what tick of the clock the damage is, in reference to the big ends.

Click to expand...

Oh, right. I didn't get that

With the crankpins at TDC, the damage on the bearing was 180* away from the 12 o'clock position,.

I will double check the crankshaft and housing for out of round.

Detonation [pinging] could possibly contribute to the damage although detonation damage on a Norton will usually show up somewhere else first.

Hardening damage from using a torch can certainly present like that. Remember you don't have to make the whole bearing hot to cause heat treat damage with a torch. Pointing a torch at a bearing surface will overheat the skin of the bearing in a matter of seconds which will make the skin soft and the bearing will fail in use . Jim

I note the various posts suggesting possible causes for the failure ......Hmmmmmmm.
Might I suggest that before every Norton expert/exspurt in the World offers reasons for the failure that you actually send the bearing to the bearing manufacturer for them to examine and obtain thier thoughts on the matter........they could well be far more accurate. You never know, they MIGHT even send you a new bearing......
The Gentleman you should send the bearing to is .....
Mr Chris Head.
Applications Engineering.
INA FAG.
Schaeffler (UK) Ltd.
Forge Lane.
Minworth.
Sutton Coldfield.
West Midlands.
B76 1AP.
U K.

As others have added their pennies worth ......a few thoughts of my own .
1. Was the bearing coated with engine oil as the bottom half was assembled thus ensuring the bearing had some lubrication upon engine start? Rub two pieces of metal together unlubricated and galling takes place which eventually as sure as God made little apples results in premature failure....as with Norton Commando cams!
2. 1 was not done and the crank was not charged with oil prior to starting the engine so the motor had no oil flying around to lubricate the bearing..... till the crank eventually received some oil .....
3. From at VERY old memory.....the timing side main shaft has a bearing seat half a thou (?) smaller than the drive side, the inner raceway of the bearing being a light push fit on the shaft and located in place by the nut on that end of the crank. I assume the shaft is of correct size so that the inner raceway was a light push fit on the shaft and the nut was correctly tightened? Just a thought....as the drive side inner is an interference fit on the shaft I would of expected different internal clearance bearings to be employed but of course they probably were when one was a ball and the other a roller....in the good olde days before the stop gap model Atlas Mk3s or Commando came into the World!!. Looking at my very olde Hoffman book internal clearance chart...30mm I.D. bearings. CN fit. Ball bearing 0.0002 - 0.0008 inch Roller bearing 0.001- 0.0016 inch. Mind you a bearing designer once wrote stating the difference was because the raceways of ball bearings deformed under load due to the smaller area in contact allowing the internal clearance to increase sufficiently for correct lubrication to take place. Hmmm, another subject for experts and exspurts to post upon??? Me ?? I know sweet bugger all about Engineering......compared to the people I have consulted over the years and still consult that is......Never ceases to amaze me how REAL EXPERTS, if approached correctly, are prepared to go to great lengths to pass on their knowledge. trouble is many are no longer with us and idiots have taken their place.

Of course before roller bearings adopted 'barrel' shaped rollers the rollers were simply straight sided with no barrel shaping with just a small radius at each end and these were the ones that failed big time after about 4000 miles on early Commandos which led to Ransom and Marles taking an old fag packet and designing a special bearing with barrel shaped rollers so that when the crank flexed all over the place the edges of the rollers no longer dug into the inner and outer raceways which helped cure the problem...along with many other changes such as new better advance retard units that DID NOT fell apart at around 4000 miles use staying fully advanced all the time etc etc..... This 'barrel' shaping (logarirthmic profiling or whatever it is called)of the rollers appears to of been adopted by all or most bearing manufacturers since then. Incidentally the origional R and M 'superblend' bearings rollers had slightly greater barrel shaping than other manufacturers later bearings. i once popped out a few rollers and ran a micrometer over them.... ALSO the original R and M rollers were narrower along with having a norrower parallel section which I suspect allowed the crank to flex even more without the edges digging in.......I have NEVER seen an original R & M roller with wear markings to the very edge but with other manufacturers later bearings.....So why did they change to FAG bearings? I heard from several NVT friends it was because they were cheaper and the FAG office was just down the road!!
An interesting artilcle to read is contained on pages 27/28 of the January 1977 issue of Motor Cycle Sport ....it discusses the SERIOUS main bearing problem suffered by a certain motor cycle manufacture somewhere in Europe
......the many causes of the problem and the many changes / fixes required to cure it. Methinks the initials at the end of the article are those of a certain motor cycle manufacturers Service Manager.
Pahaps the failure cause is a material or hardening problem but there will not be many if any posting on this and any other Norton site capable of determining that. ASSUNING the bearing is actually returned to the manufacturer for inspection it would be interesting to read the bearing manufacturers report.....

Interesting background. Thanks for posting. Nice to capture some real info from someone who was there.

Ken

I'd comment that I read that MotorCycle Sport article back then, and had a look at the rollers in the FAG bearings
that were in the wrecked 850 that I'd bought to rebuild back then.
If the rollers were anything but plain parallel rollers, the micrometer couldn't see it.
That 'logarithmic profiling' must be awful small - or BS ??
The ends of the rollers were slightly rounded (at the very ends) so they wouldn't llikely easily dig in though.

And when I went to the local Norton Dealer to buy some replacements, he tried to sell me plain FAG NJ306 bearings.
He'd plainly not read about the 'superblends' - and was offended when I wouldn't buy them.
Very big Norton Dealer too.
(He was the boss man, maybe he didn't know what the workshop knew ?!)

See also Jims take on the parallel rollers in FAG main bearings, just posted.
superblend-bearing-profile-t23638.html

Kills that myth about barrel shaped rollers stone dead.