Skip to main content
English French German Italian Spanish

Superblend barreled bearings - the myth goes on....

Forums

I noticed in a feature of the October edition of Roadholder barrel-shaped Superblend bearings get a mention.

Purely in the interests of technical accuracy please note barrel roller bearings were never fitted to Commandos - and here's why. First off - barrel roller bearings have the rollers assemled to the inner race - the opposite of NJ306E's. Second - with the inner races fitted to the crank and the outer races fitted to the cases it would be impossible to assemble the crank to the cases because the inner and outer race bearing track is curved to match the barrel rollers, so there would be no way the two components could slide together in an axial orientation.

With NJ306E's fitted it is only possible to assemble these because the rollers and inner and outer race bearing tracks are parallel to each other - not curved as in the case of barrel roller bearings.

Call NJ306E's barreled superblends if you want - but you might as well state the sun goes around the earth.....which is flat.

Permalink

While I agree that technically the rollers are not shaped like say, a wooden beer barrel. The do have radiused corners which do not dig into the races as the square corner rollers of the non superblend bearings would upon crank flex..

I had contacted FAG in Germany & also Timken in the USA & they seem to have a slightly different take on your findings.

I will try to find the emails & post them here.

Permalink

I believe the rollers used in Superblend bearings are barrel shaped. This is from reading many technical articles over the years. Maybe if you post a picture of the bearings you are talking about we can discuss those and why they are different.

Permalink

The exact shape is not that important. They call them "barrel shaped" because it is the most convenient way to describe them. I was hoping Simon would post a picture so I could help with the technical terms. What Simon is most likely talking about are "self-aligning roller bearings" which have curved races. These are not suitable for automotive applications. I don't think anyone was suggesting using an unsuitable bearing in a Commando.

Permalink

I sum it up this way super blend is street talk = super hyperbolic is bearing manufacturers description of a design/manufacture characteristic.

I read the FAG web site many years ago. I tried to reread the article more recently but you must sign up now to get access, so I didn't bother. I believe the tern is something like "super hyperbolic" to describe the shaping treatment on the end of the cylinderical roller. Barrel shaped is a huge exageration of the effective curve. You can barely measure the end treatment with micrometers... let alone SEE it.

To me super blend means super small... almost invisable.... not super big shaping like a barrel.

Otherwise the antique bearing were cylinders. Today virtually all roller bearing have the superhyperbolic effect.

added:

by definition a hyperbolic curve asymtotes a straight line but NEVER -ever becomes straight.

A cylinder has sides that are "straight" and NOT curved at all..

Permalink

FAG do not call the NJ range of bearings "barrel shaped" - they call them "Cylindrical Roller Bearings", a cylinder, as already mentioned has parallel sides.

They call their range of barrel roller bearings (beginning with part no. 202 or 203) "Barrel Roller Bearings". I guess this is to save any confusion between the two.

The barrel roller bearings cannot be dismantled, whereas the cylindrical roller bearings can - as anyone who's rebuilt a bottom end with NJ306e's will know. Other information is from FAG Standard Programme catalogue 41 510 EB.

There's more info available at www.fag.de

Permalink

If you type barrel shaped bearings andsuper blends barrel into the search box above there is a whole tankful of postings from past years giving sensible info about these bearings. Plus the odd bit of text correcting the myth perpetuators.

Permalink

Both FAG and SKF quote 4 minutes of arc allowable angular misalignment. SKF describe the profiles of their 'cylinders' as being a logarithmic curve and thus they are not perfectly prismatic cylinders.Without any historical knowledge it is guesswork - but my guess is that for many years since they were invented, cylindrical rollers were just cylinders. Like a lot of things they have become a bit more sophisticated - or at least those from the top manufacturers are. Norton were forced to get the most expensive bearings and gave them a market name. Barrel shaped may be crude but it's a gross exaggeration, not a falsehood. And a bit less confrontational than saying 'logarithmic' or 'hyperbolic'.Top bearings have always been costly and rare and, sadly for us, usually Swedish or German. They flew high speed Mosquito aircraft to Sweden during the war to buy roller bearings for Merlins. I bet they were 'hyperbolic' as well! And for the same reason.
Permalink

Previously simon_ratcliff wrote:

I noticed in a feature of the October edition of Roadholder barrel-shaped Superblend bearings get a mention.

