Everyone in Spec Miata knows that the 2015 rule set underwent some changes in the offseason. The rule changes themselves weren’t major, but the upheaval that led to them was, particularly in terms of what modifications were permitted on the cylinder heads. Internet forums lit up like Times Square at Christmas.
However, the end result was essentially a clarifying of the existing rule and allowing machine shops to debur the edge created by the plunge cuts on the short side radius of the valves.
We designed an experiment using three heads, a flow bench, a full machine shop and the services of Todd McKenzie Racing Cylinder Heads in Oxnard, Calif. McKenzie has specialized in cylinder heads for more than 30 years, having worked at Air Flow Research and Ed Pink Racing Engines, where he also did prototype work for Edelbrock when the company was developing its line of cylinder heads.
Two of the heads came from 1995 production vehicles. Both were straight and showed minimal core shift. Core shift occurs when the sand “cores” used in the casting process, which create the voids that become combustion chambers, ports, and water jackets, shift as the molten metal is poured into the mold. When the cores move out of place, you get a “core shift,” which means that you get too little metal on one side, and too much on another side of the core. On a Miata head, you can see it best in the throats beneath the valve seats.
The third head in the experiment came from a crate motor from Mazdaspeed Motorsports Development.
We had a theory that the exhaust ports on the heads we had wouldn’t benefit from the permitted “plunge cutting,” and that they would flow better than one with plunge cuts on the exhaust side.
The experiment was to flow test all three heads, then take the best-flowing head and plunge-cut the intake throats, debur the short side radii, make the legal radial valve relief cuts in the combustion chambers, resurface the deck and perform a legal valve job with three angles of 30, 45 and 70 degrees.
The second best-flowing head would get the full complement of permitted modifications, meaning it would get the same treatment as the best-flowing head, plus the plunge cuts and deburring on the exhaust side. Then, after all the machine work had been done, we’d put both heads back on the flow bench to see which one flowed the best. The head that came in third wouldn’t be used or be subject to any further testing.
Our theory stemmed from air flow principles. One, that the bottom of the plunge cuts on the exhaust side would disrupt exhaust flow as it exited the combustion chamber. Another air flow principle has to do with the very reason the rules were changed in the first place: the dynamics of air flow over the short side radius.
“Air does not like to turn, so the more gradually you can make it turn, the better off, the more it stays laminar,” McKenzie said. “If you have a 90 degree edge, air tumbles right off that edge. It doesn’t turn. It just sits there and spins.”
“Laminar” can be described as the smooth flow of a viscous liquid through a tube or pipe. The velocity of flow varies from zero at the walls to a maximum along the centerline of the vessel. When you plunge-cut the throats, it corrects for core shift, but it also makes the short side radius sharp, which can create flow problems of its own.
“You’d much rather have it not sharp than sharp, even if you have to sacrifice throat diameter,” McKenzie said. “If you were to maximize throat diameter, in what little short side radius there is on those heads, if you were to counterbore it bigger, you would have this knife edge, a 90 degree angle that the air flow would have to travel over.”
Even the 1.5 mm deburring of the short side radius that is now permitted doesn’t help air flow much, which is why we thought it best not to plunge-cut the exhaust throats.
Both heads did get the radial relief cuts on the combustion chamber walls, .7600 on the intake side and .6750 on the exhaust. This cut, which is circular and concentric to the valve guide “unshrouds” the valves, which has long been known to improve flow and unlock more power. The cuts allowed by the rules are about as large as you can make them.
“The radial relief cuts are critical,” McKenzie said. “I don’t think you can go much more or you’ll go past the cylinder head gasket and past the bore size. So if you went any larger in diameter, you would expose your head gasket.”
The rules also stipulate that “there must be a sharp edge where the valve relief cut meets the chamber. That edge must be present and unmodified.” The rule reads like that because there is more power to be had by blending the cuts and by opening up the radius a little, and it’s easy to check with go-no-go tooling.
Spec Miata rules also allow for resurfacing of the head to achieve the factory compression ratio so long as the head is no shorter than 5.245 from the deck to the cam cover rail. That leaves just the valve job, which is important on these engines because the valve lift is, well, paltry. No NA or NB Miata has more than .350 inches of lift on the intake or exhaust. “When your valve only opens .350”, port volume, size and shape really don’t have much effect on it,” McKenzie said. “You’ve got a cylindrical volume that’s .350” tall that’s capped by the valve and the seat, and that’s why getting it around the valve is so critical. That’s what makes the valve job so critical. If you’re only dealing with .350” of lift, it’s all in the valve job. If you were allowed more lift, then port shape and volume would come into play a little more.”
So, how did the experiment turn out? Our original hypothesis turned out to be correct. The head that flowed best at first still flowed better with no plunge cuts on the exhaust than the head that did have the exhaust throats plunge-cut. As you might expect, the differences were subtle, but this is Spec Miata, where any power or competitive advantages are measured in small increments. You can look at the chart below to see how the best head fared before and after the modifications.
Since we also had three intake manifolds on hand, we decided to flow test them, too. We bolted all three to the best head and flow tested one cylinder with the intake, throttle body and restrictor plate in place. We found that the intakes restricted flow numbers by about 5 percent. We suspect the restrictor plate is responsible for most if not all of that flow decrease. One of the intakes outflowed the other two by 1.5 CFM. Again, the difference is subtle, but that’s the intake that’s now on the car.
It’s interesting to note that the worst-flowing head came off the crate motor. The key to its dismal performance was the amount of material below the valve seat on the short-side radius. The likely culprit was core shift, which not only can take place side to side, but also top to bottom. The core may have shifted upward in the cylinder head, causing the excess material under the valve seat. You can see the difference in the photos below.
Though we didn’t use the head from the crate engine, we did take all the brand-new cams, springs, lifters, valves, retainers and keepers and install all of them in the best-flowing head. That way, we get more valve lift than we would with used components.
Will you get the same results if you try a similar experiment? It likely depends on the production heads you use, but our guess is that your results wouldn’t be much different from ours.
Now, as if we weren’t dizzy enough from all the testing and numbers, here’s another curve ball. We have heard around the paddock that cylinder heads that flow the best sometimes don’t make the most power. I’m sure the differences are subtle, though.
McKenzie Racing Cylinder Heads