Why ball bearing mains?

Status
Not open for further replies.
I was beginning to think my Laverda was the bees :oops: Its a bit mind blowing to think that block of Iron can spin that fast. :!:
 
That's an impressive locomotive like crank shaft Dances. Balls do moosh out of round so even if they allow more flex w/o transmitting it to cases they may fail just from pure loads sooner than a roller. I ain't no racer but know about rpm failures. I'm scared of hi rpm no matter the bearings so Peel mostly aimed at hi torque not dangerous rpms as she's got to last like 20 yr to get something out of her investment.

One point maybe over looked in what bearing to use is what cases its in. Past dragsters only had thin-ish factory cases not beefy Maney ones.

What do you press your super crank roller racer rpms too and what do ya think might be Peels 920 danger zone to avoid? The blower boost goes up non linearly so there's temptation to try 8000 like Tom Drouin did for his sales ads. I reflect that TC tried to shift at 8000 he told me but sometimes shot to 8500-ish yet is super famous for 160 runs w/o opening engines on essentially factory parts. Don't yet known his bearings but will find out along with reason and issues of his 16" rear 850 cruiser.

Roller bearings Also used where minimal side loading is present. They can carry more load than ball bearings but do not tolerate misalignment as well. A roller bearing on the nose of a quick-change ring and pinion is a popular application.

Read more: http://www.circletrack.com/drivetrainte ... z2ICtoMSxt
Click to expand...
 
john robert bould said:
I was beginning to think my Laverda was the bees :oops: Its a bit mind blowing to think that block of Iron can spin that fast. :!:
Click to expand...

From a rotational mass standpoint, I would not be surprised if this is comparable to your Laverda. What is the stroke on the Laverda.

From what I can see of this billet design and from what I have seen done by others on Norton crankshafts, this lump could probably have been a couple of pounds lighter. Quite frankly I am not sure I would want to go there as shock loading to the main bearings and cases increases with a lighter crankshaft and the bike has a nice enough drive out of the corners as it is now. There are plenty of more fruitful things to go after for improved performance.
 
[

When i first saw this crank,i thought of the Norse God of thunder ThOR ,this must be the HAMMER head used to destroy all,,I am proberly Right!
It's appearance alone chills the soul,forged in the very depths of Hell, by satin himself, and ground by Demons. sorry went a bit Hobit there! :lol:

quote="Dances with Shrapnel"]I went with the 35mm main journals on my 75mm short stroke for just the reasons Ken mentioned above (increasing journal overlap). This is nicely outlined by Charles Fayette Taylor in Volume II of "The Internal Combustion Engine in Theory and Practice", a must read for serious engine builders who want to know.
Below is a picture of the subject crankshaft:
Why ball bearing mains?

It has been through several abusive race seasons (is there any other type?) and I continue to have the piece of mind that it will not break.....which allowed me to move on and focus on other component failures. :lol:

In hind sight I probably could have secured adequate race duty durability with 30mm main journals due to 1.) superior metal quality and 2.) overlap of rod journal to main journal as mentioned by Ken above. We run roller bearings exclusively (as ball bearings are not up to the task). We never had a bearing failure.[/quote]
 
Re ball v needle rollers.

There maybe an explanation on why they prefer to use one needle and one ball, I have fitter 2 needle rollers into an 350LC engine like Yamaha did with their 350 TZ racers, where there were 2 rollers before, in the 350LC, with using 2 needle rollers you have to machine slots into the bottom of the crankcases to fit retaining collars otherwise the crank will float sideways where it did not with 2 ball bearings.
Whether that is a sufficient explanation, I don’t know, as all engines are different.
 
Ball or roller? We've examples of raced Nortons that that weren't shimmed tightly so apparently there is nothing pulling the Norton crank hard to one side, [like the cam does], so apparently the rods on journals keep crank centered to bores. So ball or roller don't matter that way in Nortons so mainly crank flex and shock loads tolerance is what matters most.

We know that some heavy twin Norton cranks/cases can tolerate 9000 rpm zone, what bearings combo do they use? I bet there are example of both ball and rollers tolerating this, so which can longer?
 
The only reason I persevered with my 63mm stroke 500cc Triumph (triton) was that it was always reliable in the bottom end. Valves were standard 650 size and the rocker gear was lightened, so it would cop 10,500 rpm without destroying itself. The bike was a nasty piece of garbage , to ride it was all t op end, and if it dropped off t he cam in a corner, and I slipped the clutch, it was immediately sideways. With a two into one exhaust or straight pipes it was bearable, but with megaphones it was a bastard. It really needed a six speed box, instead of the old 4 speed CR triumph one, and it needed a Weslake four valve hea d . If it had those two things, I believe it would have been faster than the Paton 500.
Two things will stop you doing that successfully to a Norton - the valve gear, and the lack of versatility in setting the cams independently. A more difficult motor to develop.