Purely in the interests of technical accuracy please note barrel roller bearings were never fitted to Commandos - and here's why. First off - barrel roller bearings have the rollers assemled to the inner race - the opposite of NJ306E's. Second - with the inner races fitted to the crank and the outer races fitted to the cases it would be impossible to assemble the crank to the cases because the inner and outer race bearing track is curved to match the barrel rollers, so there would be no way the two components could slide together in an axial orientation.

With NJ306E's fitted it is only possible to assemble these because the rollers and inner and outer race bearing tracks are parallel to each other - not curved as in the case of barrel roller bearings.

Call NJ306E's barreled superblends if you want - but you might as well state the sun goes around the earth.....which is flat.

Hello Now the real superblend bearing was made by R&M from 3MRJA30M the new bearing was R&M 6MRJA30B And every other make of Superblend Bearing is a Copy of R&M 6MRJA30B this is the correct information yours Anna J
Permalink

Previously simon_ratcliff wrote:

The barrel roller bearings cannot be dismantled, whereas the cylindrical roller bearings can

Good. Well then, when a chap writes in saying "Help, I can't dismantle my main bearing!" We will point him to this post. Agreed?

Although those self-aligning FAG bearings don't come in a Norton size so I don't see it happening soon.

Permalink

David, thanks for pointing out the term logarithmic profile. Although I don't really understand what SKF literally mean by a 'logarithmic roller profile' (can someone explain?), except that looking at the accompanying diagram from their website (google; skf cylindrical roller bearings and then click on logarithmic roller profile) I interpret it as a cylindrical roller bearing i.e a roller bearing with parallel sides, and the logarithmic part is the radiussed corners which spread the load due to crank flex for example, which was mentioned before.

So not barrel shaped. The clue is in the name - cylindrical. I'll repeat - if the rollers were barrel shaped then it would not be possible to separate the inner and outer race with the rollers in situ.

Permalink

Previously simon_ratcliff wrote:

David, thanks for pointing out the term logarithmic profile. Although I don't really understand what SKF literally mean by a 'logarithmic roller profile' (can someone explain?),

I think it means barrel-shaped. Races are parallel though, so any shape roller can slide in and out, for instance, when you take the bearing apart.

Permalink

Jonathan,

Barrel roller bearings run in suitably shaped (barrel possibly?) races. Please refer to previous posts (and pictures). Once again the clue is in the name. A barrel roller running in a race designed for a cylindrical roller i.e parallel surfaces would have the load concentrated over a relatively small contact area and therefore suffer from high wear rate.

Thanks for helping with the technical terms.

Permalink

Previously simon_ratcliff wrote:

I noticed in a feature of the October edition of Roadholder barrel-shaped Superblend bearings get a mention.

Purely in the interests of technical accuracy please note barrel roller bearings were never fitted to Commandos - and here's why. First off - barrel roller bearings have the rollers assemled to the inner race - the opposite of NJ306E's. Second - with the inner races fitted to the crank and the outer races fitted to the cases it would be impossible to assemble the crank to the cases because the inner and outer race bearing track is curved to match the barrel rollers, so there would be no way the two components could slide together in an axial orientation.

With NJ306E's fitted it is only possible to assemble these because the rollers and inner and outer race bearing tracks are parallel to each other - not curved as in the case of barrel roller bearings.

Call NJ306E's barreled superblends if you want - but you might as well state the sun goes around the earth.....which is flat.

Hello now FAG SKF NSK Hoffmann Tinkin .Are All Copies OF the original British made R&M 6.MRJA30 the 650 Norton Manxman for November 1960 was fitted with R&M 3.MRJA30 now the R&M 6.MRJA30 was fitted to the Short Stroke Combat Commando engine and became known has the Superblend bearings yours Anna J
Permalink

"Roll out the Barrel" was written by a Czech prior to the Second World War so the bearings were probably a foreign invention too...

While we're at it, why are Mk3 850s reputed to have 'Vernier' isolastics rather than simply 'adjustable' and don't forget the old pedant's favourites...the plural front forks and the rear pivoting fork that is always referred to as a swinging arm...but we all know what they mean.

