How To:Basic Cylinder Head Porting for Beginners

Mike,
Yes! Floor shape is critical to low lift vs high lift flow. Part of the reason you may not have found a good compromise is because it all depends on the cam profile, most specificly the rate of valve opening and closing. The slower the rate, the more low and mid lift is important. There is also a sound relationship between throat size and valve head OD, which can be influenced slightly by seat and tulip angle.

George,
Nope, name is Aaron. We use a combination of flowbench, CFD and running engine for development. For rod ratios we did some testing running from 1.5 to 2. Friction was the biggest thing influenced, with VolEff and burn rates being within the normal scatter. Honestly the baro and humidity would influence the burn more than rod ratio. Friction seemed to be the lowest around the 1.75-1.8 mark. The best tool I have ever found for power development is pressure xducers for in cylinder crank angle resolved measurement. Unfortunately not in the realm for most of us.

 

I could tell by the way you were talking you were doing a lot with CFD, thats why I asked about the JS, I thought you might be Jon Schmidt, I know he does a lot of CFD work for the OEM's. I dont know if you are aware of it or not, but CFD is making BIG inroads with most of the serious aftermarket head guys now. I dont know if its true or not, and I am not commenting on the importance of this one way or the other, but I have heard rumours that Tony Mamo is not one of the guys moving in this direction. I am just making a wag that AFR may be the main aftermarket co. you have worked with, that may explain why you have the impression that the aftermarket isn't using this technology? Not trying to put words in your mouth, but rest assured, there are other players out there using this technology.
Interesting on the R/S stuff. I usually target R/S ratios in the 1.75/1.8 range, simply because if you stick the longest rod you can in most of the stuff I'm working with, thats around where you end up, and I find long rod combos to be much easier on the bottom end, and the pistons are lighter. But 98% of the guys on here dont want to hear that, so I generally stay pretty tight-lipped about it.
I have read a fair bit about ECA using pressure tranducers, but as you say, its so far outside realm of us backyard engine builders at this point, its more of a curiosity than anything else. still good to at least have a half-assed understanding of what the real players are up to, regardless.
On the Studebaker, and cylinder wall shrouding, there were a couple other things I was going to mention, but I had to go out, and this is pertinant to the rest of the info on this thread as well.
In addition to scalloping the top of the bore adjacent to the intake valve, depending on exhaust valve size, you can either use offset head dowels to shift the head on the block deck, and move the intake valve away from the bore a bit, or, depending on overbore and sonic test numbers, offset bore the block baised away from the intake valve. Probably a combination of the above will be what you would end up looking at.
IIRC, these stude heads are for a Bonneville application, right? As peak power numbers would be your primary concern, if forced into an either/or situation for a Bonneville application, I would rather compromise exhaust flow in favour of high lift intake flow numbers.

 

I edited the words about "joint" out, this is solid gold right here. Commit it to memory...
Heres a helpful analogy for you guys, and it may help some in the way they think about shaping the short side radius, when you turn, you gotta slow down, and the sharper you turn, the more you gotta slow down.

 

George,
I have seen some of what the racing world is doing with CFD, it is good that they are finally getting into it, I just hope they don't rely on it too much. The simple 1D and 2D models don't give you much more if any information than a talented person with a flow lab. Its biggest use seems to be 3D and moving mesh work with actual valve lift and piston movement, not so much for flow development as for motion development. As of right now I believe it is only GM, Ford, FEV, GE and perhaps Lotus really getting into it (us too). There was a really interesting symposium I was at a few months ago at the local engineering college that was strictly on combustion modeling for engine development. One thing I thought was kind of funny is that with all of our technology, we still have to ask the engine what it wants in the end. The technology just gets us there quicker, and time is money of course.

On the topic of energy loss and such, if it has already been posted here please forgive be, but here are some other simple rules to live by.

1-The optimum shape of a straight port, is round (think surface to volume ratio).
2-If you have to make a turn, the aspect ratio of the port needs to change (think cobra neck).
3-Any change in direction, or velocity is a loss of energy, and the biggest loss is the air exiting the port into the chamber.
4-Actual cross section does not mean effective cross section (think boundary layer thickness)
5-There is no such thing as laminar flow in a port (it is fully turbulent)
6-Dimples and such are nice in magazine articles....(see #5)
7-Dont bother polishing a port, anything smoother that what a coarse flap wheel leaves (or a nice cast finish ) will actually increase the boundary layer thickness (see #4)
8-Ports should be quiet, if you cant have a nice conversation while your port is at max lift on the bench...something is wrong. It takes energy to develop sound.
9-There is no one size fits all port design. Every engine design will want something else.
10-Keep it fun

 

Guys,
The thread title says "Basic...for beginners". I was enjoying it up to the point that you've convinced me I need an advanced degree in flow dynamics before I can try some home porting.

