Why ball bearing mains?

I would not try o stop the crankshaft from flexing by using roller bearings. I believe the superblends allow the shaft to flex instead of being fatigued by being held rigidly at the ends, and stressing the corners of the journals. I don't believe ball races stay tight for long anyway, but I would never use two normal roller bearings on a crank. On Triumphs we always used a ball race on the timing side, and a roller bearing on the drive side and I believe even that was too rigid. These days I'd be using superblends if I had a Triumph twin racer. One of the reasons that I didn't race my Seeley for so long was that I think what is happening down there when the motor is revving hard is simply horrific. If I was really serious, I'd buy or make a billet crank, and try to move the weight towards the ends of it.

acotrel said:

I would not try o stop the crankshaft from flexing by using roller bearings. I believe the superblends allow the shaft to flex instead of being fatigued by being held rigidly at the ends, and stressing the corners of the journals. I don't believe ball races stay tight for long anyway, but I would never use two normal roller bearings on a crank. On Triumphs we always used a ball race on the timing side, and a roller bearing on the drive side and I believe even that was too rigid. These days I'd be using superblends if I had a Triumph twin racer. One of the reasons that I didn't race my Seeley for so long was that I think what is happening down there when the motor is revving hard is simply horrific. If I was really serious, I'd buy or make a billet crank, and try to move the weight towards the ends of it.

Click to expand...

Even the ball bearings do not constrain the crankshaft from flexing; maybe a double or triple row of ball bearings on each side in a heavy (I mean massively heavy and stiff) steel or cast housing but of course that is not what a Commando is. My point is if you really think there's any significant "constraint" in flexing going on you are kidding yourself; if the bearings do not accomodate it the engine cases will ....up to a point.

Billet cranks may or may not reduce the flexing and is primarily dependent upon design; the benefit is that billet has superior durability when compared to the cast steel used in Commandos. Transferring some mass from the center bob weights to the outer bob weights has some advantages but it can be overdone and you introduce a worse condition.

Roller bearings are not intended to 'stop crankshaft flex', and never were. Neither were ball bearings.
The crank has to be inherently strong enough to stop itself flexing. As best you can get, anyway.

Roller bearings simply have a higher load carrying capacity than ball bearings.
More surface area = less lb/sq in loading in the bearing.
More rollers in the so-called superblends than the earlier cylindrical roller bearing Nortons used.

My old antique early triumph has a double row ball bearing on the drive side - of the self aligning type.
So if the incredibly heavy crank ( for a slow revving 500cc single) does flex, it doesn't destroy the bearing..
It will slog down to amazingly low revs, (well, it will if I ever get it together), big torque loadings at those low revs.

lcrken said:

dave M said:

I use an old Superblend inner race for setting the clearance. The old race has been relieved on the inside with a Dremmel to give a sliding fit on the crank shaft. I can therefore try different shims until I am satisfied that I have the clearance correct before pressing the new inner race onto the crank. This saves having to either heat the cases to remove the bearing outer or else use a puller on the inner. Bearing manufacturing tolerances are such that any inner in any outer will be close enough for measuring end float. I have used this method many times and the final clearance is always the same as my test clearance with the dummy inner race.

Click to expand...

That's a clever idea, Dave, and I'm a little embarrassed that it never occurred to me. In my case, I use shims behind the outer race, but the same principle should apply if I just take a little off the O.D. of a used outer race. That would save me from repeatedly heating the case to remove and reinstall the race. Thanks for the hint.

Ken

Click to expand...

Thats actually in the instruction manual (somewhere ?) of how to set up superblends.

Since fitting and removing them is likely to damage/distort/destroy them in the process, they are a one use only affair...

BALLS , in the notorious Bonneville , it ran Ball Races both sides . Theory being less drag in ball races compared to rollers . thus freeing HORSEPOWER .

Was told it was a 1972 Crank , std sizes . But was the light one from around 68 or had that flywheel form .ANYWAY , the less drag in the Brg.s allows it
to accelerate ( the crank pick up revs ) quicker . FRACTIONALLY . V low %ages . ( but they all add up ).

