Home porting? Like Home perm?

[dd74]

Originally posted by snowman
You may not get absolute flow numbers, but you will find out if you made it better or worse.

Yeow! But by then it might be too late, right?

[Steve@Rennsport]

I've been following this thread from the beginning and its been a dandy,....

Since I've been doing this for a living (building 911 race engines) since 1976 and have a LOT of time at our flow bench, I'd just offer that its far easier to ruin a set of 911 heads than improve them, regardless of how large or pretty they might look.

I would also caution anyone not to draw too many similarities or conclusions between 911 heads and cylinder heads from other engines. There are far more variables at work here than meets the eye,...

[Wayne 962]

I'll echo Steve's comments here. I'm certainly no expert on head porting, but I did interview many experts in the Porsche field, and also did quite a bit of research on the subject. The Engine Book is light on head porting info because I couldn't really narrow down useful information to put in there. Flow rates for all heads weren't available to me, and the "black art" of head porting seems to be a well-kept secret that not too many people are willing to share (Walt didn't want to talk about his particular techniques on the record).

For those who need their heads ported, it's not really the actual porting that is tough - but knowing what and where to cut and remove. Just like an oversized exhaust can kill HP out of an engine, too big of a port job can also remove HP when you're actually trying to increase it (running too large of a carb too can cause problems). Everything has to be balanced with itself.

For me, I would never do my own porting - maybe some polishing, but not porting. There's too much risk in messing up a set of heads to make it worthwhile.

-Wayne

[jluetjen]

Lest people be confused about my intentions, I agree that guys like Steve and Walt have certainly learned a lot more about porting then the rest of us will have the opportunity to understand -- especially as Steve has said by actually experimenting with 911 heads. They've learned what they know through hard work. I certainly agree that if you want top shelf head performance for your race car -- I doubt that you could do better.

By the same token, any product or service which is shrouded in secrecy risks being polluted with "snake oil". Buyer beware! I am of the belief that a lot of people spend a lot of money getting their heads "ported" when they don't have a clue what they are getting or if it will even do them any good. There are a lot of people willing to take people's money to run a grinder around the port a couple of times and return a "ported head -- worth a gazzilion HP" when in fact the changes made have not improved the performance of the motor, let alone degrading it. Now if people want to spend their money that way because it makes them feel good -- go right ahead. I just figure that public discussion and sharing of data on the subject might be worthwhile to people who have to invest their money wisely.

[dd74]

At least with CIS engines, I believe it was Steve who said it makes a very small difference in performance when the smaller diameter heads and intake ports of the later '80-'83 U.S. 3.0 engines are ported to the larger diameter size of the '78-'79 U.S. engines or those of the ROW 3.0 engines. YMMV.

[William Miller]

John, haven't been here for a while.
Our Spam filter threw the notifications into Quarnteen. I was wondering what happened. It was like the PP board went dead.

Anyway there are many differences from the 83 US and the 83 US CIS.
The most obvious is that the US was then K-lambda with the o2 sensor and computer. ROW kept the same basic design as 79US but they continued to tweek it over the next few years.
K-lambda air meters have a cone with a single plane. ROW is stepped as most of the metering is done by the shape. It's shape is steep near idle then flattens during light acceleration and then steep again at the top where the plate rises whit higher air flow. This mixture is richened for acceleration by a vacuum mechansim in the control pressure regulator which lowers control pressure.
In the later US CIS K-lambda the fine tuning of metering is done by the computer that effects the pressure in the bottom half of the fuel distributer. The control pressure regulator was then simplified and the cone in the air meter is simplifies and does not have the steps.

I know there is a significant difference in the flows between the two as well as the 79 US. Although the 79 is simular in design the Fuel distributer does not flow as much. How do I know? When I bought my car it had the wrong FD on it it was the US 83. At the time I could not find the 83 ROW FD so I replaced it with the next best thing. 79 US FD.
This worked better than the 83US, but still ran way too lean at mid range flows. (I attemped to adjust this by lowering the control pressure, this is a little known fact that it can be done.) Even lowering the control pressure did not allow the 79 to flow enough to meet the demmands of the 83ROW air meter. My conclusion (Untested) is that the 81-83 ROW FD and Meter do flow more than both the 79US and the 81-83 US.

The 81-83 ROW maintained the larger air box (Including the cold start valve manifold) and the larger intake runners and my assumption the larger intake ports. My guess is there was a good reason for doing this.
Performance is my guess.

I do agree with the experts that maybe there is litte gained by attempting to get increases by just porting the heads, but I think there was a performance advantage in the complete system used on the 930/10 engine over the emissions restricted US 930/16.
Can anyone please confirm the port size on the 930/10 engine? I just have to know.

[Scott Clarke]

I must say, yet again, that the question that started this thread was (and remains): Is it possible to COPY what Porsche has done and is known to work? If I want an S motor, then S ports should be just the thing. Given a physical model, it seems to me that a copy could be made without a flowbench, PHD, or eye of newt.

[jluetjen]

Scott;
I think that you will continue to get a split decision on that. I know that I've shared what I know for data. So I guess the decision is now yours.

[William Miller]

It's your quest. It looks like it's possible.
Once the metal is on the floor, let us know how it went!
Good luck, I really hope it works!

[dd74]

Originally posted by William Miller

Can anyone please confirm the port size on the 930/10 engine? I just have to know.

Wasn't it 39mm?

[Wayne 962]

My book says 34mm intake, 35mm exhaust.
-Wayne

[dd74]

Originally posted by Wayne at Pelican Parts
My book says 34/35.
-Wayne

Wait, 930-10 is the late '80s U.S. engine, right? If so, yes, it's 34mm.

