Disclaimer: Take these air flow readings with a grain of salt. The readings may be 10% high. The ratings may not be comparable to flow ratings from another flow bench. Calibrations,set-up,etc. may be different than other flow benches also. I think the main purpose would be the air flow comparison between the different heads. Steve25 at PY suggested using a 1.37 correction factor to his numbers for 28" of water. Thanks-John
By Pete McCarthyFrom April and June 1991 High Performance Pontiac
Airflow testing Pontiac cylinder heads was
a truly enlightening experience.
Head Flow Charts
There seems to be a fine line between a skeptic and a cynic. A skeptic is willing to believe, but is wise enough to entertain some doubts. A cynic has reached the point where almost nothing is to be believed.
When it comes to the subject of Pontiac cylinder heads, years of bitter experience have put me solidly with the cynics. Placing trust in a stranger can be risky business, indeed. It is small wonder that most of the top-performing racers have found it necessary to eventually perform their own headwork.
Recent developments in the area of vehicle documentation, such as the establishment of Pontiac Historic Services, will very shortly put a merciful end to the confusion and chicanery that have plagued the Pontiac hobbyist for far too long. With this article, I hope to provide a similar service for Pontiac engine builders, who have been subjected, over the years, to little more than smoke and mirrors when it comes to the subject of Pontiac cylinder heads.
You may not like what you will read, but I guarantee the information has been compiled by some of the most experienced and cynical Pontiac devotees in the hobby. There has been no "figure fudging." What you will see is what we got.
The entire project would have been impossible without the support provided by a man I have known and respected for over 24 years. Jerry Goodale, the owner of Specialized Motor Service (2958 Rubidoux Blvd., Riverside, CA 92509:
714-686-3293). Jerry's shop is equipped with every modern high-tech piece of machinery imaginable. The list of name racers who parade through Jerry's shop reads like a "Who's Who" of the racing fraternity.
Tools of the trade
The equipment used consisted of a 400-cfm (cubic feet of airflow per minute) laminar flow element, manufactured by Meriam Instruments (a division of Scott & Fetzer, Cleveland, Ohio): two graduated vertical manometers (Meriam model 30 EF 25) to measure atmospheric pressure: and an incline manometer (Meriam model 40 EF 35) to measure actual airflow. The laminar flow element is powered by six vacuum cleaner motors, rated at 11/2 hp each. To test exhaust flow, the laminar element is simply plumbed backward to "push" rather than "pull."
Our "analysis team" also benefited greatly from the use of some of Jerry's homemade
supplementary airflow devices. One such device is a clever "anemometer." a propeller-driven velocity gauge that can be inserted inside the cylinder tube of the airflow bench to measure combustion chamber velocity, or "swirl speed."
Another test device is a vacuum-sensitive "manometer." whereby a plastic flexible probe can be positioned anywhere inside a port to measure relative air velocity. A few minutes spent with this toy will quickly dispatch any illusions as to how airflow really behaves. The most common utterance heard as different observers used the probe was, "Well, I'll be damned."
In addition to the access granted to all this equipment. Jerry was able to provide the services of one of his journeyman technicians, Tony Sizemore. Fortunately for us, Tony is a die-hard Pontiac fan, and was more than willing to spend much of his own time preparing the numerous heads for testing. Tony is also a student at Chaffey College, and the flow-bench information he obtained will assist him in preparation of an impressive term paper. Later on down the line, Tony's big-journal Pontiac engine will see considerable time on Jerrv's incredible, computerized color-display engine dyno. The combined efforts of the cynic and the student should make for some interesting reading!
Before proceeding with the airflow figures, a few observations about the effects uncovered with the velocity probe may be of interest. The two areas of highest velocity were the edge of the pushrod bulge and the floor of the port, where the short-turn radius makes the turn into the valve bowl. At low and intermediate valve lifts (.100 to .400 inch), the highest point of air stream velocity at the pushrod bulge was centered slightly below the halfway point (vertically). At higher lifts, the air stream begins to move progressively higher toward the roof.
It is obvious that the air takes the path of least resistance, and that path is directed initially from the center of the port opening to the back side of the valve. At higher lifts, the air path acts like a teeter totter. The axis at the short-turn radius remains constant, but the path of the high-velocity air at the port opening begins to move higher, trving to straighten out the path of air aimed at the back side of the dropping valve.