Norton Crank , in a 750 the smoother action of the superblends is liable to create a more constant motion in action of the crank . as the loadings /
dimensional stability , are more constant .
As In not exceeding the design / operateing criteria , of the bearing ( at race speeds / rpms ) where the radial loads are tightening & freeing Ea revolution
due to ( additional ) side loads in the tracks .

Thus somebody building a Cowboy 650 SS might get a bit freer operation pouring the sauce , but the ( radial deflection ) loadings become critical in a
worked 750 .Probably increased with the stiffer cases .

I consider a lateral osscilation at the flywheel counterweight contributary to failure. The notorious Crank Flex NOT neccesarilly symetrical ( or about the Axis )
but literally torsional deflection ( as well as beam deflection ) . As the Flywheel is the most removed from the support ( bearings ) and has the greatest moment ( distance to suport ) AT the COUNTERWEIGHT , it stands to reason that the conterweight point suffers a lateral reaction to the fireing impules . :shock:

then It'd get all peculiar , with natural frequencies of materials etc , so if impulses overlap , we get cumulative loadings . As in Multiplyed . :?

WHO is going to get a dungie old scrap Commando crank , support it at the main bearings , load the flywheel hydraulically at TDC untill the crank FAILS
and tell us what force it took . STATIC . 8) :?

be good to have a few ( disposable :? ) DTIs on it & force / deflection recorded . And a good high speed movie camera .

Matt Spencer said:

Norton Crank , in a 750 the smoother action of the superblends is liable to create a more constant motion in action of the crank . as the loadings /
dimensional stability , are more constant .

Click to expand...

Explain why a roller bearing would have 'a smoother action' than a ball bearing, if both have the same number of rolling elements. ?

If the heavier load capacity of the cylindrical roller bearings weren't required to ensure a suitable bearing life, designers would leap at the opportunity to use ball bearings. The cheaper the better, probably, if load capacity wasn't a constraint.

Engines back in the teens and 1920s, and earlier, often just used a pair of ball bearings on the crankshaft, and since engine outputs were low (as was lubrication), they survived just fine.

Vintage Motorcycle Club .

http://www.youtube.com/watch?v=nexai2NkM9s

Matt Spencer said:

WHO is going to get a dungie old scrap Commando crank , support it at the main bearings , load the flywheel hydraulically at TDC untill the crank FAILS
and tell us what force it took . STATIC . 8)

Click to expand...

This is a typical misconception about Norton crankshaft component failure. The test would be great if you were building a bridge or designing a beam with your Norton crankshaft :roll: but the failure mode of a Norton crank is a matter of durability .................................except for those extremely odd instances where the throttle pins open and the engine spins to infinity.

By dungie old , I mean one thats mutilated or past it . Unless someone wants to sacrifice a new one .
other option is calculate the figures . generally needs ' proofing ' in iny case .

Finding the failure load acoss the crank , say (load 0 at timing side crank pin , would get some figures to work with , regarding strength .
Most failures appear to be through Left Journal or web. which is NOT the area of least strength ( there other side is not differant ) .
Therfore the deduction is that its the area of maximum loading . which is fully obvious , both power impulses go through there , to the output shaft
/ engine sprocket . the timing side web does little more than maintain alignment ( we hope ).
Would be intresting to see the force required to produce the reputed 4 thou runout / deflection at the Flywheel .And the ultimate force required to produce failure .
If this is greatly in excess of the running loads , Cumulative forcuceing the failure , perhaps also misalignment .

Will start a thread ' Crank failures ' to try & purloin pictures of such . Break would roduce evidance of stress concentrations / load paths .

Where do you anticipate these huge forces coming from, in a running engine, to break the crank in half ??
A 100 ton hydraulic press , or even a 10 ton press, doesn't actually emulate anything in the real world ??

other option is calculate the figures [/unquote]

So someone has calculated the crank will break after 132 trillion revolutions at 3/4 throttle, or 132 billion revolutions at 90% throttle.
What now ??

Click to expand...

The 750 factory set up strains crank center with ~3600 lb @ 7000 rpm.