Permalink

In many highly stressed applications the cylindrical roller cannot be âtoo cylindricalâ â otherwise severe edge loading will occur. Perfectly cylindrical surfaces in contact always have higher stresses at the edges. Furthermore, this edge loading can worsen if the roller is slightly misaligned or if the components bend or distort during operation. By putting a tightly controlled âcrownâ on the roller (making it barrel-shaped) at the micrometer level, the loading of the bearing can actually be made more consistent over the length of the bearing.The attached figure shows how different types of surface profiles induce different kinds of contact stresses on the inner race at low and high loading.

I think the important thing to know about these bearings is that they are designed for extra load bearing capacity and only have extremely limited self aligning properties for dealing with a flexing crankshaft. The logarithmic profile being the preferred, for dealing with the elastic deformation of the roller when reacting to the centre of pressure acting upon it.

I think the main question to ask here is how the main bearing debacle came about in the first place. If you read the Norton service release bulletins, available on this website, regarding this matter it paints a very confusing picture of what must have been blind panic at the factory in 1971 and 1972. Are the engineers who wrote or approved these technical instruction still around to explain what went on at the time?

Attachments roller-profile-jpg
Permalink

Jonathan,

You are obviously confused between the difference of a spherical roller bearing (which I've not mentioned in any of the above posts) and a barrel roller bearing, so please don't make statements of what I'm refering to when you clearly don't understand.

The jpg. roller profiles previously posted clearly show the difference between barrel, cylindrical and logarithmic profiles. I agree with previous postings that the NJ range of bearings have the logarithmic profile which is a cylindrical roller bearing with radiussed corners. The vertical lines in the diagram below each roller indicate distribution of loads and shows why the logarithmic profile is the better design.

The NJ range of bearings are not designed to cope with misalignment - which is the reason for barrel roller bearings.

Permalink

It is good that contributors have helped clarify that there are three distinct types of roller bearing mentioned in this thread: Barrel (for construction installations where the shaft may not align with the housing); Profiled (for performance engines where there is flexing of the shaft that the bearing will not be able to resist); and Cylindrical (the basic version).

The Norton Superblends are the Profiled type but people have and will continue to use all three terms for them. I don't see it as reasonable to criticise anyone for that.

What I find interesting is what discussions might have gone on inside Norton when the Commando engine was developed and plain cylindrical roller bearings were specified. Engineers knew that the crankshaft flexed, but the company had got away with it so far, and reducing every penny of expenditure was critical. The recall of the Combat must have cost a bit more than the pennies saved.

Permalink

Previously simon_ratcliff wrote:

Jonathan,

You are obviously confused between the difference of a spherical roller bearing (which I've not mentioned in any of the above posts) and a barrel roller bearing, so please don't make statements of what I'm refering to when you clearly don't understand.

Hmm... You now have 2 places to direct your anger. Martin and me. At least I have spread the load.

Permalink

Previously barry_carson wrote:

i didn't know a bearing could cause so much friction.

I just checked my early Commando crankcases and there is a ball bearing in the timing side but, as we know now, they would not have done that because those bearings don't come apart.

Permalink

Previously Jonathan Soons wrote:

I just checked my early Commando crankcases and there is a ball bearing in the timing side but, as we know now, they would not have done that because those bearings don't come apart.

I'd be interested to know if this bearing is the same as the one fitted to a set of 1971 Commando crankcases I purchased some time ago from the USA on ebay. It's a MJ30 ball race and has been marked with a single circle denoting C1 or 1dot internal clearance. See attachment. The roller bearing outer on the drive side was a MRJA30 which also had a single circle and displayed heavy contact and some surface breakdown. I believe that both these bearings were the originals as fitted by the factory. The point is that these bearings with the C1 internal clearance are high precision bearings not intended to be an interferrence fit either in the crankcase housing or on a crankshaft and is probably the reason why the roller bearing was on it's way out.

Attachments img_2088-jpg
Permalink

Previously Jim Brierley wrote:

Previously Jonathan Soons wrote:

I just checked my early Commando crankcases and there is a ball bearing in the timing side but, as we know now, they would not have done that because those bearings don't come apart.

I'd be interested to know if this bearing is the same as the one fitted to a set of 1971 Commando crankcases I purchased some time ago from the USA on ebay. It's a MJ30 ball race and has been marked with a single circle denoting C1 or 1dot internal clearance. See attachment. The roller bearing outer on the drive side was a MRJA30 which also had a single circle and displayed heavy contact and some surface breakdown. I believe that both these bearings were the originals as fitted by the factory. The point is that these bearings with the C1 internal clearance are high precision bearings not intended to be an interferrence fit either in the crankcase housing or on a crankshaft and is probably the reason why the roller bearing was on it's way out.

