Engine simulators and head flows

[snowman]

Using the actual, measured head flows of both a 30mm port and 36m port, the exact same cam and compression ratio, and displacement, in a commercially availavble engine simulator program, shows that the 911S power is as advertized... But wait--!! using the lesser 30mm ported T head on the same engine gives YOU WILL HAVE TO READ IT FOR YOUR SELF power. And this type of engine simulator program is deadly accurate when used to compare power in this way.

THe first sheet is the measured air flows, actual, the second is the results of the simulation. The first is the 30mm ports, the second the 36mm ports. The last sheet is the general cam and engine specs.


Say what you will, these are the facts.

Now how much did you pay for that porting job??

[jluetjen]

Snowman;
Thanks for posting this. It's an interesting analysis and I've got a similar (abiet simplier) program that I sometimes play with (Dyno Shop V2.5.7M). I've found it useful to help to understand the trade-off's of different factors like cam timing and stuff like that. The problem that I have with both the program that I have as well as the results that you got is pretty simple. It doesn't coorelate with reality.

I don't know about other engine configurations (such as Detroit V8's), but in the case of the 911, the results don't match the output of the physical engines that we have in front of us. Unless you can input the specifics for a 2.2T and get out a very close approximation of the dyno readings of a real T, it doesn't mean much. If my memory of stats is correct, we should be able to run a "T-Test" (no pun intended ) comparing the results and get confirmation of the validity of the model. Just eyeballing the numbers, it's not a terribly good fit.

.................Porsche Published....... Dyno2000V3.08
Torque:...........130@4200 RPM..........181@4500
HP...................125@5800 RPM..........182@7000 RPM

So no matter what the computer says, it's wrong.

Why? A few things jumped out at me, some would suggest that your analysis should over estimate the power and others suggest that you would have underestimated it.

* You've listed induction flow at 800 CFM. Where did this come from?
* You have both runs using the T camshaft. If I put a T camshaft into a real S, I would expect the performance to be at the level of a T, and even a little bit worse at low rev's.
* You've spec'd the valve lift at 0.050 inches (1.27 mm) when Porsche cams are spec'd at .1 mm (0.0039 inches). So I would expect that you have in fact understated the valve lift.
* Looking at your lift quoted (11.43 mm) it looks like you may have adjusted since the quoted lift for the T camshaft is 9.8298 mm (.387 inches).
* The head who's flow you are quoting (30 mm ports) is in fact the head from a 2.4TK which used CIS injection. It is not the head from a 2.2T which used 32 mm intake ports.

At the end of the day, I don't believe that those differences account for the difference in output between your model and realilty.

Say what you will, these are the facts.

Yes it's a fact that your computer model output those numbers. No, those numbers are not the facts of a 911 engine. For those you need to look at real examples and right now they do not match your model very well. Once you get it to replicate a 2.2T accurately, you should be able to put in the spec's for a 2.0T and still have the results match real engine outputs. If it's good, it will also accurately predict the stock outputs of a 2.0 Normal and a the E's and S's too.

As such I don't believe that Dyno2000 V3.08 is an accurate predictor of a 911's engine performance.

[snowman]

Well for one it isn't a T camshaft, its an S "like" camshaft in both cases. I have put the T spec cam in and got guess what, 125 HP, very close to the actual number of 130 HP, a dyno test result. The point was to compare the HEADS not the cams ect. The intake airflow is arbitrary, I set it high as to not interfere, ie limit the response. The reason, you can put whatever intake and carbs on the engine you want, no limit on air flow. In practice you wouldn't go this large to get good performance, but large enough to get the same performance. Hope this makes sense to you. As to the accuracy of the program, I have found it to be within 5% for every engine I have ever put into it. BMW M6, Chevys, Porsche, Honda.

As to real 2.2 heads, I have previously flowed and posted the results, which are totally consistant with what I got this time. I have also previously posted the HP numbers, which were in very close agreement with the Porsche numbers, This is where I asked a question, I noticed that retarding the cam timing, using a stock T cam, increased the HP, a lot, all the way up to 150 HP with 10 degrees retarded. I have analyzed this result as well as I could and determined that Porsche simply tried to "detune" an S engine to come up with a T version. It was actually probably more difficult to detune it than enhance the performance. The T also used a cast crank, non shot peened rods, cast iron cylinders to try to lower the cost.

