Comparing CFM Flow Numbers
One question that often arises, is how to compare flow numbers taken at different test depressions. At first glance this is actually quite simple. However, (as we shall see) comparing tests made by two different sources may not always give realistic results.
When an engine part such as a cylinder head is tested on a flow bench, a constant test depression must be held or the flow numbers generated will not be standardized. Increasing the test depression will increase the airflow through the part and decreasing the test depression will decrease the airflow through the part. By holding the test depression constant, we are able to see where changes have resulted in increased airflow through the item being tested. This practice is standard procedure for anyone using a flow bench. The actual amount of test depression used however, is not standardized throughout the world, country or even local area.
Test depression is most often read by use of a water manometer. Common depressions are 10, 25, and 28" of water column, although any amount the bench is capable of generating could be used. Once a given user has decided on a depression level, it is common for that person to stick to it, so that any product tested will prove easy to compare across the board. For a single user this does not prove to be a problem. Attempting to compare products flow tested by separate sources using different depression levels however, can be difficult.
Comparing results can be made easier once it's known that air follows a square law of pressure versus flow. This means that to double the amount of CFM flowing through a passage, the pressure differential (depression) must be increased by a factor of four. A simple formula is as follows.
Flow at the new pressure drop = (the square root of (new pressure drop/old pressure drop)) times CFM at the old pressure drop.
An example of the above formula would be to convert flow numbers taken at 28" of water to those which would occur at 25" of water.
(25/28) = .89286
Using the square root key on your calculator and inputting the above number gives .94489 which can be rounded off to .945.
We can now multiply our CFM values at 28" of water by .945 to obtain the theoretical CFM values at 25" of water.
While this method is very accurate in itself, it does not account for variations in test equipment or procedures used during the tests!
Different flow benches often do not read the same when identical parts are tested on them. A given flow bench may always read low at low flow rates and high at high flow rates, or vise-versa. So long as these test conditions are repeatable, they are no hindrance to the user of that given bench, but if you are trying to compare between that bench and another one, these differences will make for an innacurate comparison. Additionally, test setup criteria can sometimes affect the results of a test. Whether or not an inlet radius is placed over a bare inlet port for instance, can sometimes make a dramatic change in the way a particular head may flow.
Our advice is to compare numbers using the above information with a degree of caution. Just like horsepower numbers (where no two dynos are likely to be spot-on), flow bench results are subject to some degree of variance between units. What is most important for us, is to have repeatable results so that comparisons made on our own bench are valid. We have tested development heads done years prior and found the numbers to repeat indefinitely. Additionally, the use of calibration orifices has shown our results to be very accurate throughout the range (less than 1%), which is considered excellent and plenty good enough for our use.
©2005 Sean Brown
All rights reserved.