Help with sizing hot sides

[S2central.net]

Right - some of you will be well aware (and/or bored rigid) of my fascination with mathematical equations surrounding turbo compressors and how to estimate airflow, power, torque and other theory on aero & rolling resistance calcs, fuel pressure & flow, compression ratio (blah blah blah) and too many hours spent on spreadsheet calculators.

My turbology paper goes into lots of detail about compressor maps, based on books I've read, websites pored over and conversations with some of the mob on here.

http://s2central.net/Turbology_v4.pdf
One thing that continues to elude me is some useful rules of thumb (or full blown theory if i can get it) for calculating the best hot-side of a turbo for a given application.

I mean its all very well saying that a certain compressor is capable of flowing enough air to generate 400bhp, but its altogether another matter to calculate if the hot-side of a K24, K26 or whatever can generate enough energy to spin the compressor wheel fast enough to produce such flow numbers.

We all know that a bigger hot-side means later and greater power delivery than a small hot-side, but I need some numbers to complete the mathematical analysis.


Paul

[Mark Halligan]

We all know that a bigger hot-side means later and greater power delivery than a small hot-side
Paul

Paul,

its worth remembering that there are other variables to consider too in your calculations (if possible), its not just the size of the exducer but there are other factors to bear in mind, i.e. the bearings used in the turbo for one.
In my turbo I have a very large exducer bore of 63mm in a #8 hot side, with what I regard to be very good if not a bit brutal power delivery, due to the ceramic BB, although my EM may have something to do with it too.

In a KKK (or any other turbo for that matter) with conventional floating bearings a large exducer would give very good top end but a fair bit of lag, however with the a ceramic BB or the roller bearing (as fitted to the GT series turbos), the power delivery is vastly different, and side by side comparisons of identical turbos with different bearing assemblies will most likely be very different in performance and delivery up to top end, and the cars very different animals to drive.

Another thing to bear in mind is the weight/mass of the turbine and compressor wheels used, Doug has recently, I believe, changed his large K29 comp wheel to a smaller K27, as his spool up was very linear and without that punch often associated and desired with our motors although his exducer/turbine is not overly large.

Hope that all makes sense and please tell me if it doesn't. I am a nuts and bolts person really, and I find that I don't have the time to get my head around the maths of it all, which makes these posts a good read for me.

I like this thread especially, as just lately it seems that the emphasis is with compressors, and not the size of the exducer which in the end dictates the max flow and power output of the motor.

Just my 2 pence worth.

Mark.

[S2central.net]

Cheers Mark - I think its safe to say the hot-side selection is more of the black art in relation to turbo technology. What a pity that Garrett and others don't provide 'turbine maps'.

As I see it - it is the speed of the compressor wheel, driven directly by the turbine wheel which draws air in. The speed at which the turbine wheel rotates is determined by A/R ratio of the turbine, the rotating mass as well as frictional and heat related aspects.

As you say - it is well known that exotic turbo bearings can significantly improve spool up time so there must a benefit from reduced friction losses.

I doubt we'll ever collect enough usable mathematical equations to quantify all of this, I just want to get some rules of thumb that help us decide on hot-side permutations for any given applications.


Paul

[Mihnea]

Paul, what you're looking for is a diploma thesis in thermodynamics and fluid dynamics. IMHO, way too much theory in there to get anything serious and useable. Experience is the best way to know this sort of stuff.

Doug's comp wheel (3272spec) was way too big, heavy and inert indeed, although his hot side was and still is the same as Tom's. Full boost at 3800 RPM in Tom's car (K26#8 hot side with 56mm exducer), full boost (2 Bar) at 5000 RPM in Doug's car, with maxxed out WGFV duty cycle and a valve spring in the wastegate. Though, theoretically, due to the same hot side, the turbo should have had similar behavior to Tom's.


So all this is really way much more complicated than simple rules of thumb, that would be oversimplified and plain wrong in most cases...

Trial and error, and constant self-criticism is the way to go. I used to think DOug's turbo would do full boost at 4k based on what I had been told and I have found out there is no way.

BTW, a compressor can't be called K29 or K27, these names are wrong. Doug's old turbo had a 3272 comp wheel in there, the new one has a 3072, same as Tom's.

The RS2 is a 2672, a stock K24 turbo has a 2470 comp wheel, an Audi 200 10vt has a 2664 comp wheel, a porsche 911 turbo has a 2772, the sport quattro has a 2750 comp wheel.

The Audi 90 IMSA GTO had a K27-13-3476HHB spec turbo. Thus, K27 hot side and wheel, 13 cubic centimeters scroll volume, 3476HHB compressor wheel.


HTH,


Mihnea

[Mark Halligan]

So all this is really way much more complicated than simple rules of thumb, that would be oversimplified and plain wrong in most cases...

Trial and error, and constant self-criticism is the way to go.

Agree with you there, particularly the trial and error part. Mind you, imagine the fun you could have researching this subject practically, although it would take deeeep pockets or parts bin to finance it all .


Mark.

[twoqu]

Hi!,
Due to the lack of data for the bigger KKK turbo's, what we need is hard evaluative data. It is unfortunate that someone has to experiment first but thats whats brilliant about this forum. We get to know what works and what doesn't!
Mihnea's and Doug's experience seem to support that too much of a good thing (in terms of compressor wheel size) isn't necessarily better! Although the 8 turbine seems to be a good compromise.
That is also the first time I have seen a list of the Comp wheel identification numbers.
Ultimately it would be great to have a dyno mule and all the possible combinations of turbine/comp wheels for a flog. There would still be all the other variables to consider, comp ratio, cams, induction, charge cooling etc.
Lets see if the numbers come up tonight!

[S2central.net]

yeah obviously the theoretical stuff is only good when tested against measured performance of given modifications. I'm not looking for five magical formulas that do all the number crunching for a 500bhp engine.

Instrumentation is the key here - I have seen no data anywhere on the efficiency rating of the S2 and RS2 intercoolers ie their ability to remove heat at a given flow volume. The colder the air at any given volume and pressure, then the greater its mass. The greater mass of air ingested by the engine means more power !

Its all very well estimating where one is operating on a given compressor map - but without factual data for VE, then its a guessing game. By measuring the temperature of air as it exits the turbo will do two things - enable reasonably accurate characterisation of the intercooler's abilities and give some insight into the actual flow rate of air - and therefore VE - at any engine speed.

I may be putting a drill to my intercooler sometime soon for the fitment of two air-temp sensors. Not sure I'm wild keen on the idea.

None of this will help the black art of hot-side selection - much of that is going to be down to experience, trial and error.

Another idea that has occured to me is to fit a 2nd MAF after the intercooler that will provide useful air 'mass' information under the prevailing conditions...Assuming I ever find a useful datasheet that will let me get gram/second info from it.

So many ideas - so little time


Paul