The velocity of air passing over the short-turn radius at all lifts is incredible. It almost tries to swallow the probe. This straight-line air path is the reason that a Ram Air IV intake port is laid out in more of a straight line than a D-port head, and the short-turn radius is low and very abrupt. A port for high-lift applications, then, is designed much differently than a low-lift head. This is the reason that high porting" works with a head originally designed for high-lift applications. If s an effective method to give the air stream the direction it wants. The logical conclusion to these observations is that increased airflow will occur only if the basic flow path engineered into the particular head is enhanced to accommodate the airflow's straight line. Enlarging a port without consideration of where the increased air volume is going is useless. I can't imagine a more senseless exercise than porting a head for a higher-lift flow path, and then using a cam with a lift too low to utilize the capability. The engine will lose whatever power it had, and will inevitably fall on its face.
Setting the scene
Gathering the heads is a story in itself. The common heads were not all that hard to find, but nailing down the more exotic heads (such as the '62-63 Super Duty, the '73 455 SD and especially the Ram Air V) was no easy task. Complicating the job was the problem of finding sets of these heads that were not modified.
Happily, the Pontiac network provided all the help needed. H-O Racing, Bruce "421 Club” Fulper, Joel Larkin, Jim Glaze, Jim Bachman, Richard McFarland and Randy "High Performance Injuns" Repp donated their most prized possessions to ensure that a wide selection of original Pontiac heads would be included in this once-in-a-lifetime event.
A number of witnesses were on hand to verify the numbers, including HPP editor Sue Elliott. She wasn't reluctant to get in the trenches and help out, as well as record some of the procedures on her trusty Nikon.
By the time the project was completed, nearly 1,000 separate procedures had been recorded. The incessant 80-decibel whine of the flow bench droned on for nearly four days.
Discovery is a powerful motivator. time after time, the cold, impartial numbers that flowed back and forth on the sliding green oil scale would blow away our long-held assumptions. In more than one case, the truth was so hard to believe that another two or three heads had to be tested to convince us that the bench was not, somehow, lying,
For longtime Pontiac "experts," the project was humbling indeed.
I had been involved in numerous flow-testing procedures before, but none had begun to approach the sheer comprehensiveness of this session. Needless to add, the degree of education that all of us obtained is priceless.
Should you port your heads?
I find it very difficult to believe that any head can be ported properly without the benefit of airflow technology. Indiscriminate grinding away inside a port with a naive notion that "bigger is better is a guaranteed way to ruin a head. In my humble experience, I have reached the unhappy conclusion that the vast majority of ported heads are automotive jewelry." They shine, they glisten, they dazzle but the airflow numbers have gone south for the duration.
When faced with the reality of diminished performance, the victim is cajoled with a time-tested litany of excuses:
"Well, what do you expect? You need a bigger cam; you need more carburetion; you need more gearing; you need better ignition; etc." It's always your fault, because you don't know anything.
Actually, the only mistake these victims made was porting the head in the first place. If I can give you any advice at all, it is this: Leave head porting for last. There are dozens of improvements to be made to any performance engine before porting becomes necessary. It is far cheaper to buy a stock performance head, even at today's prices, than to risk an equal amount of money on a ported head.
This is not to imply that all head shops are a sham. There are many reputable and dedicated establishments throughout the country. The best method to assess the ability of these shops is not by reputation, but by performance. Find out who runs a car close to your combination and successfully uses a set of ported heads. Then you've got something.
How the tests were done
All testing was done at 12 inches Hg, and airflow figures were taken at valve lift increments of .100 up to .600 inch (higher in some instances). Stringent efforts were made to control air leakage, as any leakage through the valve stem results in inaccuracy.
It didn't take long to discover that all ports are different. It was not unusual to see flow figures vary widely among ports on the same head. A reading of 65 cfm at .100 lift on
one port may be mated with a reading of 57 cfm on an adjoining port, and 69 cfm on still another. The analyst has to realize that this kind of fluctuation is reality. If a test contrasts maximum flow on one head at 150 cfm and 155 cfm on another, the proper conclusion is that the two ports are virtually identical.
I must emphasize again that these tests are relative. There is no such thing as total accuracy when dealing with cylinder head airflow. We'll get as close as we can.
Each head tested was given a grade. The grade indicates the analyst's opinion of the head with respect to the quality of the valve job, the condition of the valves and guides, the originality of the ports and chambers and the plane of the deck. A grade 1 indicates that the head is in top-notch condition in every way. A grade 2 indicates some slight air leakage (1 or 2 cfm), or some alteration within the port that is not deemed to affect normal airflow. A grade 3 indicates seat or stem leakage as high as 6 cfm, or a valve seat that could not be fully restored. Heads that did not meet these minimum requirements were set aside.
Normally, grade 1 heads will exhibit higher flow rates at low valve lift (.100 and .200), because this is the range most affected by the accuracy and originality of the valve seat. At lifts in excess of .300, the valve seat is much less of a factor in relative airflow comparison. Every effort was made to use either new or refaced valves with reasonable margin. In all instances except those noted, the correct original valve supplied with the head was used.
