Roller bearings in crankshaft?

Dec 14, 2004 19 Replies

Since the pressures are pretty fierce and it's the most hostile environment any bearing set on the engine has to endure, how come roller bearings aren't used for that application?



Des.


I guess it's cos you'd then need to have a pressed together crank like some bike engines and that's probably more expensive. They are also prone to twisting, a mate of mine twisted his GS1000 crank when he missed gear once.

Also I reckon that a plain bearing would have a much larger surface area to spread the load across.

environment

On many motorcycle engine cranks, roller bearings are used fairly often. The upside is longevity (if the oil is kept clean), but there are many downsides as far as the motor manufacturers are concerned, mostly related to cost. Most cranks are forged from one piece of steel then the shell-bearing journals are ground to size. This is a cost effective and well understood technology and the machinery is well proven. Also, if a shell bearing gives trouble, it is very straightforwward to regrind the journal and replace with oversize shells. For roller bearings, the crank has to be produced from separate sections then pressed together very accurately. To replace a roller bearing on one of the inner journals of a pressed-up crank, it can get very expensive and is often cheaper to replace the whole crank assembly.

JB

Also sprach "Designori.." :-

I'd have thought the small end was the most hostile - higher temperatures and a back and forth action which must be harder to lubricate than rotary.

Anyway - since the oil pump doesn't seem to have much trouble keeping the crankshaft afloat so there's no metal to metal contact, why change it.

They're cheap, simple and they work. And quiet, of course.

environment

The

environment

Because you couldn't do it with a single-piece crankshaft, which creates a whole load of other problems - especially in a high speed engine - and is expensive to manufacture. I did once come across a vintage single cylinder tractor engine with roller bearings and a built-up crankshaft - max speed around 1000 revs/min.

D A Stocks

environment

Because a roller bearing needs to slid over the end of a shaft, and you cant get to the end of a crank journal, because the webs are in the way!

Some engines do have roller bearing cranks (very rare) not sure how they do it, will have to be a cranks that can be split, or some sort of splitable roller bearing. Either way would be complicated to get right.

environment

They are. They are quite common in m/c engines, but in fact size for size, plain bearings are capable of taking higher loads than roller bearings. Mike.

I'm guessing here, but I can think of several reasons.

  1. White metal bearings seem to work well and are not expensive.

  1. The assembly would be a (horrendous) problem. Some single cylinder motorcycle engines have roller bearings but you can split the crankpin away from one half of the crank web, put on the bearing and put the crankpin and its nut back. Then you have to align the journals at the ends of the crank on a lathe-type machine. I suppose you could make roller bearings where the races were in two halves but the thought makes me go weak at the knees.

  2. If a roller bearing departed this life it's likely that it would happen very quickly, whereas plain bearings generally give some indication that they are tired.

Anyone got any more ideas?

Rob Graham

My bike mechanic races 1960's vintage bikes with a pressed up roller cam. He reckons that from starting to hear a roller go to the event termed as Apocalypse Now he has about five seconds and coasting the engine down from

18,000rpm (a four stroker with DOHC at 18k - scarey!) takes three seconds!

Purely down to cost and "we've always done like that".

You can get split race (and cage) ball, roller and needle bearings that would allow a one piece crank but they are very expensive.

Some of the highest revving engines in the world run on roller bearings. All modern 2 strokes use roller and ball bearing as a shell can't cope with next to no lube. 12,000 rpm was not uncommon and some

60's racers ran 18,000rpm. Model aero engines run on ball bearings - 50-60,000rpm. Turbos either use a floating bearing (shaft runs in a bush which runs in the housing so splitting the rpm between two plain bearings) or they use a ball bearing, running at 100,000 rpm and up. Jet engines run HP spool 12,000rpm on a ball bearing at the front and a roller at the rear. HP turbine on an 80K lb thrust engine produces 77,000bhp.

Back in the mid 70's a car mag set out to destroy a poor little Honda CB125S. They ran it out of oil and it kept going, so they stuck it on it's stand and held the throttle wide open in neutral. Con ron ventilated the crankcase at 18,000rpm. Roller bearing crank. Red line was 10,000rpm but I had mine out the other side to 12,000rpm in top.

Look up Tatra they used roller bearing cranks for cars and trucks.

Can't be that horrendous a problem. I had a 1966 Saab 3-cyl two-stroke with roller bearing crank. I never had the crank apart so can't comment on construction bit I think the bearings were conventional design (with very large balls). Come to think of it, aren't most/all two-strokes roller bearing? I guess in a two-stroke, because you don't have a camshaft etc to lubricate, the saving on oil pump and machining oil passages may offset the extra cost of fabricating the crank assembly? The distributor drive was in a sealed chamber lubricated by grease.

Biggles

Which is why all the Japanese 2 strokes and most others since the mid

60's had needle roller little ends. (Well not small stuff like chain saws and mowers)

As the little end motion is only about 35 degrees total (~1/10th of motion on big end or mains) on a small ~20mm dia pin (~2/5ths size of big end or mains) the velocity of pin surface to piston or rod bore is much slower. At max rpm it's comparable to running at tickover on the mains. Next to no wear is caused in a well lubed oil mist such as you find in a fully warm 4 stroke. Just go steady until you have hot oil to make that nice lubricating mist.

Excessive over revving can make little ends go oval as the inertia load of the piston at TDC stretches the little end.

Forgot to mention it's almost completely unloaded at TDC during ignition as the gas pressure on the piston balances the piston inertia. So highest load is during TDC during valve overlap and BDC on both strokes. Assuming you don't run into detonation that's just nasty for a whole lot of bits as the piston tries to make the engine go backwards while the whole inertia of the vehicle drives it forwards.

Apaprt from all the other replies IIRC the Great Western railway tried roller bearings on steam engine cranks and coupling rods and found that they were no more efficient than plain bearings and cost far more to produce and maintain. Of course the application is slightly different but the pressures involved must be similar.

environment

The pressures are too fierce for rollers to endure. Also you have to understand that the crankshaft and conrods don't actually sit on bearing shells, they sit on a thin layer of oil between the bearing journel diameter and bearing shell it'self. You also have to remember that the shell bearings used as sacrifice should things fail or wear out. If a roller bearing was used then I suggest that should it fail, any bits and peices flying around in the sump could do hell of a lot of damage elsewhere rending the crankshaft or even engine to be scrapped.

I used to think our company made big diesels untill I saw this......

formatting link

- Fatsod

Can't be that horrendous a problem. I had a 1966 Saab 3-cyl two-stroke with roller bearing crank. I never had the crank apart so can't comment on construction bit I think the bearings were conventional design (with very large balls). Come to think of it, aren't most/all two-strokes roller bearing? I guess in a two-stroke, because you don't have a camshaft etc to lubricate, the saving on oil pump and machining oil passages may offset the extra cost of fabricating the crank assembly? The distributor drive was in a sealed chamber lubricated by grease.

Biggles

All my chainsaws do have needle roller little ends.

AJH

Can not run a thin wall shell bearing on total loss lubrication and having to seal every main and big end bearing would be hopeless for seal drag. The need for a seal between each crankcase chamber also gave high drag. On racing 2 strokes seal drag was so important that they used expensive non contact labyrinth seals. Big truck 2 stroke diesels have an external pump to scavenge them so they can use a journal bearing crank and big ends with just a seal at each end of the crank.

Also roller bearings have much less drag giving a higher mechanical efficiency. This is why they tended to be used on motorcycle engines where power output is the prime consideration. Simple to install with virtually zero maintenance - witness all those wheels bearing we are running around on. No sump, no pump, no filter and no oil changes.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required