Mine has MJ30 crudely engraved in the outer race, also 97 and ENGLAND and a little circle with nothing in it. On the inner race R&M , 97 and 30M.

Permalink

Previously Jonathan Soons wrote:

Previously simon_ratcliff wrote:

You are obviously confused between the difference of a spherical roller bearing (which I've not mentioned in any of the above posts) and a barrel roller bearing, ....

Hmm... You now have 2 places to direct your anger. Martin and me. At least I have spread the load.

Jonathan,

Once again you have come to the wrong conclusion after (presumably) reading my posts and then misunderstanding the content. No where have I said I was angry because of an incorrect description of NJ cyl. roller bearings.

I stated at the start of this thread, "purely in the interests of technical accuracy...".

Maybe a bit more time spent 'Hmm-ing..' before posting would be a good idea - purely in the interests of technical accuracy, you understand.

P.S My understanding of logarithmic in the case of NJ bearings is two lines at 90deg to each other with a very small radius. Are there any mathematicians out there who can explain the actual definition as applied to cylindrical roller bearings?

Permalink

Once again you have come to the wrong conclusion after (presumably)

I stated at the start of this thread, "purely in the interests of technical accuracy...".

As long as you realize this is not terribly important, let me remind you what you really did say (purely in the interests of truthfulness):

Previously simon_ratcliff wrote:

I noticed in a feature of the October edition of Roadholder barrel-shaped Superblend bearings get a mention.

Purely in the interests of technical accuracy please note barrel roller bearings were never fitted to Commandos

Martin never mentions "barrel rollers" only you do and you should stop it.

so there would be no way the two components could slide together in an axial orientation.

Yes there is: put the bearing into the heated case then slide crank in like we always do with ball bearings!

Call NJ306E's barreled superblends if you want -

Again you are the only one to do that. Martin said barrel-shaped Superblends which we all do more or less accurately (see diagram - forget beer kegs). And so it is necessary to come to the defense of someone unfairly criticized. It was a good article but a 1972 850? is there such a thing?

Permalink

In case anyone is interested, I am uploading a logarithmic curve from Google. The key feature is it starts out very steep and then it flattens out. So from the flat side of the roller to the rolling surface it gets flatter and flatter. This diagram is deceptive though because it gets flat very quickly. The numbers on the axes (very large on the left axis, very small along the bottom axis) are scaled to hide this.

Attachments 11_8_log_graph-jpg
Permalink

As some of the contributors, to this thread, are starting to swamp it with accuracy down to a micro-degree. I just felt the need to point out that the word barelled has a double 'l' residing within. This is not a personal criticism of anybody who has thrown a few thoughts and measurements into the points pot and not spelled/spelt the word correctly. Personally the odd spelling gaff does not bother me in the slightest. I am more interested in the accuracy of the general content and welcome all contributions as long they do not degenerate into personality punch-ups....... as has happened with previous red-hot postings.

Permalink

More info on logarithmic profiling of cylindrical roller bearings.

www.ijrmet.com/vol3issue1/a1006.pdf

I haven't read this yet and wont even pretend to understand all of it.

Jonathan, the fact the curve gets flat very quickly and that there is a big difference between the numbers on the x and y axes is the nature of logarithmic graph.

P.S

Just read the abstract of the above pdf and it confirms what I, and others (no, not you Jonathan) thought is the shape of the rollers used in cylindrical roller bearings i.e parallel bearing surface (therefore cylindrical) with radiussed edges.

Permalink

Previously anna jeannette Dixon wrote:
Previously simon_ratcliff wrote:

I noticed in a feature of the October edition of Roadholder barrel-shaped Superblend bearings get a mention.

Purely in the interests of technical accuracy please note barrel roller bearings were never fitted to Commandos - and here's why. First off - barrel roller bearings have the rollers assemled to the inner race - the opposite of NJ306E's. Second - with the inner races fitted to the crank and the outer races fitted to the cases it would be impossible to assemble the crank to the cases because the inner and outer race bearing track is curved to match the barrel rollers, so there would be no way the two components could slide together in an axial orientation.

With NJ306E's fitted it is only possible to assemble these because the rollers and inner and outer race bearing tracks are parallel to each other - not curved as in the case of barrel roller bearings.