Btw you'all do realize that there is no such thing as a super blend bearing other than Norton best sales pitch to public since Fountain of Youth found in Florida myth to attract tourists. That is there was never made nor ever sold a flat roller element bearing for engine cranks that were not already barrel shaped to similar degree as mythological 'superblends". The bearing issues arose mainly on Combat and it turned out to have little to do with bearings, and about everything to do with initial bad design AAU for 4000 blow ups, then 8000 mile blows up till case bores done better, then original rollers last just fine. I had the factory set examined from 35,000 miles showing on tach pre-Peel and found almost pristine. Same with my second total factory bad stuff Combat, roller still good to go and go but replaced just in case.

except for those extremely odd instances where the throttle pins open and the engine spins to infinity.

Click to expand...

Shame on you Dances flashing me back to a horrific loss of my beloved Peels power band, but can tell ya with the DS roller/TS 11 ball set up, the crank and the things on crank end fail before the bearing do and my buddy Wes has used these over tested Peel hand me down bearings for 4 yrs & two ring jobs worth of riding so far, no problemo. But then again Peels bearings were cryo tempered and dry friction coated. Crank bent 1/16" out on DS and 1/32" on TS but after 2000 more miles before final tear down DS self corrected to 1/32" and TS t0n ~1/64". I ran 20/50 Mobil 1 back then.

If you choose to use the standard balance factor of 58%, perhaps you should use your motor in the rev range, for which it was designed ? If it is used at high revs with that factor, every revolution must flex the crank and fatigue it. If you feel excessive vibration at high revs....... I had Triumph twins for many years, and always used factors up around 78%, in motors which ran to 8,000 rpm. The stroke is only 82mm, not 89mm. I suggest using the correct balance factor is even more important in a commando engine, if you are going to rev the tits off it.

This forum thread backs up Schimdt's and me witness marks assessment to try upper 70's wet BF as that shifts the main orbital sling more horizontal as we have seen the cases tend to break in half d/t the vertical crank sling orbital component with lower BF. This goes against the mechanical logic in the BSA research article that found wet 60's BF gives least load spikes or out of round forces in 360' twins.

I've not been back to my 'puter with the article to post here that detailed the manufacture faults that created the Combat Bomb issues that reveal ball or roller bearing had little if nothing to do with the failures and superblends are just a sales pitch. In olden days of more flexy cases the balls allowed some extra crank flex to pivot on their point contacts before case strength tested. As case strength increased then barrel rollers could be used to resist cranks flex, as the rollers don't allow as much tipping action before its transmitted through the races into the case bores.

Ideally cases and crank and bearings should totally resist flex, at some cost of mass involved, which brings up question of what Sir Eddie came up with in his short stroke 500 11,000 rpm land speeder.

acotrel said:

If you choose to use the standard balance factor of 58%, perhaps you should use your motor in the rev range, for which it was designed ? If it is used at high revs with that factor, every revolution must flex the crank and fatigue it. If you feel excessive vibration at high revs....... I had Triumph twins for many years, and always used factors up around 78%, in motors which ran to 8,000 rpm. The stroke is only 82mm, not 89mm. I suggest using the correct balance factor is even more important in a commando engine, if you are going to rev the tits off it.

Click to expand...

Re; the engine balance factor, is the percentage taken with the crank full of engine oil or is the crank dry :?:

Bernhard said:

acotrel said:

If you choose to use the standard balance factor of 58%, perhaps you should use your motor in the rev range, for which it was designed ? If it is used at high revs with that factor, every revolution must flex the crank and fatigue it. If you feel excessive vibration at high revs....... I had Triumph twins for many years, and always used factors up around 78%, in motors which ran to 8,000 rpm. The stroke is only 82mm, not 89mm. I suggest using the correct balance factor is even more important in a commando engine, if you are going to rev the tits off it.

Click to expand...

Re; the engine balance factor, is the percentage taken with the crank full of engine oil or is the crank dry :?:

Click to expand...

53% is the factory recommened BF for the Norton Commando and that is quoted as dry BF. This dry versus wet is specific to each and every crankshaft so a dry BF for a billet crank will not necessarily translate directly to a dry BF for a stock crank. This is due to the considerable differences in oil cavity volume in the rod journals.

