Roller Bearings

[Martyn Nutland 1Participant @martynnutland1]

I hope you'll forgive this question because I've already solved the problem, but would like to be a wiser Easter bunny.

Does it matter (strictly speaking), in an originally designed roller race with nine roller if you have eight?

I'm sure you've guessed! Austin Seven camshaft. The centre bearing is a nine roller roller race that is an absolute design abortion of assembly technique. The stuff that unending nightmares and perpetual damnation to Hades are made of.

When I originally assembled my engine I became so desperate and frustrated I put the camshaft in with eight rollers. The great and the good said 'oh no! No way! Conscience got the better of me, I pulled it all aoart again, and after a day and a half of workshop misery my wonderful wife intervened, dextrously managed to get an elastic band around the nine rollers I had stuck in their track with Dum Dum compound and we entered the camshaft with all nine rollers.

To cut a long, tedious story short, would eight have been catastrophic?

A happy and peaceful Easter to you all.

Martyn

[Martyn Nutland 1Participant @martynnutland1]

Does one less roller matter!

[Ady1Participant @ady1]

I would think that at higher speeds and loads 9 is better and 8 wouldn't be as good

[HOWARDTParticipant @howardt]

The problem is not easy to answer. Schaeffler website should give some insight, if you can find the right article. That is one website that has got worse to navigate over the years. It will depend on actual diameters and the clearance created by the number of rollers. Less rollers will allow the shaft to flex more as will more shaft, roller, bore clearance. I assume the roller diameters are supplied within a fixed tolerance and not with a size/ tolerance choice. As I have said the Schaeffler website used to have detail of filling a gap with rollers for special uses, but it is a while since I have used this.

[Dave HalfordParticipant @davehalford22513]

A touch less valve lift on 2 & 3.

[Former MemberParticipant @formermember32069]

[This posting has been removed]

[RedsetterParticipant @redsetter]

I have, unfortunately, been involved with vintage and classic cars all my working life.

I am willing to bet that most of the great and the good who you consulted are enthusiasts and not engineers. They are probably a very nice bunch of chaps, but do not really know what they are talking about.

I doubt if the bearing in question is working anywhere near its design limits, so you would probably get away with 8 rollers if you kept quiet about it, and you would not be the first.

As to valve lift, if it worries you, you can set the clearances individually to give you the lift you want.

[old martParticipant @oldmart]

I wouldn't think the max engine revs are very high, and the camshaft is half that, so the bearing should just about last the remaining few thousand miles expected from that sort of car nowadays.The last time I rode in one was in the next door neighbours car, he spent most of his spare time taking the brakes apart, which fascinated the kids.

[Howard LewisParticipant @howardlewis46836]

With one roller missing, when load is applied, the rollers would tend to separate, leaving a gap where the ninth roller was supposed to be. The load would be taken bu four rollers, (the rollers on the unloaded side will just that, unloaded ) instead of five which to spread the load, and depending on the safety factor, would shorten bearing life.

You did the right thing by persisting and doing the job properly..

Howard

[Tim StevensParticipant @timstevens64731]

If the bearing is 'crowded' – ie the rollers contact each other and there is no cage (as implied in the description) – there is a risk that a roller could tip enough to cause problems. How much of a risk depends on all sorts of factors. The A7 is alleged to be capable of 9,000 rpm, but the camshaft only does half that.

And the idea that enthusiasts are dumb and engineers are the ones that get things right is not supported by all the evidence. Tay bridge, for example.

Cheers, Tim

[Clive FosterParticipant @clivefoster55965]

My understanding is that the continuous load capacity of a caged roller bearing is actually somewhat greater than that of a crowded roller bearing having rollers of the same size. Obviously a caged bearing has fewer rollers than a crowded one so, given appropriately hard surfaces to run on, the benefits of the extra roller would seem limited.

However crowded bearings can use longer rollers which ought help carry more load.

Whether it actually matters unless the ebaring is very marginally sized, and lubricated, for its application is moot. I have seen references implying that a crowded roller is better at the sort of high peak cylic loads seen by IC engine big ends and similar. But motorcycles generally moved to caged bearings in such applications even when crowded ones were the original design. I guess, like so many other things, its a case of learning how to get the details right.

