Quietboom;
Given the numbers that you supplied I've plugged your engine into my analysis and came up with the following results. I've assume a 6600 RPM peak HP engine speed.

Flow at maximum lift: 315 CFM or 8920 liter/minute
Time at full lift: ~ 4.20875E-05 seconds
Avg flow while valve is open at 6600 RPM: .75 liters
Cylinder capacity: .83 liters.

I suspect that you are running a high CR of at least 10:1, is my guess correct? I also suspect that my baseline analysis doesn't take into account what I suspect to be some very significant overlap in your cam spec. Would you be willing to share that?

CR was 13:1. I don't remember the exact cam specs since it was a custom Lunati grind and I'd have to dig up the cam card. Going on memory lobe separation was 108 degrees, cam was timed straigt up, and the overlap was over 40 degrees, basically it was the biggest flat tappet cam I could run before moving to a solid roller, again this is going on memory from 5 years ago. I used a ported Edelbrock Super Victor intake, custom 2" headers with 4.5" collectors. The car was a T-top body with a full cage so even stripped to the max it weighed in at 3200lbs. Best MPH was 132 and best ET was a dismal (due to traction and other issues) 11.01. The trap speed should have put the car solidly into the 10s like 10.5 or better but it was a real bear to hook up on 10" slicks and a 4spd. I shifted the car between 6800 and 7000 so your estimate of 6600 is pretty good.

As for the discussion of V8's I brought it up because airflow requirements for a given HP are the same for all engines and I just wanted to point out where I had some experience. IF I remember correctly 315cfm should have been more than enough for 600HP in theory but in the real world it rarely works that way. A quick look a 911 ports tells me that they are already a great design and snowman is right, you're not gonna get the same results porting them as you would a stock V8 head. I'd say they are more closely matched to high end aluminum V8 heads, gains can be found but with dimishing returns.

The most important thing I've found in engine design is properly matching everything, choose an rpm range and HP target and size everything towards that goal.

We're in agreement on that.

If you're shifting at 6800 - 7000 RPM, it looks like you're doing better then my model would suggest. I suspect that it is due to the large amount of overlap giving the air flow a "headstart" versus what the duration would suggest.

... I would add in an 8 cylinder engine. We're talking about a 6 cylinder engine and so right off the bat you are losing 25% of your airflow, so now we're down to 450 HP from 315 cfm per head. If we are getting 200 cfm from T/E heads with 32 mm ports (a big if), you're down to 285 HP. This is where the wheels are falling off of the wagon. I have yet to hear of anyone getting anything close to that sort of HP from a 911 motor using 32 mm ports, or even 36 mm ports. So the port size can not be ALL marketing.

200 CFM through a T/E's 32 mm ports sounds high. A 351 Cleveland V8 example that I've seen (2.041 inch ports or 51.8 mm's) shows stock port flows of 182 cfm at .550 with an improvement at lower lifts after porting. A different example with 1.655 inch ports (42 mm ports) showed peak flows of about 151 cfm at .600 inches. Both of those ports I assume are rectangular and are quite a bit bigger then 911 ports, but it doesn't make sense that the 911 port flows that much more then them when you could fit a 911 style port inside either of those examples with room to spare. If that were really the case Detroit would most likely have done that at some point during the 50 year career of these motors.

So, I guess this whole excercise falls apart for one or more of the following reasons (or maybe some which aren't mentioned).

1) HP does not coorelate well with head flow (at least not linearly).
2) The 600 HP is not a real number.
3) 32 mm T/E heads do not really flow 200 CFM
4) You can't compare V8 style heads with 911 style heads because the flows (and resulting combustion) have completely different charactoristics.
5) ...?

Either way the model does seem to fit reasonable well within the set of 911 motors. I do question if Snowman's flow numbers are somewhat optimistic. I also suspect that in fact the engines will be ingesting 100% of their volume closer to the peak torque engine speed rather then peak HP engine speed. Oh well, back to the drawing board...

Yup, I should have stated 315 per cylinder. The flow requirements of a 6 would be higher. 600HP IS NOT a real number, it's derived from MPH in the quartermile vs. vehicle weight for RWHP, I multiplied by 1.1 for an approximation of the flywheel HP. Now that's the math, here is what an expert told me after I twisted the crankshaft and I quote "you need to make in excess of 650HP to twist that crank 4 degrees!" I didn't believe it then and I don't believe it now. I have to assume that as HP and torque go up that the driveline loss as a percentage gets smaller, add lightweight or high performance diveline components I it should go down even more. Basic phyics says that I'd have to make 575HP at the rear wheels to hit 132MPH in the 1/4 mile so lets assume it's just 600 at the flywheel that's roughly 1.5HP/CI or about 88HP/liter which would be roughly 265-270HP for a 3.0 911. Now 1.5HP/CI naturally aspirated is nearly impossible to street drive, perhaps a high tech engine management system might make it better but still it would be finicky. Add to that the fact that drag race V8s have two advantages over 911s from a power output standpoint, the first being water cooled heads which allow higher compression and second they are only run for a short period and therefore do not suffer from heat soak on the intake etc.

