In the "Cylinder Head Selection Factors" issue of Tech Talk, we mentioned the media's fascination with "racing" flow benches. Here's why that type of "competition" is a poor method of predicting performance or selecting components.

 An excellent example of air flow vs horsepower was seen way back in 1981, with the 1980 Firebird Pro Stock of The M&M Boys.  Their engine was a 340CID Oldsmobile with aluminum heads, 2.08" intake valves, 1.60" exhaust valves, a fabricated intake manifold with two Dominator carburetors and a .720" lift roller cam. The combustion chambers were machined to fit the cylinder bores and the intake ports received straight forward porting to make good numbers throughout the flow curve. At .750" lift, the intakes flowed 393CFM and the engine made 640 peak horsepower.

At the same time, they also had a 331CID Small Block Chevy engine that made 690 peak horsepower on the same Dyno. After reworking, reducing, the Oldsmobile intake ports to flow 301CFM at .750" lift, the engine jumped to 705 peak horsepower. HOW? Port SIZE and SHAPE!

Every time this story is told, the first comment I hear is "Oh, making the port smaller just increased the velocity." Maybe, maybe not. Flow benches do not measure velocity, they measure cubic feet per minute. According to the flow bench, the revised Oldsmobile ports flowed fewer cubic feet per minute than the previous ports and the engine showed its pleasure with the change by making 65 more horsepower. REMEMBER: Port size and shape. The Oldsmobile's intake ports weren't just reduced, they were reshaped.

Then, attention was turned to the Oldsmobile exhaust ports. The .700" CFM lift numbers went down, from 237 to 225, but the horsepower went up to 710. All we did was to mill 1/2" off of the exhaust ports at the header flange. This illustrates how the flow benches can "like" certain things that engines won't.  Another example would be adding a pipe to an exhaust port. The longer the pipe, the more air the port will flow. Not only that, the pipe will smooth out some of the turbulence or imperfection (which you can hear and see on the manometer) that may be in the port. That doesn't mean the port is fixed though. The extension is more like a band aid.    

Conversely, flow benches typically like short intake ports. All said, two different engines cannot be evaluated on air flow numbers alone. It's not necessarily correct to say that a Hemi makes more power than a Big Block Chevy because the intake ports flow more. The Hemi intake port is shorter, from the the intake flange to the valve seat, than the Chevy's port.     

Comparing (or "racing") flow benches is just as difficult, as benches are different from brand to brand and even from model to model. Fixtures, over-the-counter or shop-built, also differ and will affect flow readings.  The engine or bench bore length can change flow numbers throughout the curve. The thickness or shape of the ideal entrance on the intake side makes a difference. Operators and the operating system can make a difference when warning up the bench or correcting for ambient air. Even the quality of the wet & dry bulb for checking the relative humidity, the barometer and thermometer can affect readings.  

To be sure our readings are consistent, we have a series of model heads for benchmarks. They have been precisely measured and we NEVER change them.  Periodically, throughout a test, we flow an appropriate model to check for changes in the ambient air and to verify changes to work piece. If we picked up 1CFM on the work piece, and the model also picks up 1CFM, then we know the air changed, not the work piece. In the end, a flow bench test is a comparative for measuring before and after work on a cylinder head.

To illustrate the difference between flow benches, let's look at a customer who called about a set of SBC heads. The manufacturer, (NOT Total Flow Products) claimed they flowed 310CFM, but didn't provide any substantiating data on bore, depression, etc.  The customer had another local shop flow the head and they saw only 298CFM. Now the customer was asking us which number to believe and what to do. In this case, the head was a good piece that was known to make horsepower in the customer's application. We simply told him that both numbers were probably correct and that bolting the heads on to his engine would make his car run just fine.   

Remember, air flow data is relative to just the individual bench and operator. To learn from flow testing, you must:

  1. Flow your finished heads on a given bench and record the data.

  2. Dyno test the engine and record the data.

  3. Run the engine in the car (multiple passes), determine the best tune and record the data.

  4. Make ONE change at a time.

  5. Flow the heads on the same bench and record the data.

  6. Run the engine on the same Dyno and compare the data to the data from step 2.

  7. Run the car again  (repeating step 3)

  8. Compare the before and after flow data with the before and after engine data.

NOW your flow bench data means something. The next time you see a magazine ad or article evangelizing a cylinder head's flow numbers, you can be confidently immune to the sermon. Those numbers don't mean anything to you.

REMEMBER: THE FLOW BENCH IS A VERY USEFUL TOOL...but its not the final word in selecting heads.  And bigger isn't always better...                           

Total Flow Products race engine cylinder head porting

Cylinder head repair shops

cnc cylinder head porting

aluminum cylinder head repair

​cylinder head porting services

Total flow Products

Cylinder Head Porting Services