 

Mike,
Quite to the contrary. Follow some basic rules and principals as laid out here, and you should be able to do as good of a job as any "rodder" would need without a flow bench.

 

I can just barely keep up here,
Have a question to the flow experts about port alignments and flow on the exhaust side on the way out and specifically the Chrysler hemi engines.
Not to turn this into an exhaust thread, but just related to the flow.

On the 392 heads, The ports are symmetrical oval shape 1-1/2 x 2-1/16 when measured straight on. However that is really an angled cross sectional measurement and much larger when measured on the angle. Most headers I've seen are a square stub to this angled port, then an immediate elbow to change the direction. And most are 1-7/8 pipe transitioned to the 1-1/2 x 2 oval. The square stub is an abrupt wall to hit on the way out, then the immediate elbow is another direction change. Now all this is happening in an area that is a large increase in volumetric area compared to the exit port. Wouldn't all those combined situations have an adverse effect on the flow out, velocity, & scavenge.

There are very few headers I've seen that have a pipe exit at the port shape on an angle that flows with the port, namely the Ramchargers gasser is about the only one.

 

Aerometal -

Not talking about cam profile at the low lifts I'm speaking of (.200" to .300") vs. .500" and higher.
Just pure measurement numbers.

You can form/shape the short turn one way and measure it at .250" and form it another way and measure it at .250" (using all the same flow bench adjustments) and get very different flow values.

Tha cam comes a little later in the overall design. Plus, in the case of the Stude engine, there just isn't many "different" cams available (though there is a few people with various roller cams I had made a while back).

Chevys, Fords and Chryslers all have their characteristics to deal with also, but those are much more a known quantity and easier to port to good flow.


31Vicky -

On most early Hemis, while yes the port exit is at an angle vs. the center of the port, the flow doesn't know this.
If I understand you correctly..As far as the abrupt change (increase in area) going from the head to the header, this is going the other way and because there are two (pressure and yes, vacuum)) actions working on getting the exhaust "out" of the port/pipe, that step also has little effect. But it does/can have a large effect, depending on the cam design, help keep exhaust from flowing back into the cylinder at low lift.
This can even be seen on high rpm Comp. Eliminator and Pro Stock engines.
The "step" header has shown this increase in volume (or square inches) to not only be true but desired.

Mike

 

Interesting that you brought up the Ramchargers Altered. They based their megaphone individual pipe exhaust on some English motorcycle technology (Norton maybe, but I'd have to check). Anyway, I always wondered what the car would have done with a 4 into 1 collector system.

 

Is there any point in an average guy playing with "good" aftermarket aluminum heads , say Dart or AFR for the street?

That's an honest question, not a snide comment.

 

I always wondered what it sounded like, up close and personal at the strip...Them megs were one step away from the Coanda experiments.....

 

Bear with me- this one pic per post sucks.

Ok here's the high and mighty.
Note the pipes are angle cut to meet the head.

 

That makes a pretty smooth straight shot out of the port.

 

Here's a shot of the later version.
This one has the straight 90* header flange
http://www.motorbase.com/profiles/p...ne Intake & Headers (2005 CEMA) N&pp=&ppage=0

 

The ruler is at the angle of the exhaust port roof.
The pipe in the background is angle cut to match that angle.
Pretty straight shot out of the port then off to a long radius gradual elbow.

BTW its not easy ovaling the pipe, cutting it at the angle. Every time you adjust everything else gets out of wack. It's a pita and easy to see why it's not done. The straight take offs are a walk in the park.

 

Here's the straight takeoffs.
The black lines on the background are the out line of the angle cut pipe in the previous pic.
Then the elbow should be closer to 90* ( didnt have one) to get the pipe over to the black lines. It needs to be to those black lines or the header pipe hits the frame.

It just looks like pure torture on the exhaust flow.

 

Ha, I'd have a few more elbows.

Almost every set of hemi headers I've ever seen have been like this.

 

Try not to get dizzy, but I though it would be helpful to see the frame dodge with the head orientated the same way as above.

 

The reason I'm asking is most of the primary sizing and cam testing is done with that goofy ass jog in the headers. To me its way beyond stupid. The primaries are over sized to compensate ? I think so any way ?

Now if I decide to screw around with the exit path, how is that going to change my other variables that there's decades and big dollars in R&D. ?

Just for reference, I can oval 4 straight pipes and fit and weld the 8 elbows for one head quicker than getting 1 angle cut pipe to fit correctly.

