Like it or not, cylinder head design is a subject that's very complex, and it's impossible to make it simple. Considering that no single component on an engine affects power output more than the cylinder heads, it pays to get educated. As a testament to the blistering pace of cylinder head development, the typical street/strip heads of today can oftentimes blow race heads of just 15 years past into the weeds. While the average hot rodder has benefitted greatly from the extremely competitive nature of the cylinder head business, the problem is that the people in the know aren't always willing to talk. After all, if you know something that gives you a leg up on the competition, why share your trade secrets for free? That's the question our sources asked us over and over again as we harangued them for info in order to compile this story. Fortunately, we somehow got them to divulge their secrets on topics ranging from port location to valve angles, and combustion chamber design, port shape, air speed, short-turn radius shape, to the importance of peak cfm.
Our daring cast of experts includes Tony Mamo of Air Flow Research; Kevin Feeney of Racing Head Service (RHS); Al Noe of Trick Flow; Tony McAfee of Dart Machinery; and Darin Morgan of Pro-Filer Performance, who's the former head R&D man of Reher-Morrison's Pro Stock engine program. As you gawk over that impressive list of all-stars, we'll take a moment to introduce you to CHP's newest department. The goal of our "How it Works" section is simple: tackle the intricacies of complex technical topics in a scope never before seen in a magazine article. Our two-pronged approach for accomplishing this is dedicating lots of space to accommodate meaty tech info and letting the real experts of our industry explain things in their own words. CHP is merely acting as the vehicle through which this otherwise-elusive information is gathered and disseminated. We're going so in-depth, that we only got to address a couple of the cylinder head topics we set out to answer before running out of space. Stay tuned for next month's issue, when we put the wraps on our cylinder head exposé. In the meantime, feel free to send in your suggestions on what tech subjects you'd like to see covered in "How it Works" in greater detail.
CFM and Velocity
Darin Morgan: "Make no mistake, velocity is the primary variable in the design of the entire induction system. I often say that my job title should be Velocity Manager instead of Cylinder Head Designer. Air speed is 10 times more important than raw flow numbers. If you kill the velocity by 10 percent, you will kill almost 40 percent of the wave and ram energy that dynamically fills the cylinder! Raw airflow cfm is an important variable as well; it's just not as important. If you buy a cylinder head that is properly sized for a flow of 400 cfm and your engine is only asking for 350 cfm, you will not only fail to achieve the power potential that the 400 cfm would have given you, you will also fail to reach the power that the 350 cfm would have given you. That's because you killed all the air speed in the induction system. If your engine is asking for 350 cfm and you give it a properly sized cylinder head flowing 350 cfm, your airflow demand is matched and your air speed is matched. You then have a chance of achieving the power potential that 350 cfm can give you.
"How much power potential can 350 cfm give? Well, that depends on a host of variables such as engine speed, overall induction system design, and piston speed. To put it in basic terms, the less restriction you have in the induction system and the more freedom you have to attain increased engine speeds, the easier it is to extract the full potential of the 350 cfm available. Most people don't know how much airflow their engine is actually asking for! This leads to builders wanting to purchase cylinder heads with way more airflow than their engines can possibly use. The end result is a low air-speed induction system that can't properly fill the cylinder by means of dynamic inertia and harmonic supercharging, which means the engine will never reach its full power potential.
"That said, a good cylinder head port design will flow a lot of air for its valve size. The bad news is that a bad port design will flow just as much if not more air! Airflow alone won't tell you if a port design will reach its power potential with 100 percent certainty. Everyone knows that it's easy to compare two 23-degree small-block Chevy heads with 220cc ports. Just pick the one with the most flow, right? That's about all the average builder can do, and in a lot of cases it's hit-and-miss. There are multitudes of ways to achieve that 220ccs. You can have a big pushrod pinch section and a very small bowl area, or a huge bowl area and a super small pushrod pinch area. One 220cc port can actually be choked off at the pushrod, short-turn radius, or throat area, hurting top end power. Another 220cc port design can have too small of a bowl area and too large of a choke and hurt power and torque equally across the entire power range. Having extra airflow isn't always bad, but it can't come at the expense of air speed. The ports must be sized properly. The amount of air Pro Comp Eliminator engines are asking for are exactly how much the heads flow, and that's not a coincidence. People want to make cylinder head design simple, but it's not. It's very complex and interdependent on a massive amount of variables."
