Ball bearing Turbochargers

[Diver944]

I've started this new topic because of some points being raised in the EBC thread.

Simon Peckham has installed a ball bearing hybrid of a K26/10 on his 317Bhp Turbo and the spool up is incredibly quick[8D] If you take a look at Simons boost chart

http://www.cannell3.co.uk/images/Dyno_Graphs/G886PUA_Boost_3.jpg

from the recent Dyno day you can see that by 2500rpm he is making 5psi of boost, by 3000rpm he is making 8psi and at only 3500rpm he is at full boost of 21psi

Compare these figures with Mark Koeberles mega monster 402 Bhp car

http://www.cannell3.co.uk/images/Dyno_Graphs/C7KOE_Boost_2.jpg

and we can see that by 2500rpm he is making 1 psi of boost, by 3500rpm he is making 1.5psi and at 3500 he is up to 3psi. Peak boost of 21psi is made at 4800rpm, but is held all the way to the redline which resulted in Marks proud ownership of a Weltmesiter 400 cap (and a big cheshire cat grin all day [])

So the question is 'Why do ball bearing turbochargers spool up so quickly?'

 

[John Sims]

Ah yes, but there are so many other factors to consider also.

They are different turbos in different cars. Mark, for example, could be getting more blow by on his cylinders and so needs more pressure to overcome the pressure he is loosing. Marks turbo could be/is bigger than Simon's and will have more inertia as a result. Marks big valve head may be loosing gas velocity at lower revs. Marks intercooler may be reducing the gas velocity through cooling and expansion.

The only way to really know what is happening is to have exactly the same turbo on exactly the same car, run it without ceramic bearings and then run it with.

That said having ceramic bearings sounds damn good and must provide less friction through all temperature ranges so must be better. [] Further noting Simon's expertise in miniature jet engines you are pretty much getting Rocket Science. [8D]

 

[sawood12]

Is the spool up speed a resultof the bearings? Wouldn't it be because of the smaller turbine side of the K26/10 turbo which can't hold the boost as it runs out of efficiency at the higher revs and can no longer drive the larger compressor? I take it that Mark Koeberles larger turbo is just laggy due to the larger turbine that is needed to drive the larger capacity compressor to hold the boost pressure all the way to the red line.

 

[Diver944]

Yes you're both right, that was perhaps a bad example to try and answer the question about ball races

Here is a slightly better example:

http://www.cannell3.co.uk/images/Dyno_Graphs/NIJ4095_Boost_2.jpg

It's Rob Hunts graph and Robs car is running on the original K26/6 turbocharger so is 'similar-ish' in size to Simons ball race version. I was going to pick my own graph from last year when I was running a K26/6 but my boost is in bar, not psi and is correlated with road speed not RPM []

At 2500rpm Robs car is making 2psi of boost, at 3000rpm he is making 4 psi and at 3500rpm he is making 5psi. Peak boost of 14psi is made at 4500rpm

 

[hartge bob]

Actually Paul, I think you may have picked a bad one in using mine []

I think I may have a boost leak at the moment, I noticed on the way back from the dyno day that it didn't seem to be boosting quite right.

If you look at my graph from last year

http://www.cannell3.co.uk/images/Dyno_Graphs/NIJ4095_Boost.jpg (sorry, it is in bar, but is against rpm)

you can see I was getting peak boost at 3500rpm, much the same as Simon (even if it's not quite at the same level).

Of course John may have hit on something too, I do think I have a bit more blow by this year too, so this may account for the difference.

 

[Indi9xx]

Oh dear... get ready for a very long winded reply )

Its going to take me some time to type it all in though !

But before I do... a couple of points need to be made.

I do not make the turbo's, Simon designs them, Simon sells them, so everything I write is 100% objective and I am not plugging his turbo's.

Simon's turbo uses a K26/6 hot side (exhaust side) housing... But its no longer a K26/6 housing, its been precision machined to accept a much larger exhaust turbine than a k26/6 or k26/8.. if there was a larger off the shelf 944 housing, it would be an exhaust housing with a designation something like k26/10.

A 26/6 housing is used because a new K26/8 housing is about £500 and even an 8 housing would have needed to be machined to accept the turbine we wanted to use... A used housing was specifically used because A:Its cheaper, B:Its already heat cycled 1000's of times and so is de-stressed and C:He had a spare one on the shelf.

A KKK exhaust housing was used because A:They fit the exhaust crossover and the rest of the exhaust system, B: They are the best most durable housings on the market, and C:They are over engineered to such an extent that they can be machined out for larger sizes of turbines.

The rest of the turbo ISNT KKK, K26/6, K26/8 or K26 anything, its all custom hybrid Garrett.

And lastly, Simons car is reducing boost at the top end at the moment for a very good reason... Its because NO ONE is able to give a good reason why a a manual boost controller has problems holding boost like an electronic one... There are lots of theories, but no good data... Simon has offered his car for experimentation into this and I am working with him collecting data about what exactly happens when a MBC looses boost, which will result hopefully in an MBC which can maintain boost almost as well as an EBC..

Some cars like Mark Dalton's 944 turbo, hold boost very well with a MBC (342BHP, 18PSI of boost and only 1PSI of creep)... Others can be the complete oposite. Simon has fitted an MBC to his car and will have one fitted until we get to the bottom of the MBC issue... Afterwhich, Simon will be fitting a EBC, another WRC run will be made with it, which will show the turbo as holding full boost from 3000RPM all the way to 6500 RPM!