When I ride my 89mm stroke Seeley Commando 850, it inspires confidence - NOT ANXIETY !

Why ball bearing mains?
 
You don't like your Triumph, and you want to develop your Commando the same way ???

Plenty of good performance Commando cams, off the shelf.
Its been done before, why re-invent the wheel ?
 
Now , this is a Skip Rope . :p :lol:

Why ball bearing mains?


Visualise the bottom as the Right hand end of the crank , and the helix as the inputs . Wont mention the flywheel :? Er . . .

Consider maximum forces / loadings as crankpins & flywheel , the helix is a rough approximation of the force vectors in
motion. perhaps , presumeably . so its a two plane distortion per ' skip rope ' .

Quote on Bearings is from Dave Minton . One of the less exciteable jouranlists .Though he had a few bothers with Nortons
at one stage ( Motorcycle Sport ) they persevered & he found compared favourably to the BMWs for Continental Touring .
His long term involvement with the saga of the marque gave him a considerd perspective of the visisatudes prevaling .
(lots of big words , should hold Rohan for a time :) :twisted: )

Presumeably the ' looser tolerance ' is say a C3 . I think a comparison with Dunstalls ' deep groove ' high capacity ball races
in that lack of consideration of flexure , tying in the ends of the crank ridgidly , causes ' something ' to give , or break . . .
 
As any wommble can see , if a bearing is intolerant of deflecton , concentrations of load are formed IF the shaft is not maintaining strightness .
In the bowing of the shat the loadings are transverse on the ball race . Out at lower , In at upper .Tight tolerance bearing binds & crank bowing isnt uniform as is the case with radial deflection tolerant ( more with open tolerance / loose clearance / OR Barreled high load capacity roller ) .

What occurs is loadings are transfered over lesser area so concentations are increased .
Or ,
As in the case of shaft deflection exceeding bearing tolerances , the deflection in the shaft is no longer uniform. The area between the alignment created at the ball race supporeted area of the shaft and the bowing causes high loadings at the area of suport , in the shaft .
These loadings exceed those formed with the axial alignment tolerant type .In the Shaft while the bearing is serviceable .
Obviously stess concentrations reduce fatiuge life .

BANG . the shaft goes /if the bearing doesnt .

Visualise a broomstick sitting on a couple of half rounds ( flat side up ) ea side of the back steps , on Conc. nib walls .

Sit in the middle the broomstick curves , the ends outside the wall being unloaded stay stright .

Now , if the ends are fixed through the Nib Wall , or ballasted so fixed . The broomstick bends cowhorn style when you sit on it .
Unless its a weak broomstick , When all the loads hit the fixed bit at the ends , it KINKS . and snaps. And you fall down, bounce down the steps ,
just as the postie comes through on his CB 110 , and you get tyre treads on the T Shirt .


Just like the flywheel at max. rpms if the shaft breaks . :lol: . is that perfectly clear ?

Why ball bearing mains?

bizarre deflection , support , stress concentration stuff .
http://www.sciencedirect.com/science/ar ... 0X06006742

Analogy . Ship = crankcase .

Why ball bearing mains?

http://www.sciencedirect.com/science/ar ... 3905000468

Something a bit less abstact .

Why ball bearing mains?


but nevertheless an oversimlification of the dimensional stability of a Norton bottom end at about its maximum safe operating conditions .
 
All interesting stuff matt but comparing a low speed marine diesel that uses babbit bearings and heavy stiff crankcase webbing with a Norton twin that uses a light weight aluminum crankcase with roller element mains and significant crankshaft span is out of context. You do not even see throw weights (bob weights) on your illustrations or the picture below.

And if I read what you wrote correctly, making a simple unconstrained beam comparison would be fair but suggesting there's any significant contraint to bending imparted by the Norton aluminum crankcases is in my opinion, a bit off.

Really completely different animals.

Why ball bearing mains?
 
Dances with Shrapnel said:
Why would anybody want to use anything but roller (Superblend) style bearings for the main bearings in a Norton big twin.

At the risk of starting an opinion war, I would like to hear some of the pro's and con's of the choices. I see the primary options are 1.) Superblend for both sides, 2.) ball bearings for both sides or 3.) a Superblend type for the drive side and a ball type for the timing side. I guess I am trying to flush out the rationale for any given choice which can range from keeping it original, this is the way we always did it, serviceability, absolute need or ????????????