Tony Mamo: "Having the most flow with the largest runner is certainly not always the best choice in cylinder heads. Perhaps a good rule of thumb or generalization is to choose the smallest head that flows enough air to meet your goals and properly feed your combination at the desired rpm range. Having more head than that will needlessly make the combination lazy and softer in the lower rpm while providing no appreciable benefit upstairs. A common issue I see a lot of is customers putting too large of a head on hydraulic roller combinations, which are typically limited to less rpm due to valvetrain control issues. A 210cc head on a 350 may work very well in a solid roller combination turning 7,500 rpm, but it's the wrong choice in a street car with a hydraulic roller valvetrain that limits usable rpm to 6,500 or so. A 195cc head would be a much better choice and would be a lot more crisp at lower rpm. It's all about choosing the right tool for the job.
"In high-rpm race applications, we aren't so concerned with velocity as much as we are with having a lot of airflow to feed a very hungry engine with a big appetite at very high engine speeds. This assumes that you have the right cam profile and valvetrain hardware to match, but the bottom line is that a head that can provide similar airflow through a more efficient design, potentially with a smaller port, would still make more average power and torque and propel the car faster down the racetrack, most notably improving 60-foot times. However, even in my high-rpm, take-no-prisoners race example just mentioned, the basic elements of total available airflow to be processed is still a balancing act that needs to be properly managed. The highest flowing head may not perform the best if it only flows 10 more peak cfm of air, but has worse low- and mid-lift flow and is 30ccs larger than a different design, which is smaller and more efficient with a stronger curve through the low- and mid-lift range."
Al Noe: "One very important characteristic of airflow is not only peak cfm, or relative velocity, but where in the lift curve those events happen. Are you looking at flow numbers at 0.100- to 0.200-, 0.300- to 0.500-, or 0.600-inch lift and beyond? Are you building a street car, or an 8,000-rpm race engine? The head has to be tailored to what it's being used on, and if this is done properly, the higher-flowing head will almost always make more power. However, we need to clarify what we mean by higher flowing. At Trick Flow, our goal is to maximize the airflow from 0.300- to 0.500-inch lift, and for typical street cars we're not concerned with flow beyond 0.700-inch lift. That's because the majority of hot rodders use cams that peak in the 0.500- to 0.650-inch range. We also develop airflow with a different flow bench from most, and we feel this makes a difference as well.
"The next thing you must consider is the valve diameter. A larger valve diameter will almost always produce higher 0.100- to 0.200-inch flow figures, which can be counterproductive to making power. Larger valves also tend to be more shrouded and have trouble with the all-important mid-lift airflow from 0.300- to 0.500-inch, but then these larger valves tend to shine at the highest lift points simply due to shear volume. For race engines we use larger valves, but are able to reduce the 0.100- to 0.200-inch flow even further with steeper seat angles. We then try to maximize airflow from 0.400- to 0.600-inch. As rpm increase to 10,000 the intake and exhaust port shapes become very critical for power production. Lower-rpm engines don't seem to be as sensitive to the port shape as higher-rpm engines. Our customers are generally looking to make peak power by 8,000 rpm or less, so the importance of flow numbers are still very relevant in their search for power.
"The velocity around the circumference of the valve is actually more important than the velocity in the port. We feel having equal localized velocities around the circumference of the valve is far more important than having equal velocities in the port, and as these velocities are equalized at the important lift points, the port will flow more air at these same lift points as well. We can measure the velocity of the air around the circumference of the valve every 45 degrees, and having these velocities equalized during the all-important mid-lift flow area yields max airflow and best power.
"The better measure of head efficiency is computing the coefficient of discharge. This is simply the airflow at each lift point divided by the valve curtain area, which is valve circumference multiplied by lift. Let's say we have two heads that both flow 300 cfm at 0.400-inch lift. One has a 2.200-inch valve and the other has a 2.100 valve. If you multiply each valve by Pi to get the circumference, you can then multiply that figure by the lift to obtain 2.76- and 2.63-inches, respectfully. Now divide the 300 cfm by each valve circumference to obtain 108.7 and 114.0 cfm/inch, respectfully. I know that is an odd unit of measure, but it is the correct terminology. In this example, you can see the smaller valve clearly has more velocity, will be easier to cam, and will generally run faster. This is the best way for the average consumer to compare two heads to one another."