Call NJ306E's barreled superblends if you want - but you might as well state the sun goes around the earth.....which is flat.

Hello now FAG SKF NSK Hoffmann Tinkin .Are All Copies OF the original British made R&M 6.MRJA30 the 650 Norton Manxman for November 1960 was fitted with R&M 3.MRJA30 now the R&M 6.MRJA30 was fitted to the Short Stroke Combat Commando engine and became known has the Superblend bearings yours Anna J

Myth What Myth I all ready told you All how the Superblend Bearing came about and there is NO BARREL SHAPED ROLLERS IN SIDE IT!! it the Roller end have more of a Radius Too them so they do not bind in the hosing at high speed so they do not get has hot, SO THEN THIS MYTH IS NOW OVER !!!! yours Anna J
Permalink

Previously simon_ratcliff wrote:

Jonathan, the fact the curve gets flat very quickly and that there is a big difference between the numbers on the x and y axes is the nature of logarithmic graph.

I do not have a dog in this fight and do not even like Superblends. I do plan to, one day, write an article for Roadholder (being an expert in my own mistakes). Although I have no fear of making a fool of myself, I'm sure there are others who have valuable experiences to share but will be discouraged. Please let's establish a precedent of sticking up for fairness.

Permalink

Previously simon_ratcliff wrote:

More info on logarithmic profiling of cylindrical roller bearings.

www.ijrmet.com/vol3issue1/a1006.pdf

I haven't read this yet and wont even pretend to understand all of it.

Jonathan, the fact the curve gets flat very quickly and that there is a big difference between the numbers on the x and y axes is the nature of logarithmic graph.

P.S

Just read the abstract of the above pdf and it confirms what I, and others (no, not you Jonathan) thought is the shape of the rollers used in cylindrical roller bearings i.e parallel bearing surface (therefore cylindrical) with radiused edges.

Well all rollers are CYLINDRICAL Shape if they where Square they would not be rollers would they and there not Barrel Shape ether , there have just that bit more Radius on the ends so they do not bind at high speed rollers have a very hard life , has in a Norton there only Splash feed with oil . I have Not a clue what some are going on about FAG bearing's or Timken Bearing's When You Can still Get the Real deal British Ransom & Marle Bearing the Original 6.MRJA30 drive side And MJ30 Ball timing side they were good enough to WIN Daytona And Pebble Beach In 1961 So they just has good today, So argument over with Yours Anna J
Permalink

Previously anna jeannette Dixon wrote:
Previously anna jeannette Dixon wrote:
Previously simon_ratcliff wrote:

I noticed in a feature of the October edition of Roadholder barrel-shaped Superblend bearings get a mention.

Purely in the interests of technical accuracy please note barrel roller bearings were never fitted to Commandos - and here's why. First off - barrel roller bearings have the rollers assemled to the inner race - the opposite of NJ306E's. Second - with the inner races fitted to the crank and the outer races fitted to the cases it would be impossible to assemble the crank to the cases because the inner and outer race bearing track is curved to match the barrel rollers, so there would be no way the two components could slide together in an axial orientation.

With NJ306E's fitted it is only possible to assemble these because the rollers and inner and outer race bearing tracks are parallel to each other - not curved as in the case of barrel roller bearings.

Call NJ306E's barreled superblends if you want - but you might as well state the sun goes around the earth.....which is flat.

Well hear is a Correction Now the Earth dose not revolve round the Sun like every one is shown , the Sun is moving in a spiral to the centre of our Galaxy and the Earth is in a Spiral motion to the Sun and every thing is in motion in spirels in a magnetic field has that what space is if you had the technology you can bend two points in space time many light year way for each other and create a worm hole and pass through it . and No its Not science fiction but Science fact, they already done the Maths on this one !

Hello now FAG SKF NSK Hoffmann Tinkin .Are All Copies OF the original British made R&M 6.MRJA30 the 650 Norton Manxman for November 1960 was fitted with R&M 3.MRJA30 now the R&M 6.MRJA30 was fitted to the Short Stroke Combat Commando engine and became known has the Superblend bearings yours Anna J

Myth What Myth I all ready told you All how the Superblend Bearing came about and there is NO BARREL SHAPED ROLLERS IN SIDE IT!! it the Roller end have more of a Radius Too them so they do not bind in the hosing at high speed so they do not get has hot, SO THEN THIS MYTH IS NOW OVER !!!! yours Anna J

Permalink

Previously anna jeannette Dixon wrote:

SO THEN THIS MYTH IS NOW OVER !!!! yours Anna J

What about the myth that you cannot install a main bearing unless it comes apart?

