Peer under the hood of the fast cars at the strip or the local burger joint and you'll see the AFR logo etched into more than a few cylinder heads. Racers demand proven performance, so it's no coincidence that many of them turn to Airflow Research for their cylinder heads. Almost forgotten due to the company's overwhelming success in the street/strip arena is its decorated history in the most competitive ranks of professional racing. Over the years, AFR has powered top drag racers such as Bill Jenkins, Warren Johnson, and Bill Glidden to the winner's circle. Likewise, NASCAR legends Richard Petty, Darrel Waltrip, and Cale Yarborough all racked up multiple championships thanks to AFR cylinder heads. In other words, these guys make some darn good stuff.
Unlike other pack leaders of the industry that prefer to stay mum, AFR has never been shy about sharing its expertise with the public. The company's head R&D man, Tony Mamo, visits popular online forums on a daily basis. In addition to battling online myth-spreading with decades of head-porting experience, he encourages consumers to independently flow-test the heads they purchase and challenge the veracity of manufacturers' claims. His candid and unorthodox approach has earned him a reputation as one straight-shootin' SOB, so we're giving him the space to do what he does best: tell it like it is.
"The first step in designing a new cylinder head is to clearly define the direction and goals of the product. For our Gen III head, we wanted to produce a more efficient cylinder head that would make significantly more power but retain the factory port locations and valvetrain geometry. However, keeping everything in the stock location poses a greater challenge in port design because you're not working with any inherent advantages such as a raised intake floor, taller exhaust port, or raised intake roof. In the case of our BBC product, we decided to raise the exhaust port height to help facilitate a port design capable of exhaling enough flow to cover the much improved intake port. This would have been impossible without raising the port, as the height of the factory exit simply makes for too sharp of a turn, and no matter who designs the port, the laws of fluid dynamics would simply never allow a substantial amount of airflow with that sharp a bend."
Gen III cylinder heads flow exceptionally well due in part to their extremely flat valve angles, but Tony feels that too much emphasis can be placed on valve angles. "I'm certainly not going to say a flatter valve angle isn't advantageous, because it is, but the chamber design itself, the port design, the valve job, and the profile of the valve will have a far greater impact on flow," Tony says. A flatter valve angle doesn't guarantee more flow or additional power. It's the package as a whole that must be taken into consideration. Port architecture and location are much more important than valve angle. Much of the airflow benefit of moving from a 23- to an 18-degree head is attributable to better port geometry and layout. "Instead of a 90-degree turn in the intake port, an 18-degree head has a near-straight shot to the intake valve in addition to a higher intake floor, a raised port roof, and much larger valves; it's just a completely superior port layout on top of the fact that it has a flatter valve angle and better chamber design. Another factor to consider is the point of diminishing returns. The factory LS1 15-degree valve angle is already an extremely flat valve angle, and moving it a few degrees flatter doesn't have nearly the same impact as it would in an application that started at, say, 23 or 26 degrees like the old-school SBC and BBC, respectively."
Performing additional port work yourself on an out-of-the-box CNC head will be a mistake 98 out of 100 times. "Unless you have a tremendous amount of experience and a flowbench at your disposal, save yourself the aggravation," Tony opines. Any changes made to a highly optimized port will more likely have a negative effect than a positive one, which is why it's as good as it is right out of the box. If you need more flow, opt for the competition version of the product you are interested in or perhaps the next larger head in that lineup. "Don't think just making a port larger will guarantee it flows more. It's the classic mistake and assumption most people make." In fact, when it comes to making power, bigger isn't necessarily better even if it does flow a little more air. "The shape of the entire flow curve and the velocity of the charge are far more important than a few more peak cfm. I have personally flow-tested dozens of 230-plus-cc heads that don't flow as much as smaller, 195-205cc heads. When it comes to port design, shape is far more critical than size."
Angle of Attack
The vantage point the incoming air has to the back of the intake valve is sometimes referred to as the angle of attack. Due to its added height, a raised-runner head simply has a better angle of attack or vantage point for a straighter shot to the back of the valve. The added height also reduces the angle of the turn that the incoming charge must negotiate. "The geometry of a raised-runner port is superior to a similar nonraised runner design, since it allows additional airflow and a higher terminal velocity before it stalls or backs up," Tony explains. Another overlooked benefit to any raised-runner design starts before the cylinder head. "Due to the higher port inlet locations, manifold runner length is naturally increased, since the space between the left and right port banks is increased, which allows the manifold designer more room for a smoother turn radius from the plenum in an X-plane-style intake. In addition to having runner shape advantages, a raised-runner intake manifold has a much better approach angle from the manifold exit to the runner entrance of the cylinder head. In order to run our 23-degree, raised-runner design, our heads require a shaft-mount rocker system as well as a dedicated intake to properly align the raised intake ports."
