Bearings

May be odd, but no doubt there are bunches of different bearings in any car. Crank, cam, wheel, tranny, final drive and so forth. My best race buds are of all things, an engineer, a former top fuel mechanic, and a machinist and at the last race we were at, LBGP, a jet engine on a airliner blew, and we got on the topic of the RPM they run over dinner. We were stymied when it came to bearings used in F1, ceramic, steel, un-optanium, who knows, thought I'd bring it up here and see what you've surmised

Edited by whitewaterMkII, 23 April 2018 - 04:08.

I don't have a clue how you put the turbine bit of a turbocharger at one end of a long shaft with the compressor at the other. But I am impressed that Mercedes-Benz made it work on their F1 cars.

Self aligning ball bearings cure many ills. http://www.nachiamer...g-Ball-Bearings . Bendy shafty no problemmo. Also very low losses.

Ta, Greg. I can see how bearings designed for use in Plummers Blocks might work at 10,000 rpm. Turbo speed? That takes a lot of confidence and testing and co-operation. And you still have the skipping rope shaft to sort out.

I found this article from SKF about their bearings for F1 that includes this basic summary:

 

 

“We use very hard powder metals for gearbox and wheel bearings, high nitrogen steels,  including our proprietary Nitromax alloy, and M50 (an extremely hard and heat resistant tool steel) for turbos,” says Jeroen Wensing the company’s Innovation Manager for Racing. “In most cases, bearing rollers are ceramic, in order to reduce friction and save power. Cages may be made from PEEK (a high performance polymer), titanium or steel. Plain bearings for suspensions and similar applications are made in steel, titanium or even aluminium, with special liners in PTFE and ceramic coatings on the ball.” 

You have to understand the rotor/bearing system well to design high speed bearings.  That means calculating the natural frequencies that occur and making sure the bearings are located at appropriate locations to control the shaft.  Running higher than the flexible critical speed is possible but you have to understand what you are doing.

Running very high speeds with oil bearings is no problem--small turbos in production cars run over 300,000 rpm. 

With rolling element (ie, ball) bearings, the parameter of merit is the "DN" value which is the bore of the bearing in millimeters times the speed in rpm.  1 million DN used to be considered a practical limit but people have gone well beyond that--it just depends on the life required and how well balanced the rotor is and how well the rotor/bearing system is designed.

commercial ball bearing turbos run in excess of 175 000 rpm comfortably..

Running very high speeds with oil bearings is no problem--small turbos in production cars run over 300,000 rpm. 

That sounds excessively high, would like to see the compressor map of a small turbo spinning at that sort of speed.

here.. smallest garret... map ends at 300000 rpm... so I doubt they run over 300000 in normal production cars.. but who knows..

http://turbochargers...trim-80-hp.html

It was around  200,000 rpm twenty five years ago, when lag was a big issue. That was when the easy go faster cheap option was a turbo.

here.. smallest garret... map ends at 300000 rpm... so I doubt they run over 300000 in normal production cars.. but who knows..

 

http://turbochargers...trim-80-hp.html

Is that for a weed whacker or something?

I looked on the Garrett site and the smallest one they sold went to 195,000

Edited by 7MGTEsup, 10 May 2018 - 13:38.

I guess it is for something like a japanese KEI car.. which are 800 cc or a snowmobile (500 cc)

the new "real man" turbos from BW, with ball bearing cores in the 500 hp range run up to 150000, however the bearing is not a limiting factor but the turbine wheel which likes to explode..

I Sometimes thonk it could be "fun" to put seperate Turbos on each cylinder. Maybe 4 of the small ones will work on 4 pot engine!

With a cross over pipe from one exhaust to a turbo sending fuel to a different cylinder... 

I Sometimes thonk it could be "fun" to put seperate Turbos on each cylinder. Maybe 4 of the small ones will work on 4 pot engine!

I'm pretty sure someone has already done it years ago on a V8 and there is the recent quad turbo (not quite 1 per cylinder but still 2 more than most people run) 2JZ in that 200SX drift car in Japan.

It was around  200,000 rpm twenty five years ago, when lag was a big issue. That was when the easy go faster cheap option was a turbo.

Go faster for about 5 min then  a cloud of white smoke. Often tinged with blue as the engine detonated.

And they all had telephone numbers of horse power and I would drive by them with 200hp less.

I inherited a race car with some balls missing from the wheel bearings to make it run freer and hence go faster.  I know cyclist do this and are positive it works but in my case it is 4 balls from each bearing. My question is simply.....will it make much difference?

Whatever small gains you make in rolling resistance will be offset by the poor stiffness in the steering system. 

I inherited a race car with some balls missing from the wheel bearings to make it run freer and hence go faster.  I know cyclist do this and are positive it works but in my case it is 4 balls from each bearing. My question is simply.....will it make much difference?

Experiment. Buy a bicycle wheel and experiment. Spin the wheel with different numbers of hub ball bearings. Use different sorts of brake and video record how the rim moves. Create more complicated scenarios and test them. You can't afford these experiments?

If you can't afford to test, assume that the manufacturer knows best.

You'll get further by preparing your car properly than looking for a quick performance fix. 

Whatever small gains you make in rolling resistance will be offset by the poor stiffness in the steering system. 

As usual, the adult in the room.

Two things, even Lego bearings can run hard. here is a Lego V-12 being motored at 40,000 rpm.

Yes I know it blew up but even so

.https://www.youtube....h?v=vv8cQbn0L9o

Also, coming back to real life, way back Pontiac came up with the bent propshaft to create a flat floor on a Tempest with a rear transmission and IRS. The shaft drove at engine speed which helped and it was deliberately very thin, presumably not hollow either. The bearings not only had to steady it at up to 6,000 rpm but force it inwards to exactly maintain the bent profile.The bearing loads must have been quite high.

Didn't help sellt the car but who said 1960's GM lacked original ideas!

most small high rpm turbos use plan bearings, usually floating, so it is not so much of a problem..

Plain*

I would like to see the spec sheet for a turbo that spins even close to 300k on ball bearings. I am aware they did 100k + on brushless DC motors and those motors had some challenges with the bearings initially.

Edited by MatsNorway, 21 June 2018 - 21:24.

280000 here https://garrett.hone...-a-turbo-works/

I was more interested in the ball bearings Greg   

If bearings can run at up to 300,000 RPM or so - the rolling elements inside the bearing must rotate at quite a few million RPM?

Depends on size and bearing type (basically a dm*N limit)

Special cage designs allow much higher speeds than standard bearings. However thermal limits can require advanced lubrication (sometimes even active cooling / lubrication of inner and outer ring - too expensive for high volume applications)