Roadholder 304

Previously wrote:

Hello, all.

Just had Roadholder 304 delivered today. Having made a start reading it, the good lady points out that the machine pictured on the back page looks familiar. Lo and behold my Atlas in all its glory at founders day! Never had one of my machines in a magazine before let alone my favourite one. I had a good day at founders day finding one or two bits and pieces for both the Atlas and Commando. However the day did,nt end up very well as the Atlas developed a nasty knocking noise under load on the way home. Made it home ok and promised the bike a thorough investigation before its next ride out. A few months later, the strip down begins. The horrible truth is soon revealed! Both main bearings are not only loose on the crankshaft but also in the crankcases. Both bearings appear to be origional being ball race on the timing side and a roller race on the drive side. My question/request to you all is would it be advantageous to fit superblends (if i can get some without taking out a second mortgage) or would you go for the origional set-up. I would think loctite would also be needed to retain the bearings in both locations. Any of your experiences in this area would be greatfully recieved. Lets hope for some warmer weather soon.

Best regards to all,

Andy.

Hello Andy the R&M MRJA30-6 or 3MRJA30M drive side AND timing side R&M MJ30 are the right bearing you need and they are not cheap.

yours anna j

I fitted superblends to my Atlas about 20 years ago and it's still going fine, despite being used to its full potential rather a lot. However, I have always fitted the original ball/roller bearings to 88/99 rebuilds successfully - well, except once I did have a drive side main fail at Fort William on a - believe it or not - warm and sunny day on my way back from Skye. Conventional wisdom has it that superblends are a must for 750 and bigger engines. I'm not entirely convinced. Main bearing failures on pre-Combat engines aren't that common. Yes, go for the correct grade of Loctite.

A couple of years ago I used bearing fit loctite to try to prevent the outer race turning in one of my cases. I did think it out and did as Mike Pemberton suggested in last month's Roadholder - heat the case first and put the bearing in very quickly indeed... so when the case warms up when running it does not expand outwards away from the bearing.

Of course I have no idea if the theory is sound...maybe one day I'll have it all in bits and then I'll find out.

Loctite don't seem to provide any advice. I assume it's usually used to fit a steel race on a steel shaft.

Previously wrote:

I fitted superblends to my Atlas about 20 years ago and it's still going fine, despite being used to its full potential rather a lot. However, I have always fitted the original ball/roller bearings to 88/99 rebuilds successfully - well, except once I did have a drive side main fail at Fort William on a - believe it or not - warm and sunny day on my way back from Skye. Conventional wisdom has it that superblends are a must for 750 and bigger engines. I'm not entirely convinced. Main bearing failures on pre-Combat engines aren't that common. Yes, go for the correct grade of Loctite.

Hello R&M MJAR30-6 is a what is know as Super blend Bearing , in other words a spherical Bearing meaning barrel shaped. roller bearing first made by R&M bearings .later copied by FAG of Germany

I think the superblends may be a bit of overkill on the pre-commando twins, but Ipersonally always install the superblends. Yes its a bit more money, but it is a better bearing. I have them in my model 77, model 7, and will have them on my Nomad also. If you go for the superblends, be sure to check your end play of the crank, you might need to use crankshaft shims if the end play is too much.I would also highly recomed purchasing the Norton Twins rebuild CD by Mick Hemmings for sale on this very site, loads of good info & tips!

Skip

If you quote "superblend" to any FAG supplier he will not have heard of the term and what we know as superblend bearings are certainly not spherical or barrelled. They are a much better quality product and that is all. The term superblend is a Norton generated myth to dispell mistrust in combat engine weaknesses.

You are correct about using the term superblend to a FAG supplier David, but our Norton parts suppliers all seem to list them as superblend, as opposed to ball & race type, so it seems that the term has become standard when ordering bearings, unless of course you went to a bearing supplier, but there are so many variables & classes, I would only buy from my trusted Norton parts dealers.

Thought I might add this quote from this very websites tech-support page under crankshafts:

"A Superblend is basically a roller bearing, but actually, its a barrel-race bearing. The 'rollers' are beer-barrel shaped and thus provide a broader surface of support than a simple ball bearing, whilst allowing for flex in the shaft they are supporting, a thing a plain roller can never do until it fails. Two Superblends offer nearly as much flex as the original ball/roller combination, but they also allow for uniform flex, along the length of the shaft, rather than isolating it towards the ball bearing end of things, like the combined bearing setup does. They each offer far greater support than a ball bearing can, but not as much as a plain roller bearing. Generally, they are a Good Thing. "

With so much information available choosing main bearings can be difficult, I do not have any experience above 650 so cannot comment on what is best for a 750.

When building my 650 I fitted roller drive side No. SKF NJ 306 ECP and ball race timing side No. NTN RL306 94-04 this is an 11 ball bearing.

I would take on board Gordon's recomendations as an Atlas owner.