Trial & Error
After establishing design goals, the next step in producing a new head is developing the most efficient, highest-flowing piece that meets the design parameters. This step requires all the intense flowbench time and constant trial and error, especially when trying to push the envelope of flow in relation to port volume. AFR does not use any computer port modeling. Designing ports is still done the old-fashioned way, by hand, which requires a lot of time, creativity, and dedication to the project. "Various valve jobs are experimented with, and every shape and contour of the port is addressed to try to achieve strong numbers at every lift point, not just at peak," Tony explains. "For example, our new 23-degree 195cc Eliminator street head (competition version) flows slightly more than 300 cfm at only 0.600 lift. Many aftermarket race heads that are 30 cc larger can barely muster that number by 0.700 lift, and some never get there at all, and that same 195cc port is already flowing 260 cfm as early as 0.400 lift. The entire curve is explosive, and that's one of the key factors in designing a cylinder head that's going to produce big power."
Since tiny variations in cylinder heads can make a huge difference in performance, maintaining tight quality control is essential. Tool life in the CNC machines is of paramount importance at AFR and is constantly monitored. An alarm will sound when it's time to change the tools, after a predetermined number of cutting hours are accrued. The carbides used on valve-job machines are also replaced at regular intervals. Valveguides are checked at the CNC machines quickly, using simple but very effective "go" and "no-go" pilots and are checked once again by AFR's QC department with a much more accurate electronic tool that records data in microns. "There are QC checks at all points through the shop, and our operators play a large role in the QC of our product," Tony explains. Anything questionable is immediately pulled off the line. Spot flow checks are also done routinely to verify that they meet or exceed AFR's advertised flow figures. The concentricity of valve jobs is also checked to make sure all the products seat the valves in their respective seats 100 percent of the time. "The bottom line is we have a lot of procedures in place to help ensure we ship a product that meets a long list of criteria. It's our opinion that QC really starts with the employees and the passion for the product they produce."
Improved software and machinery are two factors that have enabled AFR to replicate the design of a port on CNC machines much more precisely compared to just a few years ago. Once the digitizing of surfaces, programming, and tool paths are finalized, repeatability becomes the primary concern. This requires maintaining proper tool life and monitoring CNC tool and fixture coordinates, valve-job diameter and heights, and other small details that can affect the flow of the heads. The rest of accurately replicating ports is attributable to good old teamwork. "Handling most of the design chores here, I will work directly with our engineering staff during the last phase of developing the ports and CNC programs to achieve a part that's as good as-and occasionally better than-the prototype it was copied from," Tony explains. "Once the hand-ported prototype is digitized, it allows us to be creative with that optimized shape in ways that would be virtually impossible by hand. This is how it becomes feasible for us to occasionally produce a CNC part that's actually better than the original, but trial and error is still the order of the day. This last stage can still require a great deal of time, but the end result is the ability to mass-produce an extremely efficient port design."
"You have to always be looking for ways to improve your product. If you sit back thinking you have the best mousetrap, someone is going to catch you by surprise and produce a better one. We try to stay ahead of the curve by producing a truly exceptional piece that sets the bar two rungs higher, but due to the fact that most of us here are racers, we are never satisfied and are always looking for creative ways to improve upon what we already have. The redesign of our 23-degree SBC head, called the Eliminator, was a huge undertaking. Nothing from our former 23-degree line crossed over, as this was truly a clean-sheet design. The castings were redesigned both externally for aesthetics and internally for function with features like stronger rocker stud bosses and improved oil flow. We also took the recent advancements learned from our Gen III product and incorporated that into the new Eliminator line, specifically lightweight 8mm valves coupled with smaller and lighter springs, retainers, and locks. All of this helps to significantly improve valve control and performance, and extend the engine's operating range. Of course, the real meat and potatoes of any cylinder head is related to airflow, and the new Eliminator castings were shaped around all-new port and combustion chamber designs, ultimately allowing us to bring to market the most efficient 23-degree head we've ever produced. Instead of forgetting about the 23-degree segment of the market and focusing more on the newer Gen III/IV products, we embraced it and decided to swing for the fences by making it significantly better than before."
"All AFR cylinder heads start as raw ingots of A356 aluminum poured at approximately 1,400 degrees. The mold it's poured into is maintained at a specific temperature as well. The casting process is a bit of a black art, and modern foundries are very specialized in what they manufacture. Once the casting is created, the next step is to drill pilot holes that allow us to accurately affix the castings to our CNC fixtures. Every machining operation except the actual CNC porting-such as drilling bolt holes, machining rocker-stud pads, and milling deck and flange surfaces-is done on one of our very large, five-axis, dual-pallet CNC machines. The twin-pallet machines work on a dozen castings at once. Before making their way to the smaller five-axis CNC machines for final porting, hardened seats and bronze guides are installed. Next, the dedicated porting machines handle the combustion chambers and intake and exhaust ports as well as some of the cylinder head engraving. At this point, the heads are pretty much finished but still need to still get a proper valve job, which is a very critical component of flow. The valve-job shape, diameter, and finish depth are all important. From there, it's off to the wash room, where the heads are finally deburred and thoroughly cleaned. Once cleaned, they are transported to our assembly room, where our operators carefully assemble the head with the correct springs, valves, and hardware. It's here that our final QC operator checks every head before it's placed in a box, confirming that the proper parts and machining operations have all been addressed."