Given all that information I'd guess that 1.25HP/CI is a more realistic goal for a street driven vehicle, that's about 225 from a 3.0 and should be a real blast, whether the heads need ported to get there or larger than 32mm are required I really don't know. The truth is cam timing usually has a larger influence on power and drivablity that port size and volume, it is possible to undercam a large port for drivability but that doesn't make sense. My best advice would be to maximize the ports you already have for a reasonable power level, cam it correctly and install a good exhaust and intake.

Remember the 911 engine is a detuned race engine, not a V8 grocery getter, there are no huge gains to be found.

One question, the 2.7 RS made 210HP, I don't know if that was net, gross or DIN. What size ports and valves did it run? How about cam timing? I calculate about 1.28HP/CI for the RS or 77.7HP/liter that would be 233 for a 3.0 or 187 for a 2.4, seems to me just copying what the factory did would be easier.

DIN

The same as the 2.2S and 2.4S. 36 mm intake ports and 46 mm intake vales.

The exact same as the 2.2S and 2.4S.

John, just curious if you have done any number crunching on the late SC 34mm intake ports?

Tom

Argument over! LOL. Seriously from all I've read the RS was a very enjoyable car to drive so obviously it's ports and cam timing are not too radical for the street. The only other port solution that makes any sense would be a professional port and polish job that would match the flow of the S heads while also improving velocity i.e. slightly smaller but smoother ports. I had a chance to ride it a 2.4S and it was nice and probably was close on HP to my 3.0 but it definately didn't have the bottom end that my car has.

My experience with poorly flowing ports is that you have to "overcam" to get real power with them as opposed to a more mild cam with great ports, the latter has more usable HP and torque. This is why I suggested maximizing existing ports. I guess it all depends on your version of the word "streetable".

Lastly I would also be interested in 3.0 flow numbers as well as 3.0/3.3 turbo flow numbers since I'm considering low compression JE pistons (yes they will make them for me) and either a turbo or supercharger in the future.

Quite and Emcon;
You'll need to do your own work on the CIS and Turbo models because the situations are completely different in both of them. This should be a new thread for each of them.

The CIS (I believe) tends to run a slight vacuum because of the obstruction from the air meter. At peak HP I suspect this gets worse. But I don't own a CIS 911 to test this, nor a CIS anything. it would be interesting to hook up a vacuum meter and track the vacuum versus load and throttle position. It would be awsome to log it with a data logger. At a minimum it would be interesting to do a couple of acceleration runs in 2nd and 3rd gear to see what happens. Anyhow, if the pressure in the intake of the CIS cars is below ambiant, that is a whole 'nother can of worms.

Turbos on the other hand run pressurized intakes and the intake size and shape affect the turbo's reaction.

In either case I don't have much data and so I don't have much to go on. Given that, you won't get much from me beyond some vague guesses. If you'd like to collect the data on the other hand...

I forgot to comment on this earlier, Flow at maximum lift was not 315 but closer to 326 (flow at .700", the cam opened to .690") The cylinder capacity vs. valve flow doesn't seem right, a high performance engine will have a VE above 1.0 and as high as 1.5+, this is accomplished with proper exhaust and intake sizing and runner length or "tuned" ports and overlap for cylinder scavenging. It is common on high powered NA engines form them to naturally "supercharge" themselves by taking advantage of pressure pulse waves to aid cylinder filling and scavenging. As you mentioned a long overlap can allow the exiting exhaust gases to pull some of the intake charge into the cylinder and thus VE's in excess of 1 are possible.

Another thing I forgot to mention is that theoretical flow requirements can be misleading, the more important flow number would be the entire intake system as well as the entire exhaust. In other words the heads and cam may be capable of making X horsepower but are limited by the rest of the engine, this is why I stress looking at the engine as a system. For example a 40mm carb is limited at some point on a 911 and you must move to a 48mm. What size injection runners did the RS have?

Whoops I was writing as you were posting. Anyway I was just curious about flow rates of various 911 cylider heads. I've had this wacko idea of using 2 small 4cyl turbos on custom headers run into a sds or megasquirt controlled sheetmetal intake. I'm a pretty good fabricator so this would be a fun project for me but without knowing the flow of the heads I don't know if adding turbos would be worth it as opposed to just buying a turbo engine and selling my 3.0 to offset the cost.

I used to have a chart somewhere that related intake port cfm to HP but I can't find it. I do remember that I selected the porting for my drag engine based on this formula and it was just about dead on. It's been years since I've been in the drag racing loop so I'll have to dig up some old friends and see if they know.

