how many different 'best for the application',
port shapes, can there be?

with a given mechanical interfvce,
which implies a given cross-section, at the intake manifold.

it seems like '''GOODNESS''' could be seen from...
a family of curves...

airflow v. pressure drop, at several pressure drops,
up to at least Mach 0.5 airspeed.

with all that said, the more you pay, you should
get closer to exactly what is wanted.

...consider the Gen 1 SBC.........
ok, we got the intake valve in a certain position.
a dogleg to change from 23 degree to whatever.
other minor doglegs for... position, bolt holes, pushrod holes, etc.
I just don't see a billion variables there.

I kinda assume that most of the 'tuning' can be done outside
of the cylinder head. and that the 'sonic choke' is in the
external tubing,

compare to chosing the hull for a ship,
with everthing else given, you got two choices,
max speed with max engine effort, or,
better fuel economy at less than max engine.

seems to me ther would be a similar choice for cyl heads,
if a given design has best flow at a certain pressure drop,
it will not be the best choice to flow the
max at less pressure.

 

ContactPatch

I was waiting for someone to bring up those points.

The engine is a dymanic machine not a static machine but we do a lot of assumption based R&D with static oriented results. Most of us use the 28" of H2O because it is the standard by which certain flow-bench related results can be transfered into real-time dyno data. Below that 28" and the data results get skewed and it becomes difficult to identify changes in port configuration.

Reality states that it is the port velocity that creates the power curve so you have to know the cross-sectional area in order to develop that data. Now where is the engine RPM/Power going to be used at and for. This determines what you have to do with the overall head package.

We have determined over years of testing that the best torque will occure at around 205 FPS air speed, while peak power will occur at around 220 to 240 but not to exceed 260 FPS. While I have simulated many heads by calculating this data the best place to have it occur is where the manifold interfaces the head.

Next you have to realize that the engine only sees 3" of H2O, and then you have to calculate the amount of time in degrees that the intake can actually fill the cylinder, given ideal conditions.

You can see that there are a lot of varibles to conted with other than the ones you mentioned.

Denny Schmidt

 

Denny, thanks for reponding to my post.

consider the cyl-heads for other engines for cars, motorcycles, lawn mowers, that the engine does not have the complication of the 23 degree related dogleg.

what design considerations, are there?

somebody chooses the cross-section and length of the air-system,
the intake valve needs to be big enough.
the cross-section-area should match with the intake manifold.
no casting flash, mold seams, or other wierdness.

.........and the the SBC adds the complication of a sharp turn,
so that there is now a 'bowl area' of larger cross-sectiom.
............
or where are there other fundamental decisions?

 

Contactpatch
Considering your comments

The heart of any engines power curve is Intake velocity @ RPM = torque/Horsepower.

Torque is going to be at 185 to 205 FPS @ rpm
Horsepower is going to be 220 to 240 FPS @ rpm not to exceed 260 FPS

Here's what we discovered in designing ports for turbo diesels. In order to make a turn in directional air-path you have to fool the air into thinking that it is not turning. We can do that by squeezing it like you would flatten a round tube to make it an oval. Even though the cross-section remains the same the velocity changes somewhat and the air thinks its still going straight.

When the 23 style head is ported correctly that bog-leg at the push rod turn acts as this squish area helping the air to make that turn. The alternative to that, is to move the push rod out of the way and streighten out that area. Both seem to work equally well.

The way I calculate the the velocity is to use my formula to find out the gas-speed for the above at the valve. Compute the area of the valve and multiply that by 80% and that becomes the area at the mouth of the port where the manifold mates to the head. Then I recompute my velocity number for that area to see that I don't excced the 260 FPS limit.

Hope that helps you

Denny Schmidt

 

You have any experiance with 7.3L powerstroke heads?

 

Not much - I can find out - if you really need to know something.
Scotts machine shop does all the engine work for the Fayette County Transportation and school district and all they have is Fords.

Denny

 

Denny, thanks for your insight.
..............................................
OK, what I don't like is,

XXX Cylinder head company offers , 180 cc, 195cc, 215cc, intake port
volumes,

is the torque at 2000 rpm going to significantly less with
the 215cc v. the 180cc?

keep in mind only three inches or so of the intake track are in the
cylinder head,

 

I was just curious because we have done a lot of R&D on those PSD's for tractor pull competition...compare notes on swirl and porting those things...don't know if you have looked at an intake port on one but it is a nightmare.

 

MachinestOne

Actually from what I could see when the guy next door took his apart they are fairly straight forward - Not nearly as bad as the Dodge. I did a Dodge head for one of my customers but they never got a chance to run it on the chassis dyno so I don't have any real data. I'de like to see what can be done to the I.H. engine though. Can you supply some pictures of the I.H. head??? What kind of power does that Powerstrole make on the dyno??