Clive

[SteviegtrParticipant @steviegtr]

I would have thought it would maybe ok unless a roller turned slightly diagonal due to not being packed together. The result of that would be catastrophic.

Steve.

[Nicholas FarrParticipant @nicholasfarr14254]

Hi, I agree with Steviegtr on this one.

Regards Nick.

[Martyn Nutland 1Participant @martynnutland1]

Thank you very much all. I'm now much wiser. Had I been relying on eight, I like the simple expedient of adjusting the tappets to reduce the load around the bearing – of course, you would always have smart a—s telling you you had a loose tappet! I don't think a roller could get diagonal as, width-wise, they are pretty tight in the track on the camshaft.

Thanks again.

Martyn

[Raymond AndersonParticipant @raymondanderson34407]

Can't offer any advice on your problem other than Do as HOWARDT suggested and contact Schaeffler they will keep you right on this.

[Peter Jones 20Participant @peterjones20]

Posted by Clive Foster on 12/04/2020 20:55:55:

My understanding is that the continuous load capacity of a caged roller bearing is actually somewhat greater than that of a crowded roller bearing having rollers of the same size. Obviously a caged bearing has fewer rollers than a crowded one so, given appropriately hard surfaces to run on, the benefits of the extra roller would seem limited.

However crowded bearings can use longer rollers which ought help carry more load.

Whether it actually matters unless the ebaring is very marginally sized, and lubricated, for its application is moot. I have seen references implying that a crowded roller is better at the sort of high peak cylic loads seen by IC engine big ends and similar. But motorcycles generally moved to caged bearings in such applications even when crowded ones were the original design. I guess, like so many other things, its a case of learning how to get the details right.

Clive

In my understanding, motorcycles use caged bearing big ends for a number of reasons. The centrifugal loading can force bearing rollers to skid at high rpm and even if they do keep rolling you have a high friction contact as one side is moving opposite direction to the other

Honda (or maybe Suzuki? )did a training film on it in mid 60's

I doubt you've taken a Harley Davidson crankshaft apart but it's often possible to tell exactly what 'owner' was doing to motor at point of failure (wear pattern on crank-pin)

[colin hawesParticipant @colinhawes85982]

It seems to me that missing rollers could twist sideways and cause some cross locking which could result in the hard rollers skidding and cracking . Colin

[Howard LewisParticipant @howardlewis46836]

One important factor which the designer will have taken into account, we hope is the ratio between the roller and the journals.

I know of one case, ( a Lanchester balancer ) where the shaft continually brinelled under the action of the needle rollers. When I checked, the shaft was 6.1 times the diameter of the roller, so that the load came at almost exactly the same point on the shaft every revolution. Relatively small size changes removed the problem. Akin to a hunting tooth in a gear train.

So changing to a fewer larger, or more smaller rollers could cause unexpected problems,

Howard

[Brian HParticipant @brianh50089]

I'm afraid that changing size of the rollers is not an option, the 'inner' diameter is the diameter of the camshaft and the 'outer' one is a hardened steel bush inserted into the cylinder block.

I would have thought that The Seven Workshop would be a good place to contact, their site shows them at £4.50 a set inc. VAT.

Brian

[Peter Jones 20Participant @peterjones20]

Can you get the rollers in different diameters to account for reconditioning?

I have no experience of the motor your working on but I do have pretty vast knowledge of various engines. (mostly motorcycle, Aerial to Zundapp with petrol, diesel and propane industrial motors thrown in for good measure)

Harley Davidson list 3 different diameters to get correct running clearance.

There are also a few extra sizes 'aftermarket'

The only 'difficult' part is accurate measurement of bearing bore.

If it's through hardened bush it may be possible to have it ground over-size on the ID and get more favourable needle diameters?

Math is simple and you get very accurate results which will result in a better longer lasting engine

[HopperParticipant @hopper]

Eight rollers might not be catastrophic. It might just lead to rapid wear as rollers skew and or bunch up.

[Author]

Posts