911 head flows

[snowman]

The following is the measured head flows of two differen't 911 heads. Both have exactly the same valves, ie 46 mm intakes and 40 mm exhaust. One head has 36mm intake ports, the other 30mm intake ports.

The first head, a 911.104.307 dated 8-69

The second head, a 911.104.302 dated 11-72

[snowman]

The setup to measure the flows is shown below



Note the high performance valve springs used for the test, thats the one on the right.



Note that the heads flow almost exactly the same up to 0.400 inch valve lift. This is very important, especially if you refer to the typical Porsche cam specs, ie they almost never are above 0.4 in lift so why bother???

[snowman]

The clay is used to smooth the airflow into the port. It in no way inproves the acutal head flow, it just makes it as though you had a pipe hooked up to it, as in an intake. Its necessary to get the actual performance of JUST the head.

Additional info on the heads:

The ports are differen't to a depth of about one inch into the head, after that they are identical, ie valve pockets, measurements at end of valve guide (41.3mm head 1, 41.3mm head 2, same for exhaust. In other words the heads are identical, except for the entrys of the ports and a short distance into the head. All that is needed to open up the smaller head is to match the port to the larger intake gasket, with a straight in cut.

Analysis of the performance of these heads on a program called Dyno 2000 will follow.

[snowman]

The flow bench, compliments of Saddleback college, Orange Co, CA.

[jluetjen]

Jack;
Something seems confusing.

The '73TK head (911.104.302.OR dated 11-72) with the 30 mm intake ports flowed 225-227 CFM at .5 inches of lift.

The 2.2S head (911.104.307 dated 8-69) with 36 mm intake ports flowed 193-195 CFM at .5 inches of lift.



BTW, here are the nominal port dimensions for the two heads which I measured prior to sending them to you.

Inches in
from Manifold
Face...................307.OR...............302.OR
0.....................30 x 30 mm .........36 x 36 mm
.5 inches..........31 x 31 mm .........38 x 38 mm
1 inch..............32 x 32 mm ..........34 x 34 mm
1.5 inch...........35 x 35 mm ..........40 x 40 mm

After 1.5 inches the valve guide boss begins to intrude into the port.

[thor]

I don't understand it seems tge 2.2S head is less performant

[BK911]

The clay is used to smooth the airflow into the port. It in no way inproves the acutal head flow, it just makes it as though you had a pipe hooked up to it, as in an intake. Its necessary to get the actual performance of JUST the head.

The bellmouth (clay) does makes an improvement to the airflow of the head. It significantly reduces the turbulence of the entering air. In some cases I was getting ~10% more airflow using a bellmouth.

I've flowed 2.2S and 2.4TK heads without a bellmouth at 28" Hg. My numbers for the S heads are very close to yours. But for the TK head, they are not even close.

By the way, I use a superflow too.

BK

[69911e]

Jack: I sent you an email about sending you 2 'E' heads to flow (2.0, 2.4), and have not heard back. Are you still interested in flowing more heads?

[jluetjen]

BK;
Given that the intake port would normally be attached to the intake manifold, wouldn't it make sense to always use some clay to smooth the entrance of the air. Better yet, for kicks why not hook up a weber carb with a stack and alternatively an MFI manifold and see what the end result is at the cylinder? There's never a situation that I can think of where a 911 engine would be running without something mated to that surface.

[garibaldi]

this is good information thank you for sharing it.
Have you ever seen a flowbench or a dyno win a race? When all is said and done, the performance of the engine is what matters. I see so many people put their faith into these testing tools, when what really matters is the way it performs when in its natural environment. I agree, and am jsut putting in two cents.