Tony

Hi Eugene,

You should not always believe what you read on this web site !. Further on down the thread where you saw :-

A Superblend is basically a roller bearing, but actually, its a barrel-race bearing.The 'rollers' are beer-barrel shaped

Is the comment:-

Thanks to Vernon Fueston for all the good work in researching this 'Superblend' business. Now I know what it means! Essentially nothing in-so-far as bearings are concerned.

'Superblend' is the result of a PRman'slunchtime brainstorming session (total guess, no basis for this !), the rollers are not barrel shaped although they do have a slight radius on the ends.

Have a look athttp://www.santiniketanenterprises.com/cimage12/106247srcylindricalrollerbearings.pdf

And you will see what they really look like.

I used to believe all the 'barrel shaped rollers' stuff and have been guilty of perpetuating it in the past.

Tony

Hi Tony,

thanks for that. I just had a go round with a fool this weekend stating that the rollers were perfectly cylindrical, no taper at the ends etc. It seems that there were different superblend bearings also, different cages, & different rollers even. The term logarithem curve has also been mentioned on one bearing website. It seems that the problem is the auctual term "barrel", it can conjure up different images. Here is a quote from the access Norton site by Matt Spencer:

http://www.accessnorton.com/rpm-flex-bearings-cases-galore-t14516.html

The Deal is the LOAD bearing ( pun ) CAPEABILITY . Been wondering if Plain Rollers'd be fine in a Race Motor , Norton . IF CHANGED every 1.000 miles . As youd get more HORSEPOWER , as ball races absorb less . Less Drag .Young Dunstall did his big Eight Ball Heavy Duty roller Bearings , slotted to accept balls ( notched so they fed in . ) these were MORE RIDGID .So the CRANKSHAFTS BROKE . The Ransome & Marles and Hoffman BARRELLED ROLLERS allow axial misalignment ( as per Hiem Joint / Sphereical connector , to a limited degree .Therefore have the LOAD BEARING capeability of a plain roller bearing ( CONTACT Area . ) and a capeability of accepting missalignment . A Ball Race has Niether .The BARRELLED ROLLERS are Just That . Not as some berks infer ' Rounded ends to stop them digging in ' ( The Ends of the rollers apparently did , if plain roller )the infernal things are produced with a large radius , like a wooden wine / beer barrel , as are the tracks . The Crank can whip with the power impulses without cracking up the next week link . Just like the H.D. XR 750 . Therfore you need a near ridgid crank with the other bearings .

My Timken bearing rep. and I have discussed this bearing at length, and as we all know by now, the term Superblend is a Norton term not used at bearing houses.

A very good thread on the "Superblend" bearing on the Brit Bike Forum :

http://www.britbike.com/forums/ubbthreads.php?ubb=showflat&Number=444274

also concludes that the rollers are NOT perfectly flat. The term "barrel" really seems to be the issue, what someenvision as barrel shaped is different from another.Or I guess we would need toaccuratelydefine the terms barrel,logarithmic curve, parabolic, etc. Perhaps someone here can tell us thedifference?

Here is anexcerptfrom the above link:

Edited by Dave Comeau (07/12/1212:31 pm)

Hopefully this will add somehelpfulinformation to our archives on this issue.

Skip Brolund

This myth of barrel shaped rollers in FAG or similar high spec bearings never seems to go away. Much of the naff info contributed by 70s magazine atricles written by poorly informed contibutors.

A small amount of appliedsimple logic always helps with these matters.

If the rollers were barrel shaped, then both inner and outer tracks would also have to be curved to match the rollers in order togive a better loadadvantage than a ball bearing.In which case you would not be able to pull them apart as can be done with FAG/Superblends.

As Tony has pointed out, and others noticed, the FAG bearings have radiused endsto prevent them from digging in to the tracks. Also in many cases, thestandard roller bearings, they are replacing,often have slightly shorter rollers and sometimes less of them. Both of which give the FAG bearing a higher loading advantage.

To add to the confusion of barrel shaped roller bearings.......they do actually exist.....but I have never seen them in a Norton engine. They were designed to offer a 'self-aligning' bearing for application in machines where shafts were run in housings subject to flexing. The first ones that I ever came across were inside some industrial fan units bolted to the the top of some extractor ducts.

Tony and Phil are exactly right, barrel shaped rollers would be loose in the outer races until in the correct lateral position. Self aligning bearings will take up the "out of alignment" errors in some poorly engineered situations but even our Norton crankcases are not that bad. Let's just say that "superblends" are more able to deal with the Norton crankshaft flex because they are a better quality item. The only barrel shaped bit on my Norton is me.

Good point Phil, you would not be able to pull the bearing apart if they were barrel shaped. If you have a look at :-

http://www.santiniketanenterprises.com/cimage12/106247barrelrollerbearings.pdf

Then you will see this is what everyone is calling the Superblend and what I always thought they looked like (and I was wrong). The information states:-

The axial load carryingcapacity of the barrel roller bearings is limited.The bearings are not separable.