Tony McAfee: "The area of a port dictates how much flow is capable through a given orifice size. We probe different parts of port in order to measure velocity in feet per second. People often use too big of a cylinder head on a motor. Our 227cc small-block Chevy heads flow 50 cfm more than our 180cc as-cast heads, but won't make as much power on a typical 350 because the engine can't use the head's flow potential. Now if you put a bigger cam in it, raise compression, and install a better intake manifold, then the 227cc heads will work much better. Generally, as the rpm range of an engine goes up, the more air it can use. The more cycles per second a motor turns, the more velocity and airflow it needs. High-velocity ports aren't as important, but are still necessary on a high-rpm motor. After about 350 feet per second, fuel becomes detached from the air, so there is a tradeoff. Sharp areas near the valve job helps put the fuel back into suspension. This is important, because even with good atomization, some fuel still runs down the walls of the port. That said, without testing it's hard to figure out what head you really need. Your best bet is asking people with experience building similar combinations to yours for advice."
Darin Morgan: "The short-side radius is a big factor and is tied to the main overall advantage of raised intake runners. By raising the port entrance, you're taking air speed velocity load off the short-side radius. The lower the port gets, the more the air speed at the short-side radius goes up, and the more critical its shape becomes. Lower ports force you to lower the air speed to get the air to negotiate the short turn, which wastes energy. Ideally, the short-side radius needs to be shaped to have the highest air speed throughout the rpm range without disrupting the boundary layer. In essence, the short-turn radius controls torque, mid-range, and top end power. If you stand it up and make it really fast, you will increase torque and kill top end. If you lay it back and reduce air speed, it will kill low-end torque and increase top end power. Not only do raised intake ports offer more latitude in shaping the short-side radius, they also yield smaller fast-burn combustion chambers and offer a better overall induction system path and design. The higher ports move more air through the engine since the airflow path is straighter, and enables a properly tuned intake manifold runner to be aimed directly at the carburetor booster. The straighter flow path keeps the air/fuel mixture in suspension far better than any low-port flow path ever could, since there's less frictional loss and fuel fallout. An airflow mixture entering the combustion chamber from a straight high-port induction path is more homogeneous and burns faster, producing more power with better BSFC numbers.
"The first three rules to designing a port are air speed, air speed, and air speed. The forth rule is port shape. When I say port shape, this encompasses the geometric shapes throughout the port as well as port wall angles. The floor height and roof height are very difficult to define, and it changes depending on the engine and its airflow requirements. You basically choose roof and floor heights that will give you a slight converging angle to the short-side radius. The roof and floor will converge anywhere from 4 to 8 degrees on high-port heads. This gives area for port bias and a wider bowl without decreasing velocity too rapidly. As the roof and floor converge, the side wall expands and gets wider at the same rate the roof and floor converge. This is the way we change shape without decreasing velocity. Certain shapes will also let us change shape as the port gets deeper with a smooth decrease in air speed as it approaches the short-side radius. The air speed must decrease to turn the short-side radius but not too much. With high ports, we bias the ports less, but with low ports like the conventional 24-degree BBC heads, you always see a drastic bias in the port over the short-side radius."
Tony Mamo: "The bulk of the reason why a raised-runner head usually shows better results than a non-raised-runner head is that its straighter intake ports make it easier for air to negotiate the short-turn radius. What you might find interesting is our new 235cc SBC head is a non-raised-runner design, and has already made more power on the dyno than a larger-but-older AFR raised-runner head, even in an 8,000-plus rpm race engine. Our new 235cc design has an explosive curve through a moderate cross section compared to the raised-runner design with peak flow numbers of just 10 cfm less. It was a very interesting dyno test and proves that you need to use caution when making blanket statements that raised-runner heads will always make more power. Nonetheless, there is no question that a raised runner improves your ability to get more airflow from a particular head design, and it also straightens out the angle of the intake manifold exits as it directs the air into the cylinder head."