[jluetjen]

Ok Jack. I guess I'll need to wait for someone to get 150HP from a 2.2T by retarding the cams -- it's a relatively straight forward change to make. Now if you're saying that by putting a more radical cam into a 2.2T you could get upwards to 150HP, I would agree. What you're describing in many respects is a 2.2E which made 150 HP at 6200 RPM and 141 lb-ft at 4500 RPM. Sure the E had a higher CR, but near as I can tell, CR doesn't make a huge impact on 911 engines, although it will help to fill in the top and bottom extremes of the rev range when the cylinder pressures are low.

I'd also be interested in seeing an engine where someone's making 180+ HP through 30 mm intake ports, or 200 HP plus through 32 mm intake ports.

If it is so easy to do, why isn't the first and most common mod made to a T to just retard the cam and get 150 HP? The cost is almost $0.00. You would think that every T in the world would have this adjustment made to undo the detuning that Porsche made at the factory. It defies logic that everyone is missing such an easy tuning opportunity.

So in some respects we're agreeing. Imagine that! But we'll still agree to disagree on the ultimate limits that in many cases the head ports impose.

I'll just let it go now.

[1fastredsc]

Not to barge in on your discussion of early cars, which i know nothing of. But don't the late model sc's have the smaller 32mm ports, as opposed to the early sc's with 36mm ports, and make 180hp (fly) with the crappy exhaust system?

[jluetjen]

Yes 1fastredsc; I think that you might be right. The difference is the CIS system. I suspect that at higher rev's the CIS system acts as a restriction and as a result is pulling a vacuum. If this is true, then the whole equation changes because the speed of sound increases as the air pressure drops.

I don't have an SC nor any car with CIS, nor have I been able to find someone who will try to rig up a test, so there is no way to check my suspicion. If you have a vacuum guage, you can try it by hooking it up with a T connection to an existing vacuum port to the manifold or plenum. Take it out for a drive and run it up to redline a few times and see if any vacuum is detected.

BTW, didn't the early SC's have 34 mm ports rather then 32's? Or was that another typo in BA's book?

[1fastredsc]

I think the early ones have 36mm if i'm not mistaken, with the same size intake valves for both era's of SC's. That's why the early SC heads are the ones to get (or port to) if your seriously hot rodding an SC motor.

[jluetjen]

Correction -- Didn't the >>>Later SC heads have 34 mm ports rather then 32?

[1fastredsc]

I don't know, i don't have BA or WD book in front of me. I just remember that the early cars have the bigger ports.

[emcon5]

John and 1fastredsc,

According to both BA and Wayne's books, early SCs had 39mm intake ports, late had 34mm.

I don't have an SC nor any car with CIS, nor have I been able to find someone who will try to rig up a test, so there is no way to check my suspicion.

What do you have in mind for rigging up a test? I may be game.

If you have a vacuum guage, you can try it by hooking it up with a T connection to an existing vacuum port to the manifold or plenum. Take it out for a drive and run it up to redline a few times and see if any vacuum is detected.

John, not sure what you mean here. I know late SCs pulled a vac under load, the vac advance for the distributor. only works with the throttle open. Or am I misunderstanding you?

Has anyone ever punched 3.0 numbers into dyno software? I am trying to figure out a strange torque curve and would like to do a virtual port job to see if anything changes.

Do you need to be an engineer to understand it?

Tom

[jluetjen]

John, not sure what you mean here. I know late SCs pulled a vac under load, the vac advance for the distributor. only works with the throttle open. Or am I misunderstanding you?

OK, See -- you already know more about the CIS 911's then I do. What I would find interesting is to quantify the amount of vacuum across the rev range under load.

Here's a simple for example...

If a 911SC with CIS pulls .95 atmospheres (14.7psi * .95 = 14.0 psi) with the stock CSI and generates 200 HP, then the exact same motor using a mapped EFI system or MFI with no intake restriction should make (200 / .95 = ) 210.5 HP by virtue of getting a full 14.7 psi of intake charge at peak HP.

[racing97]

In the previous post may I ask what you mean by "strange torque curve"?

[adomakin]

I just wish a could make sense of all them sheets, then maybe i could join in the debate!

Andy

[350HP930]

Originally posted by snowman
I noticed that retarding the cam timing, using a stock T cam, increased the HP, a lot, all the way up to 150 HP with 10 degrees retarded. I have analyzed this result as well as I could and determined that Porsche simply tried to "detune" an S engine to come up with a T version.

While this may be correct about the effect such a change would have on power is such a thing really possible when valve clearances are taken into account?

I would guess that even retarding the cams by 5 degrees would be hard on a low tolerance motor.

[911s]

I believe the early SCs had 39mm intake ports, while the later SCs had 34mm intake ports (both had 49mm intake valves)