The load and speed rating for the NJ306E is 51kN and 10000RPM

The load and speed rating for theequivalentbarrel roller bearing is 40kN and 5000 RPM

So, the manufacturers data for the drive side roller bearing states the rollers are cylindrical.

Tony

Hi all,

wanted to "cross post" this "Superblend" info from another Norton site, I think you will find itinterestingalso, discuss:

I think to much reliance and trusting that internet forums are technically correct or even theoretically seasoned mechanics have engineering evidence to back up their preaching's. I am first to be judged, but hear me out and see if it sounds reasonable.Individual points that I pretty much accept:1. These bearing are no longer pure mathematical cylinders...even if they might have been at one time in the stone age. Current use of this term in catalogs is a VERY crude general marketingcategorydescription and has no technical value.2. They are not now and never were barrel shaped though the term is supposed to give a slight impression of the final effect...... Dangerous term to use.3. I very highly doubt the term in any engineering sense ever came up that they are logarithmic; this term has no physical shape meaning and only describes a numbering system for powers of ten.4. IIRC the term is super hyperbolic came from FAG's technical pages and the 2 joined (highly modified) hyperbolic functions describe the machined shape of 1 roller. See the general shape of a cosh (x) parabola here... the blue line. I'm guessing it should be cosh (y) to put it on it's side:http://en.wikipedia.org/wiki/Hyperbolic_function

The use of this function for the roller is more by using the graph end that more closely asymptotes a line as it approached infinity and the almost flat line becomes the basic center part of the roller. This gives a human the optical view/impression of a cylinder, which it is not. The ends are microscopically curved and I can not accept that it is a millimeter or much more than a "mil" or .001 inches. I highly doubt the visual rounding of the end of the roller is in any way part of the super hyperbolic effect. I also am guessing it is not the size of reduced diameter but rather the length of the arc which would spread the load over a much larger surface during the misalignedmoving on to norton case:5. The mains are mounted in aluminum. This is not essentially an engineering acceptable situation. I reference the bearing companies application guides.5a.This is the primary reason for the C3 bearing clearances.5b. If the bearing drops in or does not have the original shrink fit, the cases are toast. Loctite or other engineered bodge is often used.5c. When the cases heat up, the shrink fit is reduced or eliminated and the mains do spin in the cases.moving on to the Norton crank:6. The crank flexes if overrevvedbecause of overloading the material strength of the components. The force applies is from the center flywheel being centrifugally spun and being restrainedthroughthe bearings supported by the cases.6a. The crank shafts are bent out of a straight line and this angular deflection is like wise transmitted through the rollers. I have not seen it, but believe the strobe experiment was done and the final result and conclusion are IMO credible.6b. Even if the R&M "super blends" or the FAG super hyperbolic replicas eliminate main bearing brinelling failures, the load stresses appear to still transmit to the cases and crack early combat cases or even others. I've seen the results myself.7. The crankmain shaftshould a press fit in the inner race. A slip fit means a dead shaft. IMO loctite is a sorry bodge. With a small...high quality bearing puller (OTC in the USA) you can EASILY remove the inner race repeatedly without damage. IMO heat is a HUGE NO NO. For this reason I always and only put shins ON THE CRANK.back to cases:8. The cases should always be heated to install or remove mains. I have seen ball bearing both come out of the cases with the crank and some times stay in the case. The ball bearings that stay with the cases make me think the crank journal issmall.8a. Combat and hot rod RPM capabilities, makes this fur ball problem hard to precisely pin down. There are several variations of Norton heavy twin cranks and possibly as many as two dozen variations of crankcases.

The above illustrates some of the intractable problems that engineers face. Older crank case sides were probably not very stiff. They had flat side walls with a few radial webs that did not really fix the central hole rigidly to the main block. So a slight 'wobble' imparted to the case due to the crank shaft swinging out of line like a skipping rope didn't harm the bearings because the crank case allowed for the misalignemt. But constant flexing leads to fatigue cracks - so let's have stiffer cases..

Now the shaft out-of-alignment ends up stressing the bearings...

There was no such thing as finite element analysis to assist designers. But if they had used it they would no doubt have frightened themselves with the high theoretical stresses that would turn up.

The problem is that the more you stiffen one element in a structure or machine the more you attract loading into it. So the advertising says they used 'barrel' shaped rollers - as Eugene says that's too simplistic but it's probably a reasonable fair description for the fact (if it is a fact!) that the bearing rollers are not prefect cyclinders.

Bearing company web sites (Timken is an example) claim that their roller bearings are best partly because of the 'refined geometry' of the rollers. They certainly imply that they are not perfect cylinders. Rather like barrels...

And we know bearings come in a wide range of quality. About 15 years ago a friend of mine bought a slightly odd size roller bearing for a crankshaft for a vintage racing car - and was a bit stunned to be charged the best part of £100. And it came in a box saying 'not approved for helicopter gearboxes'. I bet helicopter shafts don't stay perfectly straight (since nothing is perfectly stiff).