Al Noe: "Raised intake runners make a "line of sight" port more easily obtained, which helps a high-rpm head make power. A head designed to make peak power at 8,000 rpm or less is less sensitive to how high the runner is raised. Everyone assumes higher is better, but if a port with a lower entry has better mid-lift airflow, it will usually make more power. A very interesting comparison of port entry location and power potential is a 363ci drag race engine we built that made over 800 hp. It did this with a standard intake port location 20-degree head, and also with raised-runner 15-degree head. Interestingly, these two heads were less than 5 hp apart in power production, despite the fact that most people would expect a much larger difference. As the valve angle gets flatter, it is usually necessary to raise the runner to keep high-lift port stability in check."
Bridging the "Conventional" Gap
Darin Morgan: "Cylinder head and port design is moving forward at a blinding pace, and heads are becoming more specialized. Customers want more power for less money, and this is what spurs on the development of higher-flowing and more efficient heads. The conventional 24-degree big-block Chevy category of cylinder heads has become blurred and will continue to get hazier as time goes on. Some of these 24-degree conventional BBC heads are now flowing as much as some spread-port heads! For example, I designed a 24-degree head for a class called Texas Pro Stock. Its rules call for 24-degree conventional BBC heads rolled to no less than 22 degrees, and allow for a cast single-plane intake manifold and dry-sump oiling. Most of these engines are 565 ci and the last set of heads I did-called the Profiler Performance Sniper X 24-degree heads-had 2.375-inch intake valves and flowed 385 cfm at 0.500-inch and 520 cfm at 0.900. That's way more than I thought was possible just 10 years ago. That engine produced 1,230 hp at 8,400 rpm, and propelled a 2,300-pound car down the eighth-mile in 4.52 seconds. Think about that for moment and let it sink in.
"Now let's look at the pros and cons of a set of spread-port heads versus a set of 24-degree conventional BBC heads with the exact same airflow curve. Just because the heads flow the same doesn't mean they will make the same power. The induction flow path of the 24- or 22-degree conventional BBC heads is less than desirable and the combustion chambers aren't as efficient. The spread-port heads have a straight port and a better induction system flow path aimed right up at the carburetor venturi. They also have very efficient combustion chambers. Even though they have the same flow, the spread-port heads can produce another 100-125 hp."
Kevin feeney: "Conventional heads with standard valve angles and port heights have come a long way in recent years. For instance, with today's 23-degree small-block Chevy heads, you can produce comparable power to 18-degree heads at a far lower cost since they don't require the expensive custom pistons, valvetrain hardware, and intake manifold. The downside is that these heads don't lend themselves to as small of a combustion chamber as with 18-degree style heads, requiring a tall-dome piston in a high-compression engine, which results in the need for more timing and a heavier piston. With that being said, RHS has helped bridge this gap with a few small-chamber 23-degree-style head options where the valve angle has been rolled slightly. This results in a smaller chamber while still utilizing off-the-shelf 23-degree components."
Al Noe: "A standard small-block Chevy head suffers from pushrod pinch and valve location in the bore. Compared to a 23-degree head with standard location intake ports and rocker arm and pushrod layout, an 18-degree head will make substantially more power. Most 18-degree SBC heads have been stuck in the 330- to 340-cfm range for many years, and we have yet to see anything go above that. If a 23-degree head has the valve relocated closer to the center of the bore and a shaft rocker is employed to eliminate the pushrod pinch, then the difference in power production between the two is much closer. The advantages of a 23-degree head are piston, rocker, and intake availability. When you make the jump to a raised-runner 23-degree head with relocated valve centerline and shaft rockers, you have to get custom pistons, custom rockers, and an 18-degree type of intake manifold. In this case it would make more sense to simply run an 18-degree head unless class rules dictate that a 23-degree head must be used. We chose to develop our CNC-machined 230cc Super 23 heads well over two years ago to support the big-cubic-inch SBC market. We did not want to make customers buy special rocker arms, intake manifolds, and headers in order to realize huge gains in airflow. This head moves enough air to support 650-plus hp out of the box, and does this with conventional valve and port locations."