Words And Photos: Richard Holdener
When talk turns to cylinder heads, the very first thing that comes to mind is airflow. This stands to reason as airflow is a critical element in power production and cylinder heads are one of the key components that determine the airflow into and out of a motor. Short of actual dyno testing, it is difficult to directly compared cylinder head performance, so manufacturers and enthusiasts have instead relied on airflow data. Advertised flow numbers have quickly become the primary motivational force in cylinder head selection, but does the flow rate of a cylinder head actually determine the power output of your motor? Using simple math and a few examples, we hope to illustrate that airflow alone, especially peak flow numbers, indicate only the power POTENTIAL of the cylinder heads. In truth, the vast majority of performance engine combinations are likely more of a restriction to the cylinder heads than the other way around.
Don’t get us wrong, airflow numbers are a useful tool when choosing cylinder heads. The flow numbers provide solid information on the power potential offered by the cylinder heads. Why do we use the phrase power potential? The reason is simple. As measured on the flow bench, the airflow numbers indicate only the flow potential of the port. The bench does not simulate the dynamic equation that is an internal combustion engine, nor (more importantly) does it take into account your particular engine combination. A useful tool for calculating the power potential of cylinder heads is the following formula HP=.257 x (peak) airflow x number of cylinders. If we plug in flow data from a CNC-ported small-block cylinder head that offered a peak flow of 304 cfm, we get the following: HP=.257 x 304 x 8 or 625 hp. Put another way, the 304 cfm head would support 2.056 hp per (peak) cfm.
The first question that should come to mind after using the formula is will your small block make 625 hp if you installed these 304-cfm heads? Phrased in that manner, it becomes obvious that there is a great deal more to coaxing 625 hp from your typical small block build than installing the right cylinder heads. The 625-hp heads must be combined with a 625-hp cam, 625-hp induction system and a 625-hp everything else. It is the list of “everything else” that usually causes the small block build equipped with 625-hp heads to make considerably less than the desired 625 hp. Of course the opposite is true, as it is possible to exceed the 2.056 hp per cfm power level suggested by the formula as well, though this is an extreme exception rather than the rule. The other problem with the simple formula is that it does not take into account airflow offered at different lift values. A cylinder head that offers more average flow at each lift value will likely offer more power than one with a similar maximum peak flow. Let’s take a look at a few examples to better understand the role of head flow in power production.
Test 1: Stock vs CNC-Port LS3 Heads-LS3 Crate motor (1.49 hp per cfm)
The results of this test clearly illustrate that power output of the motor is NOT determined by the flow rate of the heads. The test involved replacing the stock LS3 heads on an otherwise stock 6.2L (376-inch) LS3 crate motor from Gandrud Chevrolet with a set of CNC-ported LS3 heads. From a flow rate standpoint, the stock heads offered 306 cfm while ported heads stepped things up with 338 cfm. Using the formula, the extra 32 cfm had the potential to increase power by 66 hp, but the head swap netted just 10 extra hp, from 495 hp to 505 hp. With 338-cfm heads capable of supporting almost 700 hp, why did the LS3 make only 505 hp (1.34 hp per inch or 1.49 hp per cfm)?
The answer to the question can be found in the remainder of the components, primarily the cam profile. Though the peak flow was given at .600 lift, the stock cam used only offered .550 lift. This decreased the relative flow of the heads, but only slightly. The real reason was that the cam profile was not a performance grind capable of pushing the engine to take full advantage of the available airflow. The same can be said of the factory intake manifold and to a lesser extent the compression and displacement of the short block. With 306 cfm, the stock heads on the LS3 already offered the potential to support over 625 hp, so they simply were not restrictive to the current combination. That is why the extra airflow offered by the head swap improved the power output by just 10 hp.
Test 2: Stock vs Ported Heads-Stroker 351 Ford
This test illustrated that the combination under the heads determine how effective it was at utilizing the available airflow. The 393 stroker test motor was used to compare a number of different 5.0L Ford cylinder heads. To establish a baseline, the stroker was first run with a set of stock (E7TE) 5.0L heads. The stock heads in question flowed just 166 cfm. Using our formula (HP=.257 x (peak) airflow x number of cylinders) once again, we see that these heads will support 341 hp. Run on this 393 stroker, the stock heads produced 387 hp, or 2.33 hp per cfm! The reason for this is that the healthy stroker combination was more than able to maximize the available flow.