Although some consumers equate larger ports with larger flow, AFR strives to keep port volume and cross section to a minimum to promote air speed and cylinder fill inertia. "If you think of air as a liquid, it's easier to understand how a faster column of air will have more energy and inertia to better fill cylinders, with far less reversion issues to boot," Tony explains. "This approach to cylinder head design is especially true if we are focusing on a product aimed more toward a street/strip application than for race use. A reasonably small port that has a strong airflow curve just makes for an explosive package that you need to experience to really appreciate. If you've driven lazy big-port heads and cam stuff all your life in your daily driver and step into a package I just described, you'd never go back." According to Tony, with the proper traction and gearing, that type of combination can yield quicker e.t.'s, since so much of drag racing is about what happens in the first 60 feet.
Porosity in aluminum castings was more common years ago, and much of it was attributed to pour technique and speed, as well as how the mold was vented. "While the material itself can play a role, it was the actual pouring process that was a larger issue, as the mold must be heated prior to the pour," Tony says. "How the mold is designed also plays a role, and additional risers in the casting help provide a solid, nonporous surface. Technology and technique march forward, and while porosity may always occasionally rear its ugly head, most modern foundries have it figured out by now. Of course, in the event we machine a head that's porous, it's immediately flagged and pulled off the line so no additional machine time is wasted on that particular casting."
A flowbench is like a dyno and each can read a little differently. Most shops are honest but may not have properly calibrated equipment or the right accessories-including bore fixtures and radius plates-to ensure accurate results. The type of flowbench used is also important. "I know of a very well-respected shop using a homebuilt flowbench that uses an electric supercharger to move the air," Tony says. "The numbers on that particular bench were too stingy below 275 cfm, about right in the 300-350 cfm range, and optimistic above 400 cfm when compared to a SuperFlow SF600. Was it a useful tool in engine development for that particular shop? Very much so, but trying to compare numbers from that bench to other flowbenches would be somewhat of a useless exercise if you were really looking for exact comparisons." Even the same type of bench can read differently from one to another due to calibration differences, bore size, using clay versus a fixed-radius entry, and flowing the exhaust with or without a pipe. "Ultimately, flow numbers drive cylinder head sales, so buyer beware. Strongly consider getting your heads independently flowed on a known calibrated machine. It's far cheaper and less aggravating than trying to figure out why your car didn't dyno as high as you would have liked or get down the track as quickly as you hoped."
"Don't forget that your cylinder heads' intake ports are at the mercy of your intake manifold design. If you really want to learn about your induction system, start flowing your heads with a manifold attached, and I promise it'll be a rude awakening and you'll think a bit differently about cylinder head design. It's all about the proper combination-that's a statement most engine builders will never argue with. The wrong intake choice can hurt the performance of a combination. For instance, our new Eliminator street heads flow so well that they're best matched with a race-style single-plane, such as an Edelbrock Super Victor 2925, versus a Victor Jr., which has been a popular street-oriented intake for some time. However, packaging constraints may be an issue, and there may not be sufficient hood clearance to run a larger, taller, and generally more effective race piece. For a killer street engine, don't overlook the modern, tall, dual-plane designs, such as Edelbrock's RPM Air Gap. These have come a long way and now barely give up any high-rpm power to a street-style single-plane, while managing to add 20-30 lb-ft across the bottom and middle of the power curve."
Low- & Midlift Flow
Peak flow numbers sell heads, but low-/ midlift flow is far more important for a street motor. Much of the secret to finding good low- and midlift flow lies in combustion chamber design, valve-job profile, and the actual shape of the valve itself. "Back-cut valves are a must, and time must be invested in trying different angles as well as different widths of those angles," Tony says. The actual width of a 45-degree seat needs to be considered as well, while the short-turn height and shape also play a smaller role. "It's a give and take, and really strong peak numbers can cost you a lot of low- and midlift flow while really strong low- and midlift numbers may knock more off the peak than you may be willing to sacrifice. In most of the products we manufacture, our design goal is to produce industry-leading low- to midlift results yet retain solid, respectable peak numbers. Part of this reasoning is the fact that a lazy port with big peak numbers may ultimately be handicapped by the intake manifold choice anyway, leaving you with weak low- and midlift performance and only so-so peak flow where you were expecting that combination to shine."
You should always flow the cylinder head in question with as close to the exact bore size you plan on running in your engine combination. "While it won't make a huge difference, too small a bore can hurt numbers by 5-10 cfm, and too large a bore can artificially inflate them by approximately the same amount," Tony explains. "It's always best to be within 0.030 of your actual bore size. That's only 0.015 per side, so being a little bigger or a little smaller in that range won't matter very much. Once you have over a 0.100 mismatch, the numbers are going to start getting skewed and won't be very representative of real-world performance with your particular combination." CHP