Permalink

After carrying out a bit more research I think the clue is in the name,

I'm now of the opinion that Superblend refers to blended rollers and not logarithmic profile rollers as previously stated, as these were apparently only introduced in 1984, see 100 years of evolution of the cylindrical roller bearing.100 YEARS EVOLUTION â of cylindrical roller bearings | Evolution Online

The term barrelled is obviously just a colloquialism for a cylindrical roller that I suppose looks a bit like a barrel in a funny sort of way. When you look at a new FAG NJ306E roller you can defiantly see a different surface finish towards the ends of the roller that has been 'blended'. see Norton Commando Superblend Roller Bearing 06-4118 Drive Side NTN NJ306E C3 | eBay

So that's it; no myth just a couple of questions do you go for C3 or C2 and where can you get these logarithmic profile roller bearings?

So what's your next none confrontational post Simon? Straight oils verses multigrades perhaps??smiley

Attachments roller-profile-jpg-jpg
Permalink

Previously Jonathan Soons wrote:

Previously anna jeannette Dixon wrote:

SO THEN THIS MYTH IS NOW OVER !!!! yours Anna J

What about the myth that you cannot install a main bearing unless it comes apart?

well you can if you know what your doing, ! and I have , many times ! and a ball race bearing does not come apart , I fit the drive side bearing in the crankcase first after its been in the oven for half hour bearing fitted with loctite bearing seal i put the crank in the freezer for a hour or so then the crank fits nicely into roller bearing with a 10thou shim in brass held in place by a blob of grease then on the drive side you next fit the camshaft and the breather disc and spring with a good squirt of oil sae 50 and oil the cam lobs too and next up the timing side with ball race bearing in place with oil seal washer and per warmed cases .when fitting gudgeon pins pre warmed in old cooking oil but filler old oil first you do not want any bits of fish&chips in there. . now for cleaning cylinder heads with on valves fitted boil slowly in rhubarb duce until clean Rhubarb is mildly Acidic and will clean alloys nicely but do not let you wife catch you ! , I am OK has a live on my own no one to tell me want too do !! Anna J
Permalink

Previously Jim Brierley wrote:

The term barrelled is obviously just a colloquialism for a cylindrical roller that I suppose looks a bit like a barrel in a funny sort of way. When you look at a new FAG NJ306E roller you can defiantly see a different surface finish towards the ends of the roller that has been 'blended'. see Norton Commando Superblend Roller Bearing 06-4118 Drive Side NTN NJ306E C3 | eBay

So that's it; no myth just a couple of questions do you go for C3 or C2 and where can you get these logarithmic profile roller bearings?

So what's your next none confrontational post Simon? Straight oils verses multigrades perhaps??smiley

Straight oils Jim? Are you sure. Every one knows oil comes in a 3 dimensional container so I don't see how it can be called 'straight' as you say.

You'll be saying the Moon is made of cheese next! Cheddar or Wensleydale?

P.S Don't agree about the NJ306's looking like a barrel - not even in a serious sort of way.

Permalink

Previously simon_ratcliff wrote:

P.S Don't agree about the NJ306's looking like a barrel - not even in a serious sort of way.

It depends if you've got your beer goggles on on not!!

By the way I wouldn't fancy going the Anna's for a fish & chip supper followed by an rhubarb crumble, would you? You might lose a few gold crowns chomping on a Norton big endangry

Keep Smiling

Jim

Permalink

Could it be that what we have is a crowned roller bearing. This allows for slight misalignment as it avoids the risk of the end of the roller digging in as is the case with a conventional cylindrical roller, Crowned rollers also have an advantage in that they produce an elliptical contact as opposed to the line contact of a conventional cylindrical roller.

Given that the crankshaft production methods, particular at Wolverhampton were pretty diabolical, it was no surprise that Norton needed what we call superblend bearings.

Permalink

Previously charles_bovington wrote:

Could it be that what we have is a crowned roller bearing. This allows for slight misalignment as it avoids the risk of the end of the roller digging in as is the case with a conventional cylindrical roller, Crowned rollers also have an advantage in that they produce an elliptical contact as opposed to the line contact of a conventional cylindrical roller.

