BoostGuy951

I am trying to figure out what the stock (as well as modified) VE is of a 944 Turbo at several RPM Levels, From 3000-7000. It occured to me that you could use the stock Horsepower level at stock boost and figure out roughly what kind of VE the car has stock. Well, not having the formulas off hand, nor the patience to dig through the internet to find them, I decided to use the Turbo Map Calculator at http://www.turbofast.com.au/turbomap.html

I put in all the info required, and used 75% for compressor efficiency, and 65% for IC efficiency. I used 27 degrees celcius for temperature, and 11 psi for boost. I plotted two points on the RPM, 5800 (peak power) and 3500 (peak torque). I then played around with the VE until I got the Horsepower to read 220. It did this at 72.3%. I did a quick conversion of TQ to HP and converted the peak torque of 243 @ 3500 to 162 hp. Then I played with the VE until I got 162 hp at 3500 rpm. It happened at 88.2%.

So on a bone stock 944 turbo, it looks like this:

VE @ 3500= 88.2%
VE @ 5800= 72.3%

So exactly how much VE gain different modifications can get you?

Such as:

- a zero restriction 3" exhaust
- a bigger turbine housing
- some intake porting on the head
- a ported or sheet metal intake manifold
- a full radius valve job
- a 4-1 header like SFR's

This is all brought up to answer these questions: What do you have to do to sustain decent torque at RPMs up to say 7000 or 7500, without the torque falling off a cliff after 5500? Is it even feasible to rev to 7000 on an 8v head, or are you better off going to a 16V head?

Any input would be appreciated.

Ivan A

Boostguy

I'm pretty new at Rennlist but lets see if can get some where with your questions. Very long and convoluted answers and very hard to measure unless you have an air flow meter measuring CFM's at the different rpm's and boost levels.

After testing an 89 S we found the torque curve to be very peaky and not broad at all, as quickly as it got there it went down again. It's peak happened at 4700 not as low as you would have thought based on the info that Porsche prints. It maintained close to that level of torque from 46 to 56. The torque was about 260 at the tire which is more than the factory
claims at the flywheel.

You did not talk about cams here and I can tell you that they would have a lot to do with the VE of the engine. Remember that the motor is an air pump which is just sucking air in and pushing it out. The manner in how it does it is going to affect horsepower. NA motors can make 100% VE through some rpm ranges and turbo can exceed that for the simple reason of the force induction.

A lot of folks get hung up on compression for example, more is better. They measure the compression static, here lies the problem. Stock cam will not allow the motor to breath and fill that cylinder with air to make efficient. The dynamic CR is on the floor even if you had 15 to 1. See my point. There is no VE in the motor, you are not filling the cylinder so your are not going to have good cylinder pressure which will translate in torque. A natural aspirated engine suffers from low VE dramatically a turbo engine all it needs is more boost= more cylinder pressure and more torque. how much more torque again depends on the cam primarily ( how high and how wide) and intake flow that you have because if the throttle body and your air metering device ( MAF or air flow) does not allow the air that the head and cam combination can flow it does not matter. When I build an engine I flow intake and the head and have a cam cut for the particular application ( time and money permitting). It is the only way to know and after you are done only the dyno will tell you what you have. Most guys that go to do dyno do not use hardly any instrumentation so they can not know if they are on the right track or not. All they look at is the HP number and peak TQ. BROAD TQ band is the name of the game. Only way to achieve is with some good cylinder pressures, GOOD VE.

I hope i'm not being redundant and I helped a bit. Let me know if you like to talk about it on the phone, I'm tired of typing.

Respectfully
Ivan

Danno

Hmmm, interesting... However, you can't use 75% for the compressor efficiency. You have to use the stock K26/6's compressor map for that and I would suspect that the efficiency is in the 60-65% range depending upon which RPM-range you're looking at. Then there's the issue of the intercooler's efficiency as well. I doubt the 951's efficiency is that high in the 72-88% range. Perhaps we can start by modeling a simpler car, like a 944NA 1st and get all the givens to match the equations 1st.

