The Truth About Airflow and Horsepower
By Brian Tooley, Owner of Total Engine Airflow

Get your pencils sharpened and your calculators handy. We are about to make you not only a faster racer, but also a better consumer because you will be armed with the knowledge you need to know and understand in order to purchase the best cylinder head port work for the least amount of money. Dollar for dollar, good head porting is worth as much as nitrous or a blower.

It took years of development time to come up with the knowledge we possess. Working previously at a major automotive aftermarket company designing heads and running dynos certainly helped my learning curve. The information you are about to be given was not given to me. I worked diligently and made many mistakes along the way to get to where I am today.  Learning things the hard way is still probably the best way to expand knowledge.  What you are about to read is the latest technology available short of having your own flow bench and years of experience to back you up.  I do hope the following will help you enormously, and it will, if applied.

We have all saw the flow numbers everyone posts on the internet and in their magazine ads, 300, 305, 320, 330, 340, 350, and even 400 cubic feet per minute or cfm for short.  These numbers are obtained at .600", .700", and even at .800" lift.  Most people that flow heads do so at least every .100" of lift.  This means that there are numerous points to plot on a curve before this numerically high peak flow is achieved at the given lift.  These points before the peak are actually the most important.  You may have also heard people say that these points are more important because the valve may pass these twice, but only reach the peak point of lift once or not at all.  This is sound logic, but how do you know for sure?  We now have the technology to compute mathematically which head should run better.  This is based off a little known and seemingly secretive formula that actually computes the "area under the curve."  Computing this area under the curve is not new to mathematicians.  They have known this formula for 50 plus years.  Using it as a consumer to buy heads is fairly knew, at least I've never heard of anyone using it and I've read a lot of literature on the subject.

Comparing heads using peak flow numbers is NOT the way to evaluate a head!  The only way to evaluate one cylinder head against another is to compute the area under the curve of each.  The only way to accomplish this requires all the flow numbers in at least .100" lift increments.  If your porter cannot give you those numbers at the industry standard test pressure of 28 inches of water or if your porter dismisses the area under the curve, simply say "Thank You", and never talk to him again.  Now let us look at how to computer the area under the curve.

Let's say we have two flow curves:

	.000"	.100"	.200"	.300"	.400"	.500"	.600"	.700"
Curve 1	0	68	143	225	273	305	312	315
Curve 2	0	68	138	200	250	290	315	325

The average person would say that the bottom flow curve would make the most power, but in some cases it would not!  The formula for area under the curve is AREA=(right boundary – left boundary) x (left value + right value) / two.  Our "boundaries are the lift points .000", .100", .200", .300" etc... Our "values" are the flow at each lift point.  This formula basically makes a rectangle out of a trapezoid by averaging the two points.  The first are A1=(.1-0)x(0+68)/2 where A1=3.4.  You can visually see how this formula works by taking the two flow points added together divided by two and then draw a horizontal line through the 34 cfm point. Of you look at the two triangles created, you can see that A=B.  Let's do the same exercise for the second area, (68+143)=211, then 211/2=105.5.  Draw a line through A2 at 105.5 cfm.  The other way to visualize this is to take our new triangle "c" and rotate it counter clockwise 180 degrees to fill in the area to make it a perfect rectangle.

The second area A2=(.2-.1)x(68+143)/2 A2=10.55

The third area A3=(.3-.2)x(142+225)/2 A3=18.35

The fourth area A4=(.4-.3)x(225+273)/2 A4=24.9

The fifth area A5=(.5-.4)x(273+305)/2 A5=28.9

The sixth area A6=(.6-.5)x(305+312)/2 A6=30.85

The last area A7=(.7-.6)x(312+315)/2 A7=31.35

Now add the areas A1 through A7 and get the total 148.3 inch cfm.

This is the number you would use to compare one head against another.  If you have a cam with approximately .500" lift, then use only A1 through A5.  If you have a cam with approximately .600" lift, use A1 through A6 etc…

Now that we did it the hard way, let me show you the easy way.  The boundary is always .1, provided you get the flow numbers every .100" lift.  The bottom number is always two, so we can derive a constant or .1/2=.05.  The new formula for area is AREA=.05x((first value+second value)+(second value+third value))... etc.  Let's recompute the first flow curve using our new formula. AREA=.05x((0+68)+(68+143)+(143+225)+(225+273)+(273+305)+(305+312)+(312+315))=148.35 inch cfm

So now let's do the second flow curve

AREA=.05x((0+68)+(68+138)+(138+200)+(200+250)+(250+290)+(290+315)+(315+325))=142.35 inch cfm.

You can see that the first flow curve that had 10 cfm less peak flow has over 4% more "area under the curve."

Now let's do another exercise.  Take the second flow curve and improve the peak flow even more.  Let's make it flow a whopping 340 cfm at .700" lift and recomputed the area under the curve.

AREA=.05x((0+68)+(68+138)+(138+200)+(200+250)+(250+290)+(290+315)+(315+340))=143.1 inch cfm.  WOW!  Can you believe that?!  Our second flow curve is now out flowing or first be a whopping 25 cfm, but still has less "area under the curve."  So you can start understanding why some high flowing heads just don't seem to run the way they are supposed to.  If you were racing flow benches with these numbers, the second flow curve is the hands down winner.  If you are racing automobiles down the drag strip, the first flow curve would come in first.

Now that we have established that principle, let's talk about port volume or more accurately port cross sectional area.  Every engine wants a specific area in order to adequately fill the cylinder with air.  This can be determined only through trail and error which some people call experience.  There are enough people out there doing things to keep in mind though.  A port that has a lot of "area under the curve" tends to be more lenient in regards to port volume.  It is more "active" and less "lazy."  A lot of heads can only achieve big high lift flow numbers through larger areas or volume which is sure to kill the performance of the average street/strip car.  The area or volume needed of the head is dependent on common things like bore, stroke, rod length, and RPM, but is also very sensitive to the induction system (intake).  This is one area that we have not seen people hit on, but it vital to a typical long runner (upper/lower manifold) combination.  The longer the intake tract needs to be, including the head, in order to adequately fill the cylinder at peak RPM.