What do you think your heads flow right now? Or better yet, how much do you think your heads need to flow to make the power you want? Well, here's the ugly truth. Right now, your heads probably don't flow enough, but you still might need smaller heads anyway. Confused? I thought so. Let me explain.
On average, a properly set-up engine can make about 2 horsepower for every 1 cfm of dry intake port airflow. However, that does not automatically translate to "bolt on better-flowing heads and make more power." I've tested many different size engines with many different sets of cylinder heads and found that when switching to larger cylinder heads, the torque lost on the low end compared to the power gained at the top end might hurt overall performance. That's because our vehicles are usually heavy and they don't spend that much time running at the rpm where they make peak horsepower. They do, however, spend lots of time at or below peak torque, which is where I've seen great power gains come from running smaller heads. And what do I mean by smaller heads? I'm referring mostly to the intake ports.
A run-of-the-mill stock iron small-block's intake ports might measure around 150cc. A good stock head, such as GM's iron Vortec castings, will measure around 160cc, but can make more power. A well-designed aftermarket head will typically measure from 170-200cc and will be capable of even better gains. They all get bigger from here. There are small-block race heads out there with 280-plus-cc intake ports and bigblocks way over 400cc, but those are way out of our league for real street power.
CFM And CC
It's important to understand the difference between cfm and cc when talking heads. As on carbs, cfm is a measurement of dry airflow in cubic feet per minute, usually recorded at a standard 28 inches (water) vacuum (aka depression). It's important to mention the vacuum, the medium used for testing, and at which point the airflow is measured, because more vacuum or a different medium may alter flow. The industry standard for comparison has been a 28-inch depression of water, not mercury, so that's where you should look when shopping for heads.
This topic actually gets a lot more complicated, involving things like 45ccbarometric pressure, temperature, the specific gravity of the medium used in the tests, and even the test equipment itself. But don't worry about all those things for now. Just try to compare flow figures at equal depressions, and hopefully from the same flow bench, otherwise those results might not help.
The cc measurement is simply a volume measurement expressed in cubic centimeters as measured from a graduated burette. When comparing cylinder heads, it's important to understand these two figures, because one or the other can make or break power.
Think of a long 1-inch-diameter tube. Now take something like a leaf blower and try to push as much air as possible through that long, skinny tube. At the other end of the tube there's not going to be too much air pressure, is there? That's because most of it will get blown around the tube instead of going through it. Now take that tube, shorten it, and make it 2 inches wider. At triple the original width, you'll get a lot more air through it, and you'll feel much more pressure at the other end. Next, make that tube 100 times wider and blow the same amount of air into it. Guess what? You'll find the pressure drops at its other end. Weird, you say? No, because as the tube's cross-section increases, it gets harder to maintain a constant velocity through it. The smaller the tube, the harder it is to force a large volume of air through it. The bigger the tube, the less pressure it'll blow.
This example can be translated directly to cylinder heads. A head that's too small will not let enough air into the cylinders to make power. On the other hand, a head that has too much port volume will cause the intake air/fuel velocity to drop and not "ram" into the cylinders fast enough for the most efficient filling, which will also cost power. A very famous Pro Stock engine builder once told me, "Give me a set of heads with the smallest intake port cross-section that'll flow all the air I need." What he means by that is he already knows how much airflow he'll need based on his entire engine package. Now all he needs are the right heads to make the number.
Port Size Makes the Difference
I can think of a great example that I saw in action once. It was on a bigblock, but regardless of engine size, the principle still applies. A guy takes his heads to a local inexperienced head porter and lets him go to town on them. They were some older castings, and he really took the grinder to them. He produced a set of heads that he was very proud of and handed them back to his customer. Now, $1,500 later, the guy clamped his freshly ported heads onto a flow bench in another shop to show off the awesome porting job. The numbers read something like this:
Not too bad on paper. But then the shop pulls out a flow sheet from similar, new, off-the-shelf CNC'd cylinder head made by one of the bigname manufacturers, and our buddy just about croaked:
After his heart had been thoroughly shattered, he looked deep into the issue and just about passed out. You see, not only did the out-of-the-box CNC heads beat his ported castings, but they also did so with a 45ccbarometric smaller intake port. This is a prime example of how the right port size and shape can affect airflow. Needless to say, the guy who owned the heads wasn't too thrilled about this, but regardless, he'd spent a lot of money on them, so he bolted them on, added a big nitrous kit, and drove off a happy customer anyway.
But all this is straying away from where we need to go. Unless your engine is big, like over 500ci, street heads won't need more than about 300 cfm max. And even that's too much for most engines. One of the best combinations I've ever tested used Edelbrock's E-Tec 170cc heads on the infamous Super Chevy 355-cid "Danger Mouse." That little pump-gas Mouse made almost 460 lb-ft of torque and more than 485 hp. And it idled decently, too, but I bet it would be one helluva kick to drive.