After replacing the stock heads with a set of AFR 205s, the power output jumped to 555 hp (an extra 168 hp from a head swap)! The head swap improved the specific output of the 393 stroker from .98 hp per inch (387 hp/393 inches) to 1.41 hp per inch (555 hp/393 inches), but the hp per cfm dropped dramatically. The AFR 205 heads flowed an impressive 301 cfm, but the increase to 555 hp meant the 393 stroker was producing just 1.84 hp per cfm. This compares to 2.33 hp per cfm with the stock heads. Obviously we’d rather have the extra horsepower, but it shows that this 393 combination was not wild enough to take full advantage of the available airflow. To maintain the 2.33 hp/cfm level of the stock heads, this 393 would have to produce over 700 hp!
Test 3: 1,060-HP 565-Inch BBC
Running the stock heads in test 2 on a healthy stroker produced an impressive hp/cfm output but this 565-inch big block demonstrated what happens when you combine high-flow heads and a serious test motor. The big block was not only sporting some serious cubes at 565 inches, the combination was built to maximize ever one of the 415 cfm offered by the Dart Pro 1 355 heads. In addition to the inches, the 565 was also sporting 14.7:1 compression, a nasty roller cam offering .850 lift and a tunnel ram sporting not one but two 1050 Holley Ultra Dominators. Also included were power producers like a Total Seal low-tension ring pack and dry-sump oiling system. While the formula suggests the 415 cfm will support slightly more than 850 hp, this impressive big block produced an amazing 1,060 hp, or the equivalent of 2.55 hp per cfm. Producing 1,060 hp per inch, the 565 offered 1.876 hp per cubic inch.
From these results, it should be obvious that airflow, especially peak airflow, is only an indicator of potential power. All the airflow in the world won’t make more power if the combination already has sufficient head flow to support the existing power. The key to making big power is to take full advantage of the available airflow. When matched with appropriate cam timing and the proper induction system, head flow can be used to produce impressive gains, but if you are looking to extract every last ounce of power, you need to leave no stone unturned. Things like compression, displacement and even frictional losses come into play. Using these three examples, we have only scratched the surface of the complexity of cylinder heads. Serious engine builders understand how altering valve sizes, valve angles and even port volume affect flow, and how to properly utilize each on engines of differing displacement. Don’t get us wrong, head flow is important, but it is just one of many weapons in a performance builder’s arsenal.
Graph 1: LS3 Head Test-Stock vs GMPP CNC
This test compared a set of stock LS3 heads to a set of CNC-ported versions from GMPP.
The stock heads flowed 306 cfm while the ported version offered 338 cfm. The head swap netted an extra 10 hp, but the formula suggested an extra 66 hp. The reason for such a small power gain from the heads was not because of any problem with the ported heads but rather the limitations of the test motor. The mild 6.2L already featured 306-cfm cylinder heads capable of supporting the required 495 hp. Adding the additional 32 cfm to the mix was of little benefit since the combination already had more than enough head flow. Thanks to the stock cam, intake and short block, this LS3 crate motor offered a rather low hp/cfm output of 1.49 hp/cfm.
Graph 2: 393 Stroker Ford head Test-Stock 5.0L vs AFR 205
Performing the head swap on the 393 stroker Ford (351W) offered plenty of cylinder head data. The power gains were tremendous, as the 393 stroker was severely limited by the 166-cfm offered by the stock heads. That said, the stock-headed 393 offered an impressive specific hp-per-cfm output, as the motor produced 387 hp with just 166 cfm (2.33 hp/cfm). This illustrated is is certainly possible to exceed the numbers suggested by the power/cfm formula, but swapping on the AFR 205 heads showed that this specific output is the exception and not the rule. The head swap improved the power output from 387 hp to 555 hp, but with 301 cfm, the specific hp/cfm actually dropped to a more realistic 1.84 hp/cfm. Obviously we would choose the higher power level, but it shows that even this healthy 393 was nowhere near equaling the specific output suggested by the formula.
Graph 3: 1,060-HP 565 BBC Stroker
Achieving or exceeding 2.0 hp/cfm generally requires a dedicated race motor. This 565-inch BBC was a perfect example of how you go about achieving these elevated specific outputs. Using just about every performance trick in the book, including a tunnel ram, high compression and a dry-sump oiling system, this 565 produced 1,060 hp using Dart Pro 1 355 cylinder heads that flowed 415 cfm. That represented an amazing specific hp/cfm output of 2.55 hp/cfm.