BoostGuy951

I definitely agree that the cam would play a vital role there in keeping tq decent in the upper RPMs. I hear that Performance Developments has got a new cam in the works.

I hope we can identify some of the main items that are causing the huge tq falloff. I know some people have said that the stock turbine housing is one of biggest restrictions. Something I have noticed that also looks restrictive is the turn the crossover pipe takes just before it turns into the exhaust housing. When we make major upgrades like going to a bigger turbo, I wonder how much of the gain might be negated by the plumbing surrounding it on either side. (crossover pipe for the turbine and IC for the compressor)


Danno-

You might not have been around the forums when this happened, but several months ago, someone came up with a k26 compressor map, and surprisingly, it does run 75% efficient at its peak efficiency Island, which is around 11 psi and 3500-5800 rpms.

Obviously Peak torque is going to coincide with Peak VE, if variables like boost level and ignition timing and A/F ratio are a constant. You don't think that the 944 turbo could hit 88% VE at a measly 3500 rpm?

Wouldn't you run into problems modeling a NA 944 due to differences such as the NA motor having more cam overlap and also not having the restricive turbine housing, and exhaust plumbing?

roadrunner

In order to increase the rev range where usable power is made, you might also want to take a look at the intake manifold characteristics. Plenum volume and runner lengths and diameters will change the resonance frequency, as seen in VarioRam, to take advantage of any breathing improvements that would be done to the head and cams. BMW has pretty much set the production car standard for variable intake with the 745i...

Waterguy

Boostguy, the VE on our engines does not drop off as fast as you calculated. The stock (non-S) boost curve peaks at 11 psi, but is reduced by the cycling valve to 7.5 psi at 5800 rpm. I calculated the VE to be about 78% at 5800, dropping to about 70% at the red line (using 70% intercoler efficiency.)

When I do the same calculation for the Turbo S, which holds 11 psi to 5800 rpm, I get a slightly higher VE of 80%. Based on this, it looks like you may get a 2% VE improvement by going to the less restrictive #8 hot side. That may partially answer one of your questions.

BoostGuy951

Thats great Waterguy!

I am willing to bet that if its worth 2% at 5800, its benefits compound the higher you go in the RPM. Too bad we dont have a TQ curve of the stock S and non S that extends up that high.

I never knew the stock cycling valve did that. I bought my car with APE chips and boost set at 15 lbs.

So that 78% is with the stock restrictive exhaust system, and #6 hotside, as well as the stock airbox and subsequent plumbing. Anyone got any numbers for what a 3" exhaust from the downpipe out would be worth in terms of HP?

Going to a nearly zero restriction exhaust, and a MAP could be worth upwards of 85% @ 5800 it would seem.

I'd like to hear some arguments for why alot of people say that the VE is so terrible on these cars. It doesn't seem like it from this little excersize, but maybe we don't have the whole picture.

Waterguy

I don't think the stock air box, AFM, intercooler or intercooler piping cause a big problem. The boost pressure is measured at the intake manifold, so if you have pressure drops upstream, the turbo works a little harder. This results in hotter charge temperature, but the intercooler removes most of that heat.

The biggest bang for the buck has to be a test pipe and 3" exhaust, followed by a true Garrett hot side. Mind you, that's no surprise, the experienced guys have been saying that for years.

Tony G is the master of VE imho. The horsepower numbers he is putting out at 1 bar suggest that he must be closing in on 100% VE at high rpm. The mods sure don't come cheap, though. (And he's so humble and unopinionated! )

I can back calculate VE from a dyno chart if boost is also charted. The big assumptions (unknown variables) are intercooler efficiency (not really a constant percentage), the effect of varying timing and the effect of varying fuel mixture. I just calibrate to stock for these variables. If we can find a dyno chart for a 951 with just exhaust mods (plus chips, MAF, whatever) and has boost datalogged, we can look at the effect of exhaust on VE.