Given that the crankshaft production methods, particular at Wolverhampton were pretty diabolical, it was no surprise that Norton needed what we call superblend bearings.

Charles, what you refer to as a crowned roller bearing which copes with misalignment is what FAG call a barrel roller bearing - which is not cylindrical - unlike the NJ306 rollers with radius/blended edges and are not recomended to be used with misaligned shafts.

Any way, I''m trying to start a cheese thread.

Permalink

To clarify the question of superblend bearings, Roy Bacon in his book on Norton Twins published in 1981 and therefore quite near to the time of Commando production states, and I paraphrase here,that the effect of running the old rather flexible crankshaft in the new stiffer crankcase was was for the corners of the rollers to try, and succeed, in digging in to the track on which they ran.This + problems with jamming of the auto advance resulted in very short bearing life, 4000 miles. What was needed was a self-aligning spherical roller type but none was available which could stand the speed and load. The solution was the so called superblend bearing which had slightly barrel shaped ends but was straight in the middle.

This problem surfaced in 1972.

Permalink

Previously charles_bovington wrote:

To clarify the question of superblend bearings, Roy Bacon in his book on Norton Twins published in 1981 and therefore quite near to the time of Commando production states, and I paraphrase here,that the effect of running the old rather flexible crankshaft in the new stiffer crankcase was was for the corners of the rollers to try, and succeed, in digging in to the track on which they ran.This + problems with jamming of the auto advance resulted in very short bearing life, 4000 miles. What was needed was a self-aligning spherical roller type but none was available which could stand the speed and load. The solution was the so called superblend bearing which had slightly barrel shaped ends but was straight in the middle.

This problem surfaced in 1972.

Hello and yes is wassolved By Ransom &Marles new type of roller bearing Number 6.MRJA30. and Not FAG Bearings they just copied what Ransom & Marleshave all ready done, but FAG Bearing are Made In China and Sent to Germany for inspection they are tested to the highestSpecifications, before Marketing . But there still Not has good Has the Real Deal Ransom & Marles own British made Bearings I have a nicecollection of alldifferent makers of Bearings for R&M NTL , Hoffman Tinkin SKL PHB FAGPhoenixes Bearing and many more So you can take it from me being a Marine engineer we seen them all and Ransom & MarlesAre one of the leading Bearing manufactures in the world So keep it British andbritishjobs Yours Anna J Dixon
Permalink

Previously charles_bovington wrote:

..........and I paraphrase here,that the effect of running the old rather flexible crankshaft in the new stiffer crankcase was was for the corners of the rollers to try, and succeed, in digging in to the track on which they ran.This + problems with jamming of the auto advance resulted in very short bearing life, 4000 miles. What was needed was a self-aligning spherical roller type but none was available which could stand the speed and load. The solution was the so called superblend bearing which had slightly barrel shaped ends but was straight in the middle.

This problem surfaced in 1972.

The shape at the ends of the cylindrical rollers is a radius, later generated by a logarithmic curve. I don't see any connection with a barrel shape.

For more info regarding the 1972 problems

www.andover-norton.co.uk/Combat Sl.htmThe URL above is incorrect. It should be www.andover-norton.co.uk/SI Combat.htm

Webmaster

Permalink

Simon,

I may have missed something here. The link which you refer to seems to be non available. The explanation on the problem referred to in the section on history of the commando refers to breather problems causing loss of oil and hence bang. this seems to be at variance with every other source and ones recollection of the times. The exact form of the corners is rather esoteric. The bearing is clearly not a crowned roller bur a parallel roller with radiused ends.

Permalink

Previously charles_bovington wrote:

The bearing is clearly not a crowned roller bur a parallel roller with radiused ends.

Hi Charles

If we all use the same terminology we can at last put this post and this myth finally to bed. The so called 'superblend' rollers are not crowned, barelled, spherical, radiused, logarithmic or hyperbolic, they are blended at the ends, hence the term superblends.

In reality you would be hard pushed to measure the blended ends with a micrometer.

Permalink

i have learnt you cant argue with the facts if presented in a civil manner of course. i have been wrong on many occasions and i have admitted so. its all a part of lifes learning curve. Baz

 


Norton Owners Club Website by White-Hot Design