Anyway, now I will write the long winded answer to the advantages to ballrace turbo's. (yes, this post was short!)

 

[chrisdenton]

Blimey Jon, Its been nearly half an hour. This better be good......[][][]

 

[Indi9xx]

Sit down, get a cup of tea.. this is going to be one of those epic long "jon mitchell" posts... I'm sorry!

A ball race turbo is called a ballrace or ballbearing turbo because the shaft of the turbo runs on, is supported by and controlled by ball bearings running within an assembly (cartridge)... With a ball race turbo the shaft is always running within tolerance within a ball bearing assembly with very little friction.

A conventional turbo uses a collection of bearing journals (bushes almost) and thrust bearings (washers/p-clips almost) to stabilise a turbo.

Because of this difference, a ballrace turbo will always spool up 15% sooner than a conventional turbo due to there being 50% less frictional losses within the bearing assembly compared to a conventional bearing assembly.

A ballrace turbo requires much less oil flow and pressure than a conventional turbo Which means that your oil pump is required to pump less oil to the turbo, resulting in less oil cavitation, less oil froth within the sump, less stress on the turbo's oil seals, less likely to be oil seal failure and lower oil temperatures.

Not only this, but because the ball race turbo relies less on oil pressure, there is less float within the turbo assembly, which means that tighter tolerances can be used, which means that turbine and compressor wheels can run much closer to the turbine and compressor housing, which in turn means that less exhaust energy is lost and less resulting shaft energy is lost within the compressor side... In other words, the ballrace turbo is less leaky than a conventional turbo.

The other wonderful thing about ballrace turbo's, they are less dependant on oil than a standard turbo, which means that at start-up, where a conventional turbo will run dry for a few seconds the shaft is running in direct contact with its bearings, which will cause wear to the bearings, which in turn will cause even more float within the turbo shaft and wheel assemblies. A ball race turbo will not be affected by start-up causing such wear, because the shaft is always supported by the ball bearings, rather than the conventional turbo relying on oil to protect the shaft!.. The same is true on shutdown where we have all heard horror stories of turbo's being killed during shutdown of the engine, with a ballrace turbo, the chance of any damage during start-up or shutdown is diminished to nearly nothing.

Another benefit to ballrace turbo's is that they do not have the entire shaft running on high pressure oil, as oil even though its slippery, is viscous which means that when the turbo is spinning at 120,000 rpm to 190,000 rpm, that oil is resisting its motion.... What would be easier to turn by blowing on it... a child's windmill with a bearing involving brass bushes and engine oil, or one with a ball bearing assembly?

The main advantage that the driver will feel with a ballrace turbo over a conventional one is the way the turbo spools up If you drive Simon's car, you instantly realise that if your above 2500 rpm, the second half of the throttle travel proportionally controls boost with no lag!!!... Now I am not joking, on the first 50% of throttle travel the car drives like a 944 lux, on the second 50% it feels like a VERY VERY BIG non turbo engine that unlike a non turbo engine, does not run out of steam!.... You can play with the throttle while watching the boost gauge and the boost gauge reacts like a display of what the throttle is doing! At first it feels very strange, but after a few seconds you really get used to it. I can imagine that on a track it would be absolutely amazing, on the road its just so drivable... its like having a really massive engine without the fuel consumption!

But beware there are other turbo's which say they are ballrace or ballbearing turbos But all you need to call a turbo a ball bearing turbo is for it to have a ballbearing in the assembly somewhere, I can guarantee that all the other ball bearing turbo's on the market for the 944 turbo (that I know of), do not have a fully ball bearing supported and controlled assembly as does Simon's. They cheat by having a single ball race, they still have thrust bearings (washers) and normally they still have bearing journals as well!.

The other thing about turbo's is how they have been balanced... if it looks like the nut on the end of the turbo has been attacked with an angle grinder to balance it, it probably has been attacked with an angle grinder to balance it.... A turbo should be balanced in several different ways to finally be "balanced", all the turbo's I have fitted in the past have not been balanced as well as they should be... Which means they will not last as well as they should and they will not spool up as well as they should.

I know people should not plug their own products in here, and these turbo's are not mine (you can buy them direct from Simon), but Simon knows more about the technical details of some other wonderful aspects of his turbo's than I do, so I would love the moderators and members to give the go ahead for Simon to fill in some details that I have missed.

 

[wizard]

One aspect of Simons turbo he was keen to point out is the mouth of it is modified to enable a faster intake of air.
I remember from the 'old days' of carbs we used to use 'ram pipes' within the air filter to do same thing.

Just to add: I have never experienced a 944 turbo on full song blast past me unitil Simon did on our way up to Silverstone. Whilst it was very fast, you could just see the car trying to rip up the tarmac! this was evidence of the large amount of torque present. Very impressive

 

[Indi9xx]

Thats just one of the other secret (well not very secret now) weapon aspects of the turbo, designed to improve the feed to the front sides of the compressor wheel to again improve efficiency.

 

[wizard]

oops![&:]

 

[Indi9xx]

Dont worry, its not hush hush, just one of the developments that are part of the package.. Nothing thats really that secret, just something that no one else does on the 944 turbo.