Why test @28"?

[Brucepts]

Thought this was an excellent explanation as this question comes up from time to time;

From a post off Meaux Racing Heads forum with permission from Larry to post here
MaxRace Software
ET_Analyst ( DragStrip Simulation Software )
PipeMax ( Engine Simulation Software )
http://www.maxracesoftware.com

Why is 28" any more valid then 10" if it still isnt anywhere near real world depressions for accurate data

if every Port shape were a perfectly straight round tube that magically did not need a Valve and Stem inside, 10" WC would be about all you needed to get good accurate Data.

10"WC will still give you decent Data in most Port shapes
28"WC will give you more accurate Data in a greater majority of Port shapes
36"WC is still even more accurate, very seldom will the FlowBench fool you at 36"WC +
and from what i've seen, 45"WC to 48"WC should be more the Flow WC Standard than 28"

10" Water Column Height (WC) = 209.343 fps or 142.73 mph potential molecular speed in the centerline of a perfectly straight round tube

28" WC = 350.297 fps or 238.839 MPH

48" WC = 458.647 fps or 312.714 MPH

105" WC = 678.348 fps or 462.51 MPH


Analogy->
Suppose you are designing and testing a Formula One RaceCar , and you do all your Suspension/Tire setups and testing at 10" WC or 142.7 MPH well you will get good Data but probably not good enough to Win the Race .

at 142.7 MPH is probably going to be slow enough speed that just about any good Formula One Car can handle a normal shaped Curve without spinning out and blocking out most other Cars in that Turn.

Now suppose you Test and design for more realistic MPH Speeds of Formula One and you Test at 28" WC = 350.297 fps or 238.839 MPH and now those same Curve "Shapes" that you easily went around at 142.7 MPH, now at 238.8 MPH you spin out or barely make it thru.

i know the above is not the Worlds best Analogy to give you , but i think its the easiest to understand ?

Now suppose again you have the ability to change the Road Course into a straight line, which is analogy to straight round tube, then it becomes a DragRace and all the Cars ( molecules) don't have much a problem going straight at 142.7 MPH to 238+ MPH , so its who gets to finish line first , Wins

getting back to Curved Shapes,
lets make the Formula One Cars tow a Trailer with heavier Fuel on it ( Wet Flow ) and see how fast and how many Formula One Cars that can make it around a Curved Shape ??

FlowBench = Dry AirFlow testing , no heavier Fuel suspended in AirFlow stream trying to make a turn in your Tests.

there are some Port Shapes that definetly need to be tested on a FlowBench at 45" WC upwards to get accurate enough Data,
and
more Port Shapes that when tested at 28" to 36" WC give very accurate correlation to Real World

Another major problem with a given FlowBench Test Pressure is that Test Pressure is the theoretical average pressure thru the Port CenterLine or the theoretical average Speed thru the Port Shape,
but
because a Port has a Valve/Stem and Curves in its Shape, and changes in CSA area, the actual local FPS speeds will be drastically different than the theoretical speed.

Lets say your Flow Test Pressure = 28" and use Pitot Probe your Port Shape in many areas inside, and you can see Pitot speeds/pressures 10+" inches higher than your Test Pressure.

[Dan_DiM]

the 28inch number probably started with smokey yunick. i remember reading he tried different values and settled on 28 because as depression goes up so does flow and therefore resolution. bigger numbers make small changes easier to read. you also get a diminishing return, more depression beyond 28inches doesn't significantly increase flow like it does going from 10 inches to 28 inches, but this was on inefficient chevy heads and the cost and size of flow equipment at that time may have also been a factor in the decision. more is still better and mre efficient designs might not choke the same way as an old smallblock head. especially if turbulence changes with velocity getting near real world velocity would be ideal but it's a compromise that was chosen in a different time and it's become an industry standard. if you have the equipment to flow at 40+ inches then by all means do so, it seems like a great idea..

[Fahlin Racing]

Smokey, from what I read in one of his books came to the '28 inches' because he saw the actual changes on the flowbench once things were ported good or bad at that level of depression. Interesting thread brucepts, thank you for posting the wisdom of Mr Meaux! I learned something new as to why.

[Wheeljack]

Smokey, from what I read in one of his books came to the '28 inches' because he saw the actual changes on the flowbench once things were ported good or bad at that level of depression. Interesting thread brucepts, thank you for posting the wisdom of Mr Meaux! I learned something new as to why.

Yes, I've read the same/similar, maybe in 'Engine Airflow' IIRC, that the results at 28" depression seemed to correspond most closely with actual dyno/hp results, which is really what it's ultimately about anyway. Some would even argue drag times, etc....are the ultimate goal, but that brings in a lot more non-engine related varables. So 28" was chosen as the standard for that reason.

[powellmachine]

@ less than around 18-20" the air follows the short turn and exits the full dia. of said valve, but in running conditions it is much higher, around .05-.1 lift some estimate 80-100" then decreeses dramaticly to around the 15-20" mark as the valve leaves the seat which at very low lift acts like a venturi, at higher test pressures u get turbulence and separation @ the short turn, that's why the short turn shape is so important, just my 2 cent!!