To further expand on your statement, a flow bench in this instance also does not take into consideration piston velocity, intake manifold, exhaust manifold, valve speed, operating temperature, scavenging effects, etc. All it does is test in a controlled and static state. Not to say that they are not useful to some extent, but there is a huge difference between hooking up a shop vac to your head and holding a valve at a particular lift and recording the number, and measuring the flow comparisons at operating RPM, in its installed environment. I have seen head that have given great numbers on a flow bench lose power on the engine and some of the rattiest looking castings make power improvements. So it jsut goes to show you that you should not be fooled into thinking that we have all of this figured out- we dont

[jluetjen]

Have you ever seen a flowbench or a dyno win a race? ...

Thanks Christian;
I just had the strangest feeling of Deja Vu -- like if I'd heard those words somewhere before.

Since you obviously don't think that there is any value in this thread I guess you don't have to bother reading any further...

[Scott Clarke]

Can we assume that the charts were reversed relative to the captions in Snowman's first post?

[ChrisBennet]

Here's Mototune's take on flow benching..

Excerpts:

It turns out that a flowbench measures the least
important aspect of intake cycle efficiency !!

It's true !!

When you stop to think about it, a
flowbench only measures the efficiency
of the "suction phase".

That's not the right
thing to measure.

Why ??
Because, it doesn't matter how well
the cylinder is filled at that point in
the intake cycle !

What ... that sounds crazy !!!
It's 100% true ... it's simply a matter of the sequence of events ! The success of the last event, the charging phase, determines the success of the entire intake process.

What happens when a low velocity port fills the cylinder really well, but too early ?? The result is a slower intake charge that stops flowing into the cylinder. Then some of the charge gets pushed back out of the cylinder and into the port as the piston returns up the bore during the intake charging phase. The gain in flow doesn't offset the loss in port velocity.

Here's a real "out of the box" idea: high flow ports also flow really well backwards !!! All that work on the flowbench comes back to haunt you when the piston now has even less resistance to pushing the intake charge back into the " high flowing port " !!!!!!! When you begin to consider the consequences of all this, the whole idea of "more flow is better" comes crashing down like a lead balloon.

What's The Secret ??
Using the race analogy, if you increase the Port Velocity, by making the port smaller, the intake cycle "race" starts out slower, gains momentum and makes a tremendous charge on the "last lap" to overtake the high flow port and win the "race". The interesting thing is, this type of port will always lose in a flowbench contest !!!

[jluetjen]

I don't know about anyone else, but I'm safely past the "More is better" phase. By the same token, too little won't work either. There are limits such as the speed of sound that limit how fast you can flow your intake charge. Designing in an intake where the charge goes sonic at 1000 RPM (or even 7000 RPM) is not a recipe for success.

There is a "happy medium" and the question is where is that "happy medium" and why is it there rather then at some other number.

Now if all of you nay-sayers would get out of the way, we're trying to increase our understanding here!

[garibaldi]

What are you getting all bent out of shape for? No one is bashing or "Nay saying" your thread. So relax. Like I said, good info. I will definetely question its valvidity though- all of this testing is done in a controlled environment in a room, on a flow bench and if you are going to argue that all of that doesnt change once that same head is on a motor and operating, then go ahead. I question how valid the numbers that your flow bench will yield once you are in operation? My expereince is that it doesnt matter much. I am not trying to discourage you from experimenting and trying to learn, by all means! But the reality is that at the end of the day, the head is going to be functioning on an engine, with manifolds, and a cam, at operating speed, and with temperautre, with varying throttle positions, and varying loads, with varying injection cycles and methods and so on , and so on, all of which do not get considered, and all of which WILL have an effect on the "flow" numbers in and out of a given head. It was stated earlier in the thread- what was the flow difference when a bead of clay was applied to the perimeter of the port opening? That was a strip of clay, what happens when you put the motor together and run it- how does that effect flow? A bench cannot tell you that. Thats all I am saying- again, its not bad in my opinion, I just put my faith in another number.

