Approach to engine, cylinder head development - Cyl head enging development

[FTCH]

Hello everyone, I am extremly please to have found such a great forum. I feel I will benifit alot from here so it is only fair I try to contribute. It is quite clear other have far more experiance than I do.
Back ground: I am from Australia, I have been port heads and developing engines for 25 years. I only have a small super flow, flow bench, which has served me well. I built a large cenrifical fan and shroud which runs to do other testing with, however I never completed it. I raced in Australian Pro Stock (Small block chev 340 cubes) in the late 90s early 2000s. Always qualified. The last big race meeting here was our winter nationals (June 2008) in queensland. Two cars with my cylinder heads ect, set new national records. I have reasonable experiance in blown alcholo and other forms of motor sport. In recent times I have had great access to a dyno so we can now test more ideas ect.

My approach. The first thing I ask a customer is, what is the bore, stoke, rod length, (cubic inches) and what rpm do you want to make peak power. This then gives the numbers to try and caculate the minimum cross sectional area. On a small block chev its the push rod. All of my engine design stem from there.

Stepping back 20 years. When I first got my flow bench I figured I had all the answers. CFM was everything. I like so many here and those new to this game starting reading everything I could. I kept grinding and flowing. We had no dyno access and I figured there must have been a better approach. Wanting a better system I started looking at computer programs. The programs I first used were Engine Expert and Dynomation. The orginal authors were a great help to me and hopefully I provided them with some good feed back as well. I feel the information books from both of these were a massive help before even loading the disks. I also appreciated the limitaions that simulation had. The best leason I learnt, was to take a mathmatical approach. In saying that my maths is extremley limited (english even worse just check spelling). I started making port molds, then cutting them every 1/2 inch, stamp on graph paper and started counting squares. OK now I had thought I had all the answers!!! (ah wrong).

Moving on: I read an artical in a very famous USA magzine from a VERY VERY Famous engine builder. The artical was about airflow through a port. The problem was the artical stated that the air in the intake port was fastest on the outside of the bend as it had to travel the longest distance. The same as the water flowing down a river always erodes the outside bank ect. When I use a pitot tube in the inlet it was the opposite to what I found. I genuianly felt I must be wrong, because ther is no way this artical would be wrong. Over time I couldnt stand it any longer so I devised a test that measured presure all the way around the short turn and long side radius. This test put me on to presuure gradiants or and different airspeeds in different parts of the bend. It is these differences that I believe cause many of the problems young head porters come up against when they see flow drop off, tumble,flow bench pulse at high valve lift. Normally above 25% l/d ratio.

(Will continue writing some more in a few days, its getting late down under) cheers[B]

[larrycavan]

Most likely, all of us started out with the more CFM has got to be better mind set. I remember when I started porting heads. Looking back on what I thought was right then and comparing it to what I understand now, it's an entirely different approach.

I had a similar situation to yours in both the flowbench and in misinformation. I was told the fast air was in the roof of the port and flow along he floor was slow / dead.

Running my engine configurations through PipeMax has gotten to be my routine initial analysis. From there I jump over to some spreadsheets I made for port mapping, cam profiling, etc.

Having an accurate, repeatable flowbench with digital setup, various software and good pitots to test with allows me to exercise my thinking and engine building approach in different ways than I did just 5 years ago.

In some ways, it's been like walking down a long dark hall with lights that get turned on as you travel. The floor is full of pitfalls. You can risk traveling in the dark, apply the take your chances approach and carry on with that. It sure is a lot easier journey when you find another light to turn on so you can clearly see the path ahead though. Eventually, darkness sets in again and you're looking for another light.

A guy named Larry Meaux wandered over here a couple of years ago and flipped on several light switches. What a difference that made!

During the same time frame I discovered the Flowperformance FP1. What a tool that turned out to be!

Now with the new lights turned on and the new flow computer to use, I needed some software to work with on the flowbench and pitots of the size that fit my motorcycle head porting.

I sat down and created my FPexcel program and Bruce got busy making the pitots.

Things began to happen in leaps and bounds all of a sudden and my enthusiasm for porting was not only restored, but elevated to a new level. Now I had the tools to exercise my twisted mind and more data to contemplate than I ever had.

That's all a little off topic. Just a filler till we get this thread cranking.

Larry C



Edited By larrycavan on 1216970761

[106-1194218389]

[larrycavan]

In our other discussion we touched on the TDC overlap time frame. Taking things into consideration of the events taking place during that portion of the cycle, in many circumstances, it stands to reason that high velocity exhaust ports can be benificial. Again, another balance factor between flow & velocity to kick around in your brain.