The 53% BF is near ideal from an engine stress and strain perspective for the Commandos and was made practical once they stepped away from the solid mounts and started using the isolastic system. Quotes and mentions of higher BF use is an artifact from the solid mount days and is appropriate if you are not using an ISO system. The actual BF to use for the solid mounts is a compromise and specific to each bike.

hobot said:

This forum thread backs up Schimdt's and me witness marks assessment to try upper 70's wet BF as that shifts the main orbital sling more horizontal as we have seen the cases tend to break in half d/t the vertical crank sling orbital component with lower BF. This goes against the mechanical logic in the BSA research article that found wet 60's BF gives least load spikes or out of round forces in 360' twins.

Click to expand...

The only way these crop circles would possibly be useful is if you did a before measurement and wanted to replicate the movement because you were happy with it. You will find that each bike/engine configuration will have a slightly different "happy spot".

hobot said:

In olden days of more flexy cases the balls allowed some extra crank flex to pivot on their point contacts before case strength tested. As case strength increased then barrel rollers could be used to resist cranks flex, as the rollers don't allow as much tipping action before its transmitted through the races into the case bores.

Click to expand...

Your statement above about "barrel rollers could be used to resist cranks flex" alomost looks like a fabrication of yours. There's really no resisting going on as the forces at speed will overcome just about anything the Norton cases can dish out. Again, there really is virtually no constraining the flex, only accomodating it in the bearings and cases.

Dances with Shrapnel said:

53% is the factory recommened BF for the Norton Commando and that is quoted as dry BF.

Click to expand...

Recorded Commando factory BF is 52% WET.

nortonspeed said:

Dances with Shrapnel said:

53% is the factory recommened BF for the Norton Commando and that is quoted as dry BF.

Click to expand...

Recorded Commando factory BF is 52% WET.

Click to expand...

I was referring to factory recommended whereas you cite the factory recorded - big difference.

Thanks for straightening me out as well as the record. Yes 52% dry. Been so many numbers thrown around lately I am loosing track.

All listed BF are dry state. Factory size oil mass wet raises BF about 9-10%

Yes Dancer each chassis and pilot and purpose would each have a more optimal BF so Jim and I are just speculating with assumption that a bit more BF for a bit more horizontal component for a bit more circular motion may average down the vibration better especially in the upper rpm. Ms Peel is one of the few equipped for fairly easy BF changes. Past Combat Peel flat disappeared to pilot sensation around 2000 but heavier 920+ Drouin 15 lb is an unknown but going by the 920 racers with JSM piston kit reporting in delight definitely smoother Peel may still disappear better than flying carpet with the dang fringe flap annoyance, closer to saucer ship but retaining the sense of acceleration rush.

As to crank flex tolerance in barrel rollers, Yes compared to model T flat roller axle bearings, barrels allow more crank flex w/o binding by allowing some shift of load contact along lenght of shaft on rollers, but No barrels tolerate less flex as ball bearings can.

Dilemma ? * is it better to allow crank to flex by thin case sides and ball bearings or prevent it from flexing by thicker case and longer bearings? Keeping things in line obviously better so Peel's philosophy is go for the Torque with stiffer heavier cases and barrel roller and hold the rpms down to last a long enough to get some return on my wasted time+money. The oil pump snout and the cam tensioner are at risk if TS crank end whips about much so barrels on that side too. If I'd put the blower on 750 Peel as originally intended I fit with JMS kit and BSA lifters and stuck with the Combat flexy cases and 11 ball on TS and shoot for 9000 rpm as often as I could.

With my Seeley, the choice of balance factor was easy. It is intended for racing with revs between 3,000 and 7.500. It literally shakes at 3,000 and is extremely smooth at 7,000. If you rev a 53% BF standard commando crank at 7,000, I suggest you are looking for trouble. The crank would be well out away from ideal balance and every part of the centre of it would be stretching. The loads due to out of balance at high revs must be much higher than if it occurs at low revs - simple laws of physics. I think if I had a road bike, I would use a BF at about 65 to 70% and tolerate the shake, when I was riding in slow traffic.. The Norton Atlas would have been around that figure.