[Tom Vaught]

The Ford Motor Company did all of the Flow Bench and Engine modeling work on a bench that could accurately test heads at high test pressures (greater than 67" water column), (using a orifice matrix grid in a large orifice tank). The walls of the tank were 1/2" thick and the tank weighed several hundred pounds. The tank had an "access door" which gave me the idea for the 1st Super Flow 1200 "Port Hole" modification. Now everyone uses the "Port Holes".

So back to the Test Pressure deal. Ford could test at higher than 67" of water column. But everyone had an opinion on what was the correct test pressure? Obviously the guy who came up with the first Super Flow 110 could not test at + 67" test pressure. But at least he could find some info at 10". Holley carburetor always tested carbs and induction systems at 20.4" of test pressure. Smokey liked 28" as was mentioned.

So Ford, since the early 60s, has tested at ALL of the test pressures: 10, 20.4, 28, 36, 48, 67, and on lower flowing heads up in the 80s.

Many times there is not one "magic" test pressure but each test pressure capability will provide more info for a given head.
Bob Mullens, (did a lot of flow work on the Chrysler Hemi heads of the 60s, which worked very well). had a bench that could only do 6" of test pressure but a lot of flow.

So good luck with your investigations. larry M is one sharp dude.

Tom V.

[Old Grey]

I remember reading on Speedtalk a long time ago - naturally details are going to be sketchy because of my memory - that a guy a Mopar worked out that a running engine has an average depression of 56". Now because 56" benches were so expensive, he came to the conclusion that half the depression, 28", afforded the best value while producing close, to better results, than a 56" bench. His reasoning at the time was that if you port at 56", it's so sensitive that grinding anywhere in the port produced a gain - they obviously didn't know much about velocity and port areas at the time -, and if you chased the gains, your port became an oversized drain pipe.

The half 56" came out good, if you do direct velocity measurement at 28", because all you have to do is double it for actual running air speed.

[early Hemi]

old thread but worth revisiting- the reason the 28" standard was chosen, was at that pressure drop, changes and improvements made on the flow bench using bare heads, correlated to actual horsepower increases as measured on a dyno, when the engine was assembled- in back to back testing. a direct link was found between flow bench numbers being increased via porting, and horsepower numbers on the dyno going up when crosschecked on the dyno.

28" is the "no bs" number

at 8" or 10", you can make substantial changes in the port, and it not even show up on the flow bench at all. but these changes would show up really quick at the higher pressure drop of 28", be they improvements or detriments

to make a really blunt comparison, flowing at low pressure drops like 8" or 10", is like playing a 78rpm record on a phonograph, at 33rpm or 45rpm. sure it will play, and you're hearing the song and information off it, but it's playing too slow, and it's distorted, and you can't make out all the words in the song...you're not getting the jist of the recording. when you switch it up to the proper speed 78rpm, suddenly it all comes into synch and makes sense, and sounds right, and you're hearing the notes and words how they really should sound. it's the same record and phonograph, but it has to be played at the right speed, to work correctly, and make sense.
in the 1970's and 80's the recording companies tried high speed dubbing, where tapes would be recorded at 2x or 3x normal speed, both source and dupe tape. then the duplicate tape would be slowed down to normal speed to be played back, by the consumer. it was a way to cut time and costs making millions of copies. the problem was, the playback speed was never quite correct, if the tape was high speed dubbed, and some of the frequency response was lost. the tapes tended to sound tinny and sounded like they were playing slightly fast, too much margin of error was introduced.
this same phenom occurs to some extent with flow benches, but in a different way. slowing down the air to do the tests on a cheapie bench, then extrapolating the results back up to 28", a lot of the flow test information and characteristics of the port, are lost in that process. the bench has to be run at the right speed to be useful, in this case air speed.
another phenom is, the faster a record or tape is recorded on the media, the better its fidelity and frequency response when played back. the same applies to a flow bench. slowing down bench and air speed, reduces the fidelity and accuracy. when that ported head is "played back" later in the assembled engine, it's not going to "sound" the way it did, when you ported and flow tested it at 8" or 10"

the flow bench is a more efficient flow measuring tool, at 28". no professional engine builders worth their salt, who care about their reputation, would even think about flowing at 10" and extrapolating up to 28" using math. some of the front runners flow their heads higher than 28", for instance Warren Johnson. it would severely affect their reputation amongst their peers and customers, and raise doubt whether their work was reliable or not, if they used a cheapie low powered bench, and extrapolated the numbers to 28" standard

[DaveMcLain]

Some of this might be a little misleading too in that back in Smokey's day he didn't have electronic manometer devices with the resolution of the devices of today which would have reduced the resolution he could get when using lower depressions than 28 or so inches. If 28 is such a standard why does Superflow rate the SF600 at 25 inches of water?

[Brucepts]

If 28 is such a standard why does Superflow rate the SF600 at 25 inches of water?

Probably because they might not have the power to get 600@28" would be my guess?