Swirl seems to be very important in deisels - the Dodge head has purpose built swirly intake ports very similar to that strange "Vortech" Chevy head with the wing in the bowl area. What we have done to some of them is build a combustion chamber in the head very similar to a 348/409 Chevy. The Dodge Seems to like that, but you have to readjust the injector depth or go around it because there is some power in finding the correct depth.

ContactPatch
The cc volume of the runner in the head is actually meaningless, because it's the velocity that controls the power curve. Consequently you could have a 315cc port that had the correct veolcity's for the RPM you desired and it would run fine. The reason that we cam up with the cc volume in the first place is because it is real hard for the average guy to understand what I explaind above, let alone try to compute it without the aid of a computer back in the 70's. Now if you have the formulas and a computer it's cinch. I still do it by hand on my calculator.

Where are the bananas???

Denny

 

Some of us are..

 

Very nIce post.. !!!!!

 

Powerstroke's have not yet reached the power of the dodge...our pull truck makes about 670-750 depending on which tune is in the computer, there are only about 5 trucks in the country that I know of that are over 600hp on diesel only (no NOS)...the problem with the ports is that they have to make a sharp dog-leg around the large HEUI injector bore in the center of each chamber and on top of that the port on each end of the head has a significantly smaller cross-section than the two ports in the middle.
Some pics of a recent build....

http://photos.thedieselstop.com/show...i=machinistone

Specifically the port on a head we cut...
http://photos.thedieselstop.com/show...t=1&thecat=998

 

Great Pictures

Is the early head the same as the powerstroke? (Early meaning mechanical injectors)

Have you tried porting the heads?
Putting bigger valves?
Getting a cam ground?
From the pic's it looks like you could put a chamber in there like I do in the Dodge's

Q - How little compression does it take to get it to deisel? Why?
Because I think that you can make more power if you can lower the C/R and add more boost.

Boost equals CFM
CFM = CID *RPM/3456
7.3=446 CID 446*4000/3456=516CFM

516 CFM = 1BAR (15LBS atmospheric pres) 2 BAR=516*3 (1548CFM to achiev that pres) So it looks like you're going to need a turbo that puts out at least 1600 CFM at 30 lbs boost.

But several things you're going to have to look at:
1) Cam timing events
2) Air quality Vs boost pres
3) Some type of clean exhaust system (headers)

We found out with the Dodges that we could put 70 PSI in there and hurt the power, because of poor air quality. All the really fast turbo cars have chillers to create high quality air density.

What I don't know is if there is enough room in the truck to do all of that?

Denny

 

Early 7.3L IDI(indirect-injection) heads are very different and the intake ports are a straight shot to the valve.

We spent 60 hours porting the first set and that is purely bowl work and matching port volumes...dropped 10 lbs boost and picked up 50hp at only 50psi.

We have to run a large single turbo (2.8" inlet) to stay within rules, and our turbo that makes the best torque maxes out it's map at 60psi. Others have tried reducing compression and they are making less power than we are, not saying that it's the wrong thing to do, if we were able to push more than 60 psi I would drop the compression to 15:1 instead of 17.5:1 and turn the boost way up. This is what you mean by chamber, as a friend had done to his heads...



We have a very large CAC that keeps the inlet temps after turbo to near ambiant. The turbo is a hybrid switzer that is loosely similar to an L-14 Cummins unit, it flows lots of air but going bigger makes spooling to laggy even after dropping exhaust housing size. If we could run compound turbo's we would not have a problem.

Turbo on the left is ours, similar in size to a Garret GT42, stock turbo is far right


One company has tried building stainless headers, but they seem to crack I have heard of no serious power gains, and from everything I know about turbo diesel's we should make cast iron manifolds that flow like headers but will retain the heat needed to run the turbo...heat=power.

 

Yes, Yes & Yes

In the pic with the chamber, which is a step in the right direction, you need to get rid of those sharp edges around the valve area. To do this make some valves that have knife sharp margines so you can get right down to the seat areas and smooth all that up. Scribe the bore onto the deck and lay out each chamber by hand so that you can get everything even. Then when you set the valve-job depth in your Serdi machine all the chambers will be within + or - 2cc's. Valve job 35 top 45 seat - 60 or 65 bottom use .045 seat width - put the largest valves that you can find even if you have to get someone like Ferrera or REV to make them for you.

As far as the headers are concerned its not so much the primary tubes its the collector that is most important. Get Hugh Macinnes's book (ISBN 0-89586-135-6) on turbocharging and look on the back cover - thats a prime example of what turbo headers should look like. To retain the heat you can get them coated inside and out and then wrap them with heat retaining tape from Swain Coatings.

You didn't answer the question about the early heads - will they bolt on the powerstroke engine? If they do then use them instead, and switch over the the mechanical injectors. Port the whole head, intake, exhaust, chamber and fix the back side of the valves just like you would for a gas engine.

Next: I found a lot of info about deisel cams specific to powerstroke on the net by typing in "Cam Design" !!!!!

Denny