Tomas L

You can't use a program like that to calculate VE. First there is to much uncertainty in the input data like IC efficiency and actual boost pressure. Second, the program makes assumptions that aren't displayed, for instance BSFC. A change in BSFC from 0.5 to 0.55 will make a VE change of 8%. Third it seems to disregard of IC pressure drop, although this is not an influential parameter. All of this makes it a bad idea to try to calculate VE this way, a program like this should be used to see the effects of changes, for instance "if I can increase VE (or boost, or IC efficiency, or...) by 2% how much more power do I get?", not to get absolute values.

Since we measure boost pressure in the intake manifold and use it to calculate VE, any changes to the intake system before the manifold can't by definition change VE. They will though change the air density in the manifold and therefore change engine power, but with this definition of VE they can't change VE.
I have made an excel spreadsheet to calculate the intake side of a turbo engine and I have come to some interesting conclusions.
Only looking at the intake system effects Waterguy is correct in that the pressure drop in the AFM, IC or piping only makes the turbo work a little harder which causes a little more heat that is mostly removed from the IC. For instance a 3 psi pressure drop in the IC gives a theoretical power loss of 3 hp.
Another interesting thing I've noticed is that just considering the intake system there is almost no gain in switching to a bigger compressor in a mildly modyfied 951. If you calculate the power difference caused by the difference in efficiency at 350 hp for the K26 (appr 69%) and a K27 or TO4e-57 (appr 75%) you will get something like 4 hp depending on what parameters you use. Obviously the K26 is almost at it's limit and can't flow much more but this indicates that if you are under 300 rwhp there is almost no gain in switching compressor. I'm not sure that I believe that this is the complete truth.

My theory is that when the compressor passes peak efficiency it requires more power to drive it, the turbine has to work harder causing increased exhaust back pressure which lowers VE.
The same thing would happen with high IC pressure drop, 3 psi pressure drop would mean at least 3 psi increased exhaust back pressure.
Since the increase in exhaust pressure is not matched by increast boost pressure you'll get an unbalance between intake and exhaust. The effect of this unbalance would be even worse if you run a cam with some overlap.

It would be very interesting if someone have dyno charts for K26 and K27 (or TO4) on the same engine with no other changes between the tests than the turbo.

Tomas

tazman

Maye this will help you here is my dyno sheet for my turbos with ape chips and SFR 3" cat back.

TurboTommy

As far as IC efficiency is concerned, one clue might be the info on Lindsey's intercooler testing. I believe the stock IC was around 75% on a warm day at moderate speeds.

I'm still wondering how much VE gain you will get from improving the exhaust (turbo hot housing included) if there is apparently no valve overlap. You need some valve overlap for good scavenging (this would definitely increase VE)

Waterguy

Just the kind of data I need Tom. Nice curve, 16 psi boost at 3500 rpm certainly suggests that the catback is helping the K26-8 spool! You do get a slight boost drop at higher rpms. Even so, I would have expected a higher HP peak than 5200 rpm. Chips/AFR? This shows up in the back calculation as reduced VE, since I assume BSFC is constant and a constant 12.5:1 AFR.

For those interested, using Taz's curves I get:

RPM ____ VE

3000 _ 77%
3500 _ 87.5%
4000 _ 86%
4500 _ 82.5%
5000 _ 83.5%
5500 _ 79%
5800 _ 74%
6200 _ 69%

As I say, I question these results a bit as I think the power was falling off at high rpm more due to fuel ratio than VE.

Thomas, I agree with your comments about needing a better program to calculate VE. I use a spreadsheet I developed myself, and incorporate the stock intercooler pressure drop. I got similar conclusions to you regarding IC pressure drop and turbo efficiency.

tazman

The AVC-R helped a lot more then the cat back did with the spool up. The AVC-R was new to me at the time and since then I have ben able to get 16psi at 3200rpm and I think the boost drop was due to me not revving the engine that high often enough for the AVC-R to learn the proper duty cycle. I had the APE chips with an adjustable FPR and there was nothing I could do to get the fuel curve any better except program the AVC-R to hit 17psi then drop down above 500rpm to 12 psi and lower as the rpms increased. I should have some more data for you soon if you are interested with a bunch of different mods if I could get my car back together.