Keep in mind that the Superblend style solution was either not known, not needed and/or not available in the early years of the Norton big twins.
Click to expand...

This is really simple, I wrote a technical article in the Norton Notice back in 1996 or so. You Can search the Norton Notice and find it. In a nutshell, if you wind it up, the crank will flex a LOT more than what will be tolerated by the Superblend bearings. If you have a Mk3, you will break the crank cases, particularly if your crank is less than perfect in the fit of the those six little rubber band bolts holding the whole shebang together. If those are not a drive fit, keep your revs under 6,000. Better yet ream the crank so you have tight bolts. The racers pretty much like a Superblend on the drive side, the roller does carry more load, and a Max Ball on the timing side. That's the bearing with an extra ball in it. If you have your crank out and you're serious, have someone put as much Mallory metal (tungsten) as possible in the cheek pieces and remove metal from the central flywheel. There is WAY to much weight in the middle. The more you can move to the cheeks, the better. While you're at it, check the balance factor, you might even be able to remove more from the cast flywheel. If the flywheel grenades, the damage is impressive...you might save the head, but even that is doubtful. Nobody but a bean counter would actually choose to make a flywheel like the Norton uses out of cast iron. Any materials engineer would shudder and whisper "steel is much better".

The race guys also like the older cases as they allow more crank flex without breaking. Of course, flex on the drive side has the alternator rotor hitting the stator, flex on the timing side isn't good for the oil pump drive etc. It's best to know how well fitted your crank bolt are and use a 6,000 RPM red line. If you want to rev faster than that you need a steel flywheel and probably ought to put in longer rods and do a 90 degree crank. THAT get's rid of a LOT of weight in the center.
Stevan Thomas
World's Straightest Commando
http://www.vintagenet.us/phantom/wsc.html
 
nortonmargarita said:
This is really simple, I wrote a technical article in the Norton Notice back in 1996 or so.
Click to expand...

I finally found that article back in 1996 or so. Interesting historical perspective on the fellows opinions. This article had already been summarized earlier in this thread and it was nicely stated by someone else earlier in this thread that other people have different experiences. There's plenty of inconsistencies on what works and why and what should be done.

Bearing catalog information lists roller bearings having about 1/3 the tolerance for misalignment as that of comparable ball bearings (4' versus 12'). At the center of the Norton crankshaft this roughly equates to between 0.003" and 0.010" respectively of total deflection; not a lot. It would be interesting to get a crankshaft into V-blocks and measure/chart load versus deflection; not necessarily to destruction but just to get a sense of load versus deflection. We know out of balance forces are in the tons and I suspect these loads will result in deflections greater than 0.010". May do this myself when I get to the shop and find a load cell.

Clearly there are some trade offs as the roller bearings have roughly 50% greater load bearing capacity than comparable ball bearings. What we do not know is the difference in bearing behaviour beyond the recommended limits. As an example, catalog information for roller bearings recommends consulting with bearing manufactures when greatly exceeding allowable angular misalignment whereas single row deep groove ball bearings seem to have a hard limit for angular misalignment.
 
Bearing catalog information lists roller bearings having about 1/3 the tolerance for misalignment as that of comparable ball bearings (4' versus 12'). At the center of the Norton crankshaft this roughly equates to between 0.003" and 0.010" respectively of total deflection; not a lot. It would be interesting to get a crankshaft into V-blocks and measure/chart load versus deflection; not necessarily to destruction but just to get a sense of load versus deflection. We know out of balance forces are in the tons and I suspect these loads will result in deflections greater than 0.010". May do this myself when I get to the shop and find a load cell.
Click to expand...

Well there ya go Dances, this is how I understand ball vs roller in our cranks, though not to the N'th degree you filled in. For rollers to take the same length of shaft flex loads they 'have to be a lot stronger than balls' as they bind on crank and case bores trying to resist its flex, which we Know noting can stop it flexing in red zone. Oak tree vs Palm tree in hurricane though either can be blown away at some flexing amount and force. Too bad Norton ain't around to do some destructive testing to find the facts for us. Anywho I've got two spare bent cheeks from past late Peel + maybe a cast iron flywheel and a fella with dozer shop with a 30, 000 lb press we used on a swing arm spindle removal, so could do your thought experiment for real to full destruction. I think the factory cranks flex in center to over .010" while still in the 7000's rpm.
I saw figures in NOC posts that piston jerk down applies 3600 lb at 7000 rpm. My smart ass brains say build it to take 8000 and never run it to that point, so ball or roller works as well till something else lets go for dramatic endings.
 
Status
Not open for further replies.
Back
Top