Circular ports are infinately superior to any rectangular port in terms of airflow for a given cross sectional area. Air flowing is a lot like water flowing. Ever wonder why water flows so much faster out the drain once the swirl forms? The math is horrendous, but can prove the point. Same is true for microwave wave guides. Circular polorized waveguide is much much more effecient than rectangular guide, should be the same for any other fluid in a pulsed or periodic system. Bottom line, you cannot compare chevy rectangular head ports to round Porsche ports, directly.

In any case air flow is air flow, easily measured, on an air flow bench. There is no magic, if you get enough air, then all you have to do is to add fuel in 13:1 proportion, no magic there, and you get the corresponding power.

In a fixed size engine, given the displacement, you get such so much air each time the piston goes down, to suck more air you just do it more often, ie higher RPM. The limit is either the head or intake tract, or the basic displacement and RPM. IF the head isn't limiting the flow, then it may be other things, ie the carbs, the intake runners, the air flow meter, whatever, or the RPM. That simple. In the case of the 911 heads ITS NOT THE HEAD.

You question the air flow measurements. OK whats the basis? Do you have other head flow measurements, are they with a commercial air flow bench? Do you have more than one? Do you have the math to state that it isn't possible? IF so then I will be challenged to look into my measurements further.

jluetjen

I don't want to hijack the thread, but in context of your comments on flow rates, valve overlap and duration, I have a question on cam timing. On a 3.0 CIS motor (78 Euro SC), are there really any perceivable differences between the cam timing specifications for the different production years as far as peak HP or torque range goes? I have to install Carerra chain tensioners and will reset the cam timing, so I have an opportunity to experiment. What does your analysis predict, and can you offer any insights? Common knowledge suggests just setting the cam timing midway between the early and later years' advanced/retarded settings, but it would be interesting to see if your calculations show up anything else.

If this subject has been beaten to death in other threads, I will apologise and disappear into cyberspace.........

Thanks

LeRoux

Porsche played around with moving the cam timing around on these 3.0L engines, moving it and then moving it back. I don't think there's any large, hidden HP in changing the cam timing, but then again, I haven't dynoed all of the range...

-Wayne

Quiteboom;
I would suspect that you are getting a VE of greater then 1:1, but most likely at your peak torque engine speed, not your peak HP engine speed. If you are thinking about fabricating a Turbo inlit, the parameters are going to be completely different from NA inlets. To be honest I don't know enough to give you an opinion.

Snowman, The 911 head flow data that I have is what you published and what BK911 published. Unfortunately they don't match too well as I mentioned although BK did test some of my heads and some of his, and the data was consistent between the two. I gave you my calculations already. But if you don't have anything new to bring up don't bother researching it further, I'm willing to let it drop at this point.

Rule of thumb for advancing or retarding cam timing:

1 degree advanced moves the powerband 100rpm lower in the rpm range, 1 degree retarded moves it 100rpm higher. Advanced is typical for a street vehicle and many cams for cars other than 911's have 3-4 degrees advance built in.

John,

THis head flow data shouldn't be a big secret, Hopefully some others who have access to a real flow bench, can pitch in with their info to the benefit of all.

A limited time offer. I will flow test and publish the results of anyone who is willing to lend me their head for 2 weeks. It must be done before the end of May 2004 or waiting until Sept of 04 when I can again access the equipmet. Cost, nothing. I will post flow results, along with a photo or two.

All I would need is your head and an intake and an exhaust valve. I will test it and return it with data withing a 2 week period.

Send me a private e mail if interested.

I just pulled apart an 2.7 RS spec motor and was surprised to see that the intake ports were a whopping 41.5 mm. If I recall correctly, the exhaust ports were 35mm which was the stock size for the RS. The prior owner absolutely raved about the motor and indicated that the 3.6 he installed didn't have as much torque as the RS (maybe he has a tired old 3.6?) Anyway, the porting job on the intake is beautiful and I want to preserve it for possible replication in future rebuilds by creating a mold. It would be interesting to assess the flow characteristics of these heads. The heads are at German Precision right now getting a full valve job. I'll contact jack to determine if his plate is full yet.

Not yet.

I just received 2 heads today and will be flow testing them first thing after spring break. I can do maybe one more that week (mar 21-26) After that it would be the last week of April before I could do another, none in May thru first of Sept. We will see after that

I will send the results to the owner of the heads first, wait for a response and post pictures and results when I get confirmation that the owner has received his results. Not asking for permission, just a courtesy.

I have received the heads, see photos below. Note the big difference between S and T intake port sizes. Note also this the ONLY difference, ie same valves same port sizes, except at interface.

The Superflow bench that will be used to measure the exact air flow of each head.


http://forums.pelicanparts.com/uploa...1079508292.jpg http://forums.pelicanparts.com/uploa...1079508306.jpg http://forums.pelicanparts.com/uploa...1079508323.jpg http://forums.pelicanparts.com/uploa...1079508336.jpg