 

Honestly, its a little beyond my rather limited grasp, but Jon and some of the other guys were discussing moving mesh syms , whatever the hell that is...Your ten points above, I am up on all of that except point #5?? Can you elaborate? Maybe pm me again, so we dont upset anyone who just wants to grind iron,and doesn't want to understand what happening in a port...

On the dimples, someone asked if it had ever been used inside an intake runner, I was going to post a photo(not that I am endorsing the idea) but haven't got around to it yet, if I remember I will post it tomorrow.

 

One thing I can tell you, and I dont need a flow bench or any syms for this, avoiding any sharp bends for at least the first 4" of the header primaries will pay dividends in power every time. In fact, I have seen it more than counter the difference between a primary that is 1/8" too small, and one that is optimum.

 

That's what I always thought too.
I never paid too much attention until I needed to do it.
It just seems plain wrong to do this unless there is no alternative.
Plenty of BIG buck cars with custom made pipes and they are all like this.
One ginormous kink followed by a 90* elbow immediately after the port.

 

George and all,
Ahh point #5. Simple enough to explain, velocity determines if the flow though a port ( or runner or exhaust pipe ) is laminar, partially turbulent, or fully turbulent. The velocity through even the largest port is well into the "fully turbulent" area, so things that work at low velocity, are pointless. Dimples work at the 100mph (ish) and under mark, but most ports are over 200mph. As for the low valve lift flows vs cam profile, indeed there is a lot to be gained or lost there. Probably not as important in say a drag application, but certainly in a road course or street car application. That being said, its probably a mute point as we are stuck with the cams from a catalog, but still something worth thinking about. If you have an engine design that has a fairly flimsy valvetrain, with a mild cam, you might be better off working on the 1-8mm lift range flow values than some high lift number that you wont spend any time at on the induction stroke.

Oh and on that exhaust header situation, all I can say is.......yech! Exhausts should be a seamless extension of the port. You can make a bend fairly close to the head, but the port has to be designed for it. The situation you have......is anything but.

 

Well I'm just a regular asshole that likes to think about stuff. Especially when it's cold and cabin fever sets in.
I've never sat in front of or stood next to a flow bench, but those header designs look good but the dynamics of them go against everything I understand to be correct. Apparently since the 1950s, I'm the only one who thinks its a problem or willing to think about it. Just seems strange that the big buck gurus haven't done something about the obvious " YECH" here.

 

There might be bigger "yechs" in the engine design than the exhaust, and so it is not the limiting factor. If it is big enough, it does not have to be as efficient to not hurt power too much

 

Something like this?

 

Well thats not a early hemi, and I have no idea what port or port roof looks like on that. That's all I can say on that one.

Just compare the pic on 165 to 167 and then look at 172

 

The bitch of it is the packaging.If we weren't stuffing big Ol Mister Hemisphere into a tiny little Model A or something equally as 'real estate limited',life would be easy for us.

Using tall motor mounts like many did back in the late 50's/early 60's can provide a lot of clearance between exhaust ports and those damn frame rails.

Sometimes running the exhaust over the frame instead of inside it would really free up room and provide a more unrestricted flow path,but not everybody wants their pipes running outside the engine compartment.

I have to admit,when I sat my set of 555's up on the bench,that's the first thing I thought of when looking at the exhaust ports: "Damn,these things don't exit on a 90* angle to the flange at all..."
When I get to mocking it up in the car I'm going to give careful consideration to the exhaust when setting the engine location.

I'm not big on compromising,so I might be pretty grouchy by the time I settle on a location....lol

Scott


Posted using the Full Custom H.A.M.B. App!

 

The way i see it :
The first compromise is the 90 take off, but its way easier to get started.
If you don't go down that rabbit hole there's no need for the other compromises to pay for the first. No secret that oversize primaries rob you of plenty. How do you over come a 3 way kink? With an over size primary is one way I guess. Fuck around and oval shape slash cuts for outside or oval elbows for inside is another.

I really find it difficult to worry about short sides and roofs and smooth radius ports then shove a potato in it-

So how does the flow thru an oval react vs an identical circumference round?
I'm sure the oval flows less and the drop is on a pretty big bell curve with respect to the width vs heigh ratio of the oval? Any rule of thumb for figuring such? Thanks

 

31Vic-
Glad you asked, when you use anything but a round cross section, you need to look at the equivalent "hydraulic diameter" for the shape you are using. Google it and I bet there are some calculators out there to help with this. As long as the effective or "hydraulic diameter" stays constant throughout your pipe, you wont suffer much loss. This is why when you change the aspect ratio of a duct to go around a bend, you need to make the area bigger. Its not so much "slowing it down" as it is maintaining a constant effective diameter. Hope this helps.