It was not possible to collect every stock Pontiac head. What was done, therefore, was to test every head that fits a general category. For example, the 1965 No.77 head was a functional duplicate of the 1965 No.76. The only difference in these two heads was a slightly smaller chamber volume in the No.77, for use on the 10.75:1-compression 421 HO and GTO engines. This difference in chamber volume would not affect airflow through the ports. All of the "performance" heads were tested individually.
Each airflow test at a given lift is accompanied by a corresponding figure indicating the mathematical exhaust-to-intake port ratio. For example, if an intake port flows 150 cfin at .400 lift and the exhaust port flows 112 cfm at the same lilt, then the Exh./Int. ratio is .746, or .75.
There is disagreement among engineers as to what the ideal flow ratio should be for pure air testing. However, .75 seems to be a number that most would agree is acceptable. This does not imply that a higher number is better. A number higher than .80 can mean that the exhaust is too efficient, with the result that part of the intake charge is scavenged out of the chamber before optimum combustion can occur.
The percentage of exhaust efficiency can vary greatly at different lifts. Most Pontiac exhaust ports seem to be deficient at very low lifts (.100 to .200). However, the
exhaust ratio usually improves to acceptable numbers in the .300 to .500-lilt range. At higher lifts, the ratio often rises well above the .75 level. This does not necessarily mean that the exhaust port is particularly efficient, but rather that the intake port has "tapped out" and will produce no further increase in airflow.
Low-lift exhaust deficiency has been a Pontiac curse for many years. That is why all the Pontiac stock cams, with the exception of the very first profile developed (No.518111 "A"), are dual pattern, with a longer-duration exhaust lobe.
The following graphs reflect the airflow numbers obtained by testing Pontiac cylinder heads in their stock, unaltered form. The flow numbers are listed in one of three ways: single-port, indicated at the top of the chart as S; dual-port (one interior port and one end port), D; and 4-port, or multiport, MP. In some cases, all four intake and exhaust ports have been analyzed. Each listing will be self-explanatory
Each head is identified by year, casting number and general description. The exact date casting is printed for specific identification.
In our last issue, we began our foray into the world of airflow, as
it occurs inside Pontiac cylinder heads. We were armed with a 400-cfm laminar flow element (powered by six 1.5 hp vacuum cleaner motors), two graduated vertical manometers to measure atmospheric pressure, an incline manometer to measure actual airflow, and a velocity probe. To test exhaust flow, we simply plumbed the laminar element backward to "push" rather than "pull."
Many of you may not have liked what you read in April about 1958-70 Pontiac cylinder heads, but we guarantee the information was compiled by some of the most experienced and cynical Pontiac devotees in the hobby. There was no "figure fudging."
And now we're back, with flow numbers for 1971-79 heads, and some conclusions about our test results.
As it was not possible to collect every stock Pontiac head, we tested every head that fit a general category. All of the "performance" heads, however, were tested individually.
Again, we must thank Jerry Goodale, the owner of Specialized Motor Service (2958 Rubidoux Blvd., Riverside, CA 92509; 714-686-3293), who allowed us the use of both his shop and one of his invaluable technicians, Tony Sizemore, an irrepressible die-hard Pontiac fiend.
How the tests were done
To rehash, all testing was done at 12 inches water, and airflow figures were taken at valve lift increments of .100 up to .600 inch (higher in some instances). Stringent efforts were made to control air leakage, as any leakage through the valve stem results in inaccuracy.
It was not unusual to see flow figures vary widely among ports on the same head. Therefore, if a test contrasts maximum flow on one head at 150 cfm and 155 cfm on another, we must conclude that the two ports are virtually identical.
I must emphasize again that these tests are relative. There is no such thing as total accuracy when dealing with cylinder head airflow.
Each head tested was given a grade. The grade indicates the analyst's opinjon of the head with respect to the quality of the valve job, the condition of the valves and guides, the originality of the ports and chambers, and the plane of the deck. A grade 1 indicates that the head is in top-notch condition in every way. A grade 2 indicates slight air leakage (1 or 2 cim), or some alteration within the port that is not deemed to affect normal airflow. A grade 3 indicates seat or stem leakage as high as 6 cim, or a valve seat that could not be fully restored. Heads that did not meet these minimum requirements were set aside.
Normally, grade 1 heads will exhibit higher flow rates at low valve lift (.100 and .200), because this is the range most affected by the accuracy and originality of the valve seat. At lifts in excess of .300, the valve seat is much less a factor in relative airflow comparison. Every effort was made to use either new or refaced valves with reasonable margin. In all instances except those noted, the correct original valve supplied with the head was used.