[jluetjen]

Christian;
If you do a search of this BBS for "porting", "harmonics", "cams" and such you'll see many of the subjects which you mentioned examined and discussed. We may not be PhD's, but I think that there is a lot of shared learning going on. On this particular thread, we're looking at head flows. It's agreed that it is not the be-all end-all of HP, but it is one component of the equation -- and one that is definitely not well understood by most of us who are spending the money for improvements. So in my case I'm trying to push the boundries and learn what I can from everyone else so that I can make more informed decisions about my 911's modifications.

[KevinG]

The Wizardry / Alchemy / Artistry of porting, polishing and flowing heads is way beyond my ken, but I can visualize quite a bit of what is presented in this thread. Please keep up with the experiments and presenting the data!

Is it even possible to use the bench to check the changes caused by using MFI components (stax & tb's) vs. equivalents for carbs? That could prove interesting as well.

[snowman]

The 36mm port flows the most, ie 225 cfm, the smaller one less. I have corrected the data sheet lables to reflect the correct head numbers. THEREFORE IGNORE ALL SWITCHED HEAD REMARKS.

As to the value of flow testing heads, it IS what wins races. The air flow is directly proportional to the power the engine can produce, more airflow, more power. In fact, since we already know that the desired Air fuel ratio will be about 12.5 to one we can calculate the exact ammount of power that can be produced. However there is an OPTIMUM size port, ie you can have to much flow, and to little velocity. You have to know this to make a good flowing head that makes good power. But given a specific design, and going racing usually means that you are going to spin it up as high as possible to get all the power possible, consequently you will need much more flow than a stock head, You also don't care if it runs below 4000 rpm. Not so street friendly, but a race car. for a street car porting usually means matching the gaskets, and maintaining the general size and shape of the ports.

And yes you can hook up the entire intake and exhaust system and flow test it. You start, one piece at a time, with the heads. If you can't get the air into the head you don't need to flow any better anywhere else. ONce you have done this you add each component, and test it individually, then togather. How accurate is the power calculation using the head flow, it can be withing 1%, which is pretty darn closel All other things, such as ram tuning can be predicted with similar accuracy. Cam timing is critical but a seperate issue.

By the way the similar flows at lower valve lifts make perfact sense. At lower flows, the intake valve is the limiting factor, ie the port is bigger than needed at the low valve lifts. Only when the valve is open far enough for the port itself to become the limiting factor is there a difference in flow. Note that the valve spends most of its time at lower lifts, and very little at high lifts, consequently the heads will perform nearly the same. In fact the head with the smaller port will produce better low end torque, due to higher intake velocities.

And for referance I measured the intake flow without the clay, its measured at 0.5" lift and is on the right side of the data sheet.

[jluetjen]

Uh Jack -- I think that you did.

[snowman]

Originally posted by jluetjen
Jack;
Something seems confusing.

The '73TK head (911.104.302.OR dated 11-72) with the 30 mm intake ports flowed 225-227 CFM at .5 inches of lift.

The 2.2S head (911.104.307 dated 8-69) with 36 mm intake ports flowed 193-195 CFM at .5 inches of lift.



BTW, here are the nominal port dimensions for the two heads which I measured prior to sending them to you.

Inches in
from Manifold
Face...................307.OR...............302.OR
0.....................30 x 30 mm .........36 x 36 mm
.5 inches..........31 x 31 mm .........38 x 38 mm
1 inch..............32 x 32 mm ..........34 x 34 mm
1.5 inch...........35 x 35 mm ..........40 x 40 mm

After 1.5 inches the valve guide boss begins to intrude into the port.

I measured the port at the end of the valve guide, which is where the port turns, and got the exact same size on both heads at that point, 41.3mm. Of course the valves are identical.

at o depth intake 29.6mm vs 35.8mm and exhaust 33.3mm vs 34.9mm

at a depth of 18mm into each intake port I got 31.9mm vs 37mm

at a depth of 18mm into each exhaust 33.8mm vs 35.1mm

If one were to use a drill and went straight into the port it could be opened up to almost identical dimensions to the larger ports.

You will have the heads in a few days, please verify part number vs size so the data sheet can be corrected if necessary.