I see a lot of good OEM motorcycle head configurations with I/E ratios that are considerably below what was considered optimal in years gone by. These are multi valve engines too.

[FTCH]

Thanks for your comments and thoughts guys, I am going to seek out threads, software and contributors to the forum over time.

Adding to my previous post
Once I established the importantce of the short turn radius, and having a better understanding as to which parts of the port and short turn effected airflow. I set about trying to work with it. I took the view as a starting point that any short turn radius should be as large as it could be. So I would measure the hieght of the short turn from the bottom of the valve seat to the begining of the s/t . For this example lets say its 1.000 inch. I took a small piece of sheet alluminum lets say 2.500 inch square. I found the centre of the plate bored a hole in the same size as the radius. (2.000 diam , 1.000 radius). You end up with what looks like an oriffice teat plate. If you have the plate marked /divided into 1/4 before you start it is very easy to make a template that leaves you with the perfect short turn maximum radius. Does that make sense? You cut away the 3/4 you dont need. using the template on the intake short turn you can see the shape of the turn so much better, it may have a high sharp point, or something else wrong. Next if you know where the air is fastest you can consider how can I slow the air back down, and where it slow what should I do. I found even streamlining or shortning the inlet valve guided can effect the air sometime negatively if the whole turn is not right. Another good study I made was what effected different valve lift points, had on air flow. My current test sheet gives the flow at different valve lifts, then the average flow of all the lift points, the av low lift, the av mid lift, the average high lift. It makes it very easy to compare one head to another and maybe where to start looking.

cheers for now
FTCH

[68Corvette]

I'm pretty new porter so there might be still some development how I should approach cylinder head porting =)

In most cases I want to know about everything about engine, car and usage. Then I figure out if combination sounds feasible and possibly propose some changes to cam etc. if I see that car would be over cammed for stall, rpm limit, weight, etc.
When we have mutual understanding of engine combination I run it through pipemax and my own spreadsheet to check that everything is in the line with the goal.
If it is, then I map cross section areas from intake and exhaust sides. Measure shortside dimensions, valves, guides and sometimes port volume if customer is interested about that.
Nowdays I trust pretty well to pipemax and use CSA values from there as where I am currently flowing heads does not have well working pitot probe =/
But thats one reason why I'm building my own flowbench =)

I enjoy doing other than US V8 engines more, as flow numers have been inflated and everybody just wants big flow numbers.
If I ask "do you want well runnig engine or big flow numbers".. the answer is many times big flow numbers =(
Thats not usually case in european cars because people do not have much to compare, mostly only dyno results.
That leaves more space to do fast FPS runners which may be flowing lower max CFM numbers.

Shortside shape depends a lot of usage.
When some engine has low ss apex and using lot of cam / high VE% goals I tend to make shortside flatter to get FPS down.
But if engine has small cam and runner is allready too big for the application I make shortside a little fatter to boost mid lift flow numbers even if the head backs up after maximum lift.
Also if head has very high skijump style shortside I tend to make port a little larger than necessary to keep FPS more down to get over shortside better.
Or if combustion chamber allows I might also keep it little too tight on longside side to try to keep flow attached better on the shortside.

I guess there are too many options to describe if your dealing with many differents brands and heads =)

[FTCH]

Thanks 68 Corvette, sounds very good approach. I will be investgating and purchasing pipemax, everypne speaks highly of it.

The exhaust port:
I always measur the top and bottom of the intake and exhaust runner, and use it to get my average length. It is my opinion that gas should not expand at more than 6 to 7 degrress included angle of taper. I measure the exhaust valve example 1.600 then a starting point might be a bowl of 87% = 1.392. If the average exhaust port length is 3.000 inches. I then take 3.000inches X.120 = .360 + bowl 1.392 = 1.752 . The .120 is the maximum expansion rate per 1 inch of length that I would ever use. Then if I add the bowl I have my maximum exhaust port exit size. I can now see how this compares to my first step in the exhaust header.
Back to the valve bowl. Most porter tend to cut the valve bowl/ throat with a machine, it means the bowl is concentric to the seat, taking to much out of the back of the bowl is something I feel is a common mistake. If I was shooting for a bowl of 1.392 I would machine the bowl to a proximately 1.340. I then grind up an old valve in the lath to 1.392 diam make it thin. I then shape the back side of the ground template valve so it leaves more material in the back of the bowl and then hand grind the bowl. hmmm did that make sense?

cheers
FTCH Australia