From mc-engine list, 4/12/2012 8:37 AM: to put in our pipes to smoke on...
I've been a spectator awaiting someone to state the perhaps real story behind superblend bearing usage and the probable fact that C3 clearance bearings are only specified to enable a good interference fit of the bearing outer race within the cases to ensure they are still a reasonable fit when the engine is up to opertating temperature.
Sure, barrel shaped roller bearings are supposed to be better because they anable a certain amount of crank flex to occur without placing out of spec loads in the nearings and mainshafts. But I seriously doubt that is the main purpose.
Just look at the super-rigid bearings and pairs of bearings on mainshafts that have been used over the years, some of those cranks dd not event pretend to be allowed to flex like a butterfly!
Yes, it was claimed that the Superblend bearings solved the Norton crank breakage problem. But an article I read recently in an old (Classic Bike?) magazine (still looking for it to 100% confirm my memory....in which I am 99% confident anyhow) that was written by one of the factory test riders stated that Norton had intermittent crank failure problems with their road test bikes.... they changed to superblends but, at the same time, more or less, someone discovered the real cause.... it was the radius on the crank step. It was found that the crank breakages were cyclic...they would be fine for about a week then several would come along...then they'd be fine for another week...and so on.
It was discovered it was the same timing as the cycle that the crank grinder re-dressed his grinding wheel. He re-deressed the wheel about weekly and not put a suitable radius on the wheel edge. Over the following days the wheel would get a reasonable radius on their edges....until it cam time to re-dress the wheel....and that was found just after the superblend bearing was introduced.
The first weeks that the bearing was fitted the crank problems continued....but soon after the grinding wheel radius issue was addressed and all the crank problems stopped.
Superblend bearings were declared the solution.
That "conclusion" was either out of ignorance of the facts, or it was decided by management that it was a palatable solution to some problem that was perhaps not deemed to be an outright mistake by the factory?
Whatever, the saviour superblend was suggested to be superfluous ...... but it's legend lives on.
Without checking I think that according to manufacturer's specs, a ball bearing on a rotating shaft with a stationary outer race needs to have no interference between shaft and inner race...it is supposed to be a sliding fit(?)
Therefore all a bearing's inner clearance that reulsts after installation is: Initial bearinginternal clearance - a percentage of the interference fit of the outer race to the housing.
Of course, the alloy crankcase outer housing has a higher coefficient of expansion than the bearing steel. Therefore as the temp of the assembly rises, the interference fit decreases....and it may get to a point where the interference fit is not sufficient to hold the bearing within its housing and slippage occurs.
As far as I am able to ascertain, the use of C3 bearing for crank mainshafts, is to allow a good interference fit of the outer race in the cases to ensure a reasonable fit after the whole assembly reachers operating temperature.
However, I've measured quite a large difference in interference fits used by manufacturers for the outer race when C3 bearings are specified.
The differences have been large enough to suggest that different rules of thumb had been used.... I am 100% confident that the differences in interference fit have not been due just to manufacturing tolerances and wear.
Therefore all I was able to conclude is that various designers have differnt ideas as to what initial interference fit the bearing race is within it's crankcase housing.
That didn't help me...as I found everything I measured to be different to what bearing manufacturers published as guidelines for the bearing installation.....!!!
So, what do I do?
I use a slightly tighter 20 degrees Centigrade interference fit for the bearing outer race than what bearing manufacturers specify because that is in agreement with the overall trend of the several different (modern) production engines thse that I've measured.
Also, I note that some people are referring to preload.....and I am not 100% certain of they mean interference fit of the bearing within it's housing or if they mean axial preload of the crank within it's bearings.
Sure, the Velo with it's tapered needle roller bearings had axial preload of the crank, but I dount if any other engine using rollers(!) lipped rollers, or ball bearings used axial preload.
But I am certain they all used interference fit of the bearing within their housings and low or size for size fits of the bearings on their mainshafts.
For racing engines I far prefer (where possible) to fit a top hat type steel sleeve in the cases in which the main bearing (still C3) is fitted.
Having said that, I've not had one problem or even evidence of, a roatating main bearing outer race in any of the engines I have built.
So maybe I am doing the right thing.
In the meantime I am surprised that on-one has offered the bearing manufacturer's calculations to illustrate the remaining interference fit such a bearing has within it's housing at opertaing temperature.
I have done it...ages back....with help..... but it would be illustrative to see it presented again....and if it does, I'll definetly dig mine out and see if it is the same.
Cheers and goodnight,
Greg Summerton
http://www.eurospares.com/greg.html
http://www.youtube.com/user/grannyjumast
Phone: +61 (0)8 82639018
Mobile: +61 (0)422 580 642