Each airflow test at a given lift is accompanied by a corresponding figure indicating the mathematical exhaust-to-intake port ratio. For example, if an intake port flows 150 cfm at .400 lift and the exhaust port flows 112 cfm at the same lift, then the Exh./Int. ratio is .746, or .75 (rounded).
The graphs reflect the airflow numbers obtained by testing Pontiac cylinder heads in their stock, unaltered forms. The flow numbers are listed in three ways: single-port, indicated at the top of the chart as S; dual-port (one interior port and one end port), D; and 4-port, or multiport, MP. In some cases, all four intake and exhaust ports have been analyzed. Each listing will be self-explanatory.
Each head is identified by year, casting number and general description. The exact date casting is printed for specffic identification of each head.
If any observation can be drawn from this chronological analysis, it is that if a head casting number was changed, something was done to the head. These changes, in many cases, were extremely subtle. Between the flow numbers and close inspection of the heads themselves, it has been possible to identify some of the more obvious directions that the factory took from year to year, as the chief engineer's office gave the designers their "marching orders."
Let’s dissect a couple of the more apparent differences. The 1971 No. 197 455 HO head (referred to in error as the No.191) was detuned slightly from the 1970 Ram Air IV heads. It is not a “big-chamber" version of the Ram Air IV head, except in general appearance. The intake port volume was reduced some l2cc's. The result was a diminished airflow capacity above .300 lift.
What was the reason for the reduced airflow? The engineers knew the camshaft to be used was going to be reduced in lift. The port was therefore engineered to provide maximum performance at low lift. The relatively high numbers at .300 to .500 indicate that the Ram Air IV cam, with 1.5:1 rockers, may have been considered as a real possibility before the smog lobby prevailed.
The exhaust port throat in the 1971 HO head is noticeably different from the Ram Air IV. The bottom of the HO exhaust valve guide (and roof) is at least 1/4 inch closer to the valve seat than that on the RA IV. The radius, and the turn into the exhaust port itself, is lower, and the exhaust port volume is some 8cc's smaller. This was done to increase the off-the-seat velocity, and was engineered to complement the somewhat reduced intake flow. This effort was not particularly successful, having had the principal result of killing off whatever flow numbers there were above .300 lift.
The 1972 455 HO head was altered even more significantly in the exhaust port. At this juncture, the engineers were certain that the days of the .500-lift cam were over. The exhaust throat in the '72 HO head was re-engineered to produce even more velocity off the seat, and the high-lift capacity (over .400) was even further reduced. The intake port, however, was left alone.
Among the extreme subtleties is a difference between the 1969 Ram Air IV head (No.722) and the 1970 Ram Air IV head. The factory managed to increase maximum airflow in the 1970 intake port by some 20 cfm, and got it all at .400 lift, by increasing the port width framing the short-turn radius by a mere .017 inch on either side. This startling lesson in airflow behavior indicates how naive it is to begin grinding before one understands what port shape is all about.
I have no doubt that these figures are going to cause some consternation among a number of HPP's readers. The "I told you my head is better than yours" contingent should have a field day.
Before readers jump to conclusions based on airflow numbers alone, however, they must take into account all the other factors that compose a total engine combination. A well-set-up 400-cid engine with big-valve No.16 or No.48 D-port heads is going to run just as well as another engine with Ram Air IV heads, if the tuner doesn't have enough cam and gearing to utilize the higher-lift airflow
I'm sure our readers are quite capable of drawing their own conclusions with respect to some of these heads. I cannot resist the temptation, however, to list a few of my own opinions (see “Pete's Picks"). Whether you agree with these assessments or not, I would imagine that most Pontiac die-hards will find the numbers researched to be of great interest.
We'll get into the complexities and complications of porting various Pontiac heads in the near future. I promise you that the surprises will be many.
Best performance head 1970 Ram Air IV
Best intake port 1969-70 Ram Air V
Best exhaust port 1 968 1/2 Ram Air II
Best D-port head No.16, No.48, No.12 (tie)
Best low-compression D-port head No.96(1971)
Best low-compression post-1972 head No. 6X
Best balanced head (exhaust to intake) 1963 421 SD
Best low-lift (under .400) head 1967 No.670
Worst exhaust-to-intake port ratio 1969-70 Ram Air V
Worst intake-to exhaust ratio 1968 1/2 Ram Air II
Biggest surprise Intake port, No.17 350 head
Biggest disappointment 1969-70 RA V, 1973-74 455 SD (tie)
Biggest "sleeper" 1975 No. 5C
Most undercammed 1963 SD, 1971 455 HO, 1973-74 455 SD (tie)
Most underexhausted 1964 GTO (No.9770716)
Most potential for porting 1973-74 455 SD, 1968 1/2 RA II (tie)