I thought he was asking about the ROW late '80s 3.0.

[Scott Clarke]

I have a plan! I'll find a junk T head and a S head to use as a model. I will port the T head, and have both flow tested. Either it will work or it won't. This scheme allows me to gracefully bow out if the going is rough, or if I get poor flow. Anyone have some junked heads?

[snowman]

Originally posted by dd74
Yeow! But by then it might be too late, right?

No, because if you follow all the advise you will learn that you only make small changes, slowly, so any downside will be small. Also air flow is air flow, and all cars are similar in this respect. What you may find on an english MG does apply to a Porsche or Ferrari if its the same situation, eg port mismatch. ( The Ferrari ports will not likely have this problem to begin with but if they did...)

On T heads, I would suggest flowing them first, you will most likely find you don't even need to port them! (unless your building a race engine that turns over 8000 RPM)

How do we know this? The air fuel ratio is well known, for max power on a fuel injected engine, its about 13 to one. That known, the engine displacement known, the max RPM known ( you know this answer based on your budget and what you are doing for rods, pistons, valves....and almost certainly it is less than 8000 RPM for a street car) and you know EXACTLY how much air you need. If your heads flow this much air, your done. Knowing the HP produced by this quantity of fuel will also give you an almost exact ammount of HP you will make. A program like Engine Dyno 2000 is most usefull for this kind of thing.

To many people hide behind the smoke and mirrors and snake oil used to sell this kind of stuff, its not magic, and once you look close at it, not all that difficult to understand, after all even an automechanic can understand it. ( Slam was toung in cheek fellas)

Those that take the time to investigate and learn this stuff will be amply rewarded, the rest will continue to pay thru the nose. And in the case of head porting you may NOT get what you pay for.

I am going to give one of the porting secrets away now. Even though they make lots of bucks doing it, the people that do it, and are very good at it, and give value for the buck, HATE doing the job, probably even more than cleaning the bathroom, even for all that money.

[911s]

Ok, if porting is viewed as a sin, but port matching is perfectly acceptable, is it possible to match the inake port of E head to that of an S? Or how about matching the port of E head to that of an S manifold? Would this be considered an acceptable practice that stands to gain more than it can damage? Afterall, all I'm trying to do is match the port to the 36mm size.

[snowman]

Mismatched port sizes are always bad. An absolute mininum would be to match the ports at the junction. A better solution would be to use the same size ports thruout. If you want to see what your ports look like make a silicone mold of them like this:

This is a full race intake port for a Porsche 912 or 356 engine. It flows 195 cfm at 28" of water at 0.5" valve lift Thats enough for almost 200 HP at 8000 RPM for a 1725cc engine.

This kind of porting isn't for a street engine, unless you don't care if you can't use your engine below 3500 to 4000 RPM.

To answer your question directly, DON'T mis match ports. Use the same thruout and don't do it yourself, unless you have a flow bench, homemade or otherwise. The S ports are 9 tenths myth, and one tenth performance. Trying to make an E port an S port, will most likely yield worse performance, even if you do it right. Street perfromance depends on the low end, not the 8000 plus performance.

additonal info, it would be better to match a S manifold to an E head as the port velocity would increase, rather than decrease as it entered the cylinder.

[jluetjen]

additonal info, it would be better to match a S manifold to an E head as the port velocity would increase, rather than decrease as it entered the cylinder.

Jack, that's a nice opinion, but it flys in the face of each of the the 911 port designs that I've looked at, plus the 912 head casting that you showed. My advice is to be very careful accepting "conventional wisdom" at face value when it is contrary to facts. I wasn't there, but the data suggests that Porsche spent quite a lot of time getting their head and port designs right and optomised for the application.

Note that every stock head that I've measured (and the 912 head that you showed) gradually expanded from the manifold face to the bowl behind the valve. Everyone who suggests a straight plunge (parallel port walls) or walls that converge at the bowl (as you describe) is suggesting something very different and apparantly feel that they know something that Porsche didn't. Or maybe they are just missing some subtle points which Porsche knew. Why did they do that? I don't know but a couple of SWAGs are...

a) They didn't want the velocity to be so high going around the bend in the bowl
b) Cross Sectional area. In order to maintain the cross sectional area in the region of the valve stem and boss, the overall diameter needs to be larger.
c) Flow. If you are going to expand the diameter, you can't do it suddenly lest you stall the air near the boundry layer, so a gradual change in cross section is required.
d) I've got a couple of other hair brained ideas too which I'll keep to my self lest I confuse others as much as I've confused myself.

Check out the video(s) in this post to get an idea of the types of things going on in the ports in regards to pressures, speeds and timing.

[Grady Clay]

I’ll third Steve and Wayne. John and Jack have great information.
Bigger is not always better, at least on ports and associated plumbing.

Remember, the flow bench measurements are static, continuous flow and are indicators. The air/fuel flow in an operating engine is discontinuous pulses. Does the “closing of the valve” pulse reflect off the transition from the intake runner to the airbox? What are the intake reversion issues? Additionally, the CIS (kontinuierlich) continues to flow fuel from the nozzle even when the valve is closed and there is no air flow. Where does that fuel go and how is it re-atomized?

Who has insights on the dynamics of the situation?

Best,
Grady

[Grady Clay]

Everyone should look at and study John’s animation at:
http://forums.pelicanparts.com/showthread.php?s=&postid=1143536

This gives some idea of the complexity of the dynamic situation.
Good post John.

Best,
Grady

[jluetjen]

BTW Grady, I can't take credit for the automation. Only finding it on the web after reading the article in "Race Tech Engines". I'm not smart enough to have figured it out in the first place.