porting and flow testing
So we've established that the porting and flow testing of the cylinder head is a critical part of making a six cylinder engine perform. For the most part port development follows standard procedures but there are a couple of things peculiar to the little Holden head that are important to keep in mind. We'll have a look at them now.
intake/exhaust flow balance
A lot of people still follow the old rule-of-thumb that says that the exhaust flow should be around 70 to 80% of intake flow. The standard Holden head roughly follows this ratio and in fact it works pretty well for low-output engines of say 0.7 - 0.8 hp/cu.in. But as the specific output increases we find that we need to use a lower proportion of exhaust flow to make power efficiently. For outputs of 2hp/cube for example an exhaust flow of not much more than 50% of intake flow will be entirely sufficient and will make more power than an engine with an overported exhaust. Also remember that a big port that flows very well - especially at low lifts - will usually be just as happy to flow backwards as forwards. The end result of this of course is that reversion and charge short-circuiting will become increasingly problematic as rpms and power levels increase. Both these conditions will probably be very familiar to anyone who has done much work with the little sixes. This tendency towards overexhausting also has implications regarding camshaft and exhaust system selection, and may partly explain why the engine has a reputation for not responding well to larger exhaust systems on street cars. We'll look more closely at this in the cam section, but broadly speaking a dual pattern cam (with a little more exhaust duration) works well on low powered engines, while for mid-range engines (say 0.9 to 1.25 hp/cube) a single pattern cam would be preferred. Higher outputs still may work best with a reverse pattern cam with up to 10 degrees less duration on the exhaust lobe than the intake. Provided the port flows and cam profiles are appropriate, the exhaust system can then be proportioned following "standard" performance formulas without overexhausting.
It's interesting to see just how differently the intake and exhaust ports capacity relates to engine power. On the intake side it's a pretty simple relationship - as the horsepower target increases so does the required intake port flow. Doubling the horsepower means doubling the flow capacity. It was once thought that the same relationship also applied to the exhaust port, but it appears to work completely differently on the hot side. Because the exhaust port is basically "self powered" (lousy description I know) by the energy of the very gas it is venting, its capacity doesn't need to be increased in line with power increases. In fact the optimum capacity appears to be based on the cylinder displacement more than anything else, and it seems this figure remains fairly static despite huge variations in RPMs and outputs. In terms of hard numbers, you could reasonably expect exhaust flows of only around 110-120cfm @ 28" to work very well with 202s ranging in outputs from say 150hp to 300hp+.
The one big exception to this rule is with mechanically supercharged engines; the flow capacity of the exhaust needs to be increased in direct proportion to the increase in mass flow. In other words you'd ideally have roughly double the exhaust flow capacity with an engine running at 14-15psi. Obviously the ideal numbers will almost certainly be difficult or impossible to achieve (and you'll have some unorthodox intake/exhaust balance figures) but any increases in the exhaust flow will help. Turbocharged engines are a different matter; the pressure of the gas between the valve and the turbine means the volume is much smaller and hence the need for overlarge exhaust flows is much reduced.
flow testing
I've already mentioned how important it is to be able to test and verify port work. I can't stress this enough - trying to get good results from a cylinder head without flow testing is somewhat akin to trying to paint a car while blindfolded. You might get acceptable results, but it's very very unlikely. Not only is it absolutely critical that we test and verify on the flow bench, we must also make sure we use a sufficiently high test pressure to ensure our flow numbers are meaningful. Converting from one test pressure to another works just fine with modern port designs but it just won't cut it with the Holden six. Why is this? Why can't we just test at say 8 or 10" and convert the numbers to the defacto standard 28" using conversion tables? The answer to these questions is in the design of the ports. To explain, let's first look at a more modern port, like the SOHC Falcon six head for example. The intake port on these heads is basically round in cross section and very gently curved. There is no short side radius to speak of and the port resembles a gently curved piece of pipe. Flow testing this port we see that the flow velocities are fairly uniform throughout the port cross section, even at very low and very high pressure drops. We can test at say 6" and then convert to 28", and the converted figure will match pretty closely what we get if we test at 28".
Now let's look at the Holden heads. The big difference between the Holden and the Falcon port we looked at is the very sharp angle in the port - the air has to make a turn of almost 90 degrees into the valve bowl. The short side radius is very pronounced as is the difference in length between the floor and roof. Now, if we test one of these ports at low pressure (and this applies to both 9 and 12 port heads) we will find that nearly all of the air will be flowing along the floor and out the short side. Flow along the long side will be somewhat turbulent and minor by comparison. If we test again at a much higher pressure drop however, we get a completely different picture. Much of the flow will be skipping across the back of the valve and out the long side. There will be turbulence and possibly even areas of reverse flow around the short side. In other words we have two entirely different flow patterns, and this is why it is unrealistic to expect flow numbers from testing at say 8" to have any relevance to what happens at say 28", or to what happens in a running engine for that matter. Bottom line: if you're going to work with old-style ports like these, then test using at least 28" of pressure drop. If you use much lower pressures you may be fooling yourself into thinking the ports are better than they really are.
testing ports with the manifold runner attached
One other thing that's worth mentioning is the importance of testing with a runner attached to the head port. This is pretty good practice with any head but it's even more important with our Holden heads. As was mentioned earlier, the flow velocities are nowhere near being uniform throughout the ports cross section and there are areas of definite flow bias. Because of this, the ports are fairly sensitive to the shape of the manifold runner as well as its approach angle. If you are limited for whatever reason to a particular manifold, then that's the one to use for testing the head. Otherwise test to find the best runner size, shape and approach angle and use these for both the manifold design and head testing.