As a follow-up test, I bolted on Edelbrock's bigger E-Tec 200cc heads and got my butt handed back to me along with all the torque I'd just lost. Overall, torque fell off by 24 lb-ft. But interestingly, peak horsepower stayed right at 485, although it moved 200 rpm further up in the powerband. That'd be nice if I were planning on racing in a tirelimited class and didn't care so much about torque, but I wanted this to exemplify the best combinations for the street.
That day's testing really taught me a lesson about proper head management. I learned that the E-Tec heads are a great design overall, capable of making good power. But the smaller of the two, the 170cc, was better suited to overall running around, at least on that little 355-inch Mouse.
Chamber Shape & Flow
Air and fuel flow through an intake port at a pretty good clip. I don't know exactly how fast, probably slower than the speed of light but faster than grandma getting off her rocker. I'm sure there's a study on intake air velocity somewhere, but that's not important for you to consider when choosing a set of heads.
In a test I did a while ago, I dyno'd an engine with a pair of mildly CNC-ported heads right out of the box and then sent them for all-out aggressive CNC porting after the day's tests were done. When I got them back, the port volume had increased by 12cc and dry airflow dropped off at the bottom of the lift curve, but it picked back up at the top, surpassing the original set. More importantly, the shop that did the CNC work hit the chambers with their mill and reshaped them into what they believed to be a better design. The differences were subtle, but you could see it. Anyway, I bolted the heads back onto the same engine, went back to the same dyno, and retested it with no other changes at all. And I found more power. Not a huge amount, about 2.5 percent average overall, but it was there. After talking with the CNC shop, they pointed out that airflow measured on a dry flow bench does not indicate where the air is going as it enters the chamber.
My longtime friend and cylinder head expert, Joe Mondello, has developed a "wet flow" bench to find out what the air and fuel does as it enters the chamber. Many bigtime engine builders and cylinder head companies are using it to find things they'd never thought would occur. That's because air and fuel mixed together become a fluid, and fluids can move in strange ways.
Much of this effect comes from the chamber's shape and how it relates to the intake port. When the air/fuel column is rushing past the intake valve, especially at higher rpm, it's important for the cylinder head's combustion chamber to give it some direction. Note in the photograph how far back the chamber has been milled away from the intake valve. The arrows indicate the direction of air and fuel travel as it enters the chamber. You can see that the chamber's shape will at least give the mixture some direction of travel. This typically creates "swirl," which has been determined to be a good thing in combustion chambers.
Now, check out this stock chamber from an older iron head. See how it's got no shape to induce swirl? Nothing in this chamber is designed to direct the column of air and fuel anywhere. Its only purpose seems to be keeping the combustion pressures in the block.
For the longest time, it's been said that matching your intake manifold's ports to your cylinder head's ports is the only way to make real power. And that's not far from the truth, but it has to be taken in context. If you're building an engine to make max power at high rpm, then you'd better have your ports aligned perfectly. But on a street engine where you're trying to get the best power on a budget, it's kinda tough to perfectly match the ports. You can get close, but that's usually the best you can do.
Port-matching a set of cylinder heads is really not matching the ports completely. Instead, the ports are typically matched to an intake gasket size, and then the intake manifold can be opened up to match that same gasket, or left just a bit smaller. But by doing so, you could effectively put your head's performance out of the range you need.
What I like to do for stock or ported heads is to take the existing port shape in the cylinder head and straighten its walls and smooth in the entry just a little bit. Don't worry about opening it up to match a gasket exactly unless you're going all out. Then I do the same thing on the intake manifold and bolt them together. It seems to work pretty well and is a lot less hassle.
This is a perfect example of why you should clean up your intake ports. These stock ports wouldn't really be the best for performance, but cleaned up like the ports in the next photo, they'll make more power. Note how the cleaned ports are not really much taller or wider, except at the top. This is an easy way to make power, and fortunately, most aftermarket cylinder heads already come this way so you don't have to mess with them.
Shown at the top of the next page is what I'd call a fully ported intake port- almost too much for the application. Note how the top of the port has been opened up to the point where the gasket doesn't have much to seal against. What this guy really needed was a raised-port cylinder head instead of trying to make one himself. By raising the intake port's entrance in the head, you get a straighter shot at the intake valve, and the air/fuel column gets to make a less steep turn into the chamber. But raised port heads also require a properly matched raised port intake.
Average Flow Is The ReaL Deal
When shopping for heads, one of the best indicators for performance potential is average flow data. Take this example:
If you were reading an ad for both of these heads, Head A would probably proclaim "260 cfm @ .600-inch." And Head B could only meekly shout "245 cfm @ .600- inch!" But which one do you think might be the better head? Looking at intake flow numbers alone, I'd be pressed to choose Head B over Head A. That's because its average, not peak intake, flow is better. In fact, it flows much better down low in the scale, which is a very important thing to look at. But then I look over at exhaust and see that Head A flows better. Tough decision to make, isn't it?
Well, these flow figures aren't made up, and here are the test results from both heads run on the same engine using the exact same parts, except for the heads, of course. This engine was a stout 427ci Mouse, and the difference is slight but notable. Head B made more peak torque than Head A, but it started out making less. Average horsepower and torque went to Head A, so I'd call Head A the winner.
But Head B features an 180cc intake port, while Head A was more than 200cc, so which one do you think might work best on a smaller engine? It's a tough call, and because this engine had a large displacement, it could make use of the bigger intake ports in Head A. Had it been a good ol' 350-inch engine, Head B's 180cc intake port would respond very well on the street.
How Everything Else Affects Flow
Now we're going back to that thing I like to preach about: building the entire combination correctly, not just randomly bolting on parts when you can afford them. When choosing a camshaft for any engine, the first thing I look at is the airflow from the engine's cylinder heads. I pay particular attention at what valve lift peak airflow occurs, because it tells me how much lift needs to be put on the cam.
So if our heads are done flowing at 0.600-inch valve lift, it doesn't make sense to try and stuff a cam in that has 0.750-inch lift. Sure, you can go a little over the top, say 0.050-inch, but don't try to run too much more valve lift than your heads can flow. Also, pay close attention to the low-lift flow figures, because that's where your engine spends twice as much time as it does at max lift. A head that flows good low-lift figures will respond well, especially with a good dualplane intake and a smaller carb. Then look at the exhaust flow figures and compare them to the intake flow figures. You'll want a head that can flow a good ratio of exhaust to intake; somewhere in the 65-80 percent range is a good figure to shoot for. Anything more won't hurt, but anything less will.
Ways To Improve Performance Without Porting
While new heads are cool, there are other ways to improve power using the heads you already have. A while ago, Joe and I tested a rather mundane small-block iron headthat'd spent more than one day on the flow bench and Serdi valve machine being modified and tweaked in steps to measure which mods worked and which didn't. We did things like a three-angle valve job, back-cut the valves, port and polish the bowls, grind and polish the ports, and finally add new high-flow valves. What we found was that just by performing a threeangle valve job along with adding a 75-degree throat cut to smooth the transition from the end of the port to the valve seat, the head picked up as much as 6 cfm at .500 lift.
When you swap rocker arms to a higher ratio, you'll typically increase airflow through the head due to the additional valve lift and duration the higher ratio creates. But make sure you're not increasing lift to the point of pushing it past the head's optimum performance level, i.e., opening the valve past .600-lift when the head only works to .500-lift.
Comp Cams supplied the chart on this page that shows how changing from the stock small-block Chevy ratio of 1.52:1 (Comp says it's actually around 1.46:1) to a true 1.60:1 will increase your total valve duration by almost 20 degrees. This geometric increase is true for any cam in any engine.
Factory vs. Aftermarket heads
Earlier I alluded to a big-block example of hand-porting's downfalls compared to CNC porting. But there's not always a CNC-ported option for every head. Hand-porting, when done properly by an expert, is still a great way to improve a stock cylinder head's performance. I mention stock heads here because I don't recommend hand-porting a set of aftermarket CNC'd heads unless you can't afford a new set. The cost of porting a set of older heads versus the cost of a new set of CNC heads might work out to be about the same amount. Sometimes it'll cost you more to port the heads than buy new ones, so check all your options before plopping your money down.
You'll often overhear gearheads talking valve angles, as in, "I'm gonna buy a new set of 18-degree Chevy heads." It seems that the aftermarket is changing valve angles faster than some of the related parts can keep up, and many are finding out too late that the new splayed-valve heads they just nabbed are going to cost a lot more to run than they thought.
The valve angle changes I'm referring to here are the angle of the valve's face in relation to the piston's flat surface. Way back when, in order to fit the wide 90-degree V-8 inside a stock engine compartment, OEM engineers angled the top of the valve stem in toward the intake manifold, compromising flow. Race engine builders then started angle-milling cylinder heads to move the valve stem back out toward the exhaust side, which puts the valve in a better relation to the piston. This produced a better-breathing cylinder head, and as such more power came along with it. But angle-milling the heads meant that you also had to move everything else to match.
Today, cylinder head companies are simply moving the valve angles within the heads, but the intake manifold and rocker arm companies have a tough time keeping up with the varying angles. And so far I've only talked about rolling the valves toward the motor's exhaust side.
Canted valve heads' intake and exhaust valves are at different angles, and splayed valve heads' valve centerlines point out from each other as viewed from the top of the heads. This allows large valves room to open in small cylinders. Actually, canted/splayed valve technology is nothing new. Big-block Chevys have had it since their birth, which might explain why a stock oval-port big-block head can flow just about as much as many race small-block heads. The most important thing to keep in mind here is that whenever you move the valve angles around, it's going to cost more money to build the engine.
This illustration highlights how complicated the end of a normal V-8 valve can become. Study it carefully. All these angles, widths, and dimensions can be modified for increased performance. It takes days, even years to develop the perfect setup, and only the guys with the biggest budgets have the resources to find minute differences in power here.
The three-angle valve job you've heard so much about refers to the angle of the seat cuts on the valve and in the head. Typically, the valve actually has only two of the angles (highlighted in the circles), and the head has the third, a transition cut commonly called a "throat cut," used to blend the valve seat into the cast area of the port's bowl. It does not have a corresponding cut on the valve face, and the valve will never touch this area.