One could ask oneself: why bother testing heads without a manifold as you can’t really use them without. Well for one it allows you to isolate the source of the restriction. Is the manifold just bad or is there something else going on. Vizard talks about valve limited and port limited. This implies that there is a crossover point somewhere, As you can’t separate the port from the valve seat area without a hacksaw ( not that that makes much sense anyway) I thought of adding a bit of port in the form of a manifold. The manifold should introduce a flow loss that is somewhat independent of the valve and seat area. So by vaying the amount of port (and keep the seat area unchanged and vice versa) you should be able to figure out where the crossover point between seat dominated flow and port dominated flow is (at least in theory). (I have since found the figure in the Vizard 3rd edition yellow book. He measured air velocity across the port and seat and averaged that. Figure 9.23 page 204 plots seat velocity vs port velocity crossing over at whatayaknow 0.275 inch lift or 7 mm).
This method basically does the same thing in a different manner without having to use pitot tubes and averaging of a load of velocity data.
In this case the setup is unchanged: the valve with the narrow margin and tiny back cut is still used but now the manifold ( a minispares medium bore item) including a large radius intake is added. As with all tests, it is corrected for temperature, barometric pressure and relative humidity. As it turns out it does absolutely nothing on low lift flow,
until 6mm lift when it starts to diverge and causes a constant loss of flow starting at about 7 mm lift.
You could interpret this as follows: The flow to 6mm or 236 thousands of an inch is seat/valve limited, after 7mm/275 thou it is port limited.
If we look at the following:
blue and green are NO-MANIFOLD and MANIFOLD respectively and the red line is NO-BACKCUT/NO-MANIFOLD. The gains of the valve are still there regardless of the manifold loss up to a lift of about 8.5mm .
Now .. all together now. The graph is getting a bit crowded but bear with me. It are two pairs NOMANIFOLD/MANIFOLD resp Green&Blue and Red& Black. Note that the only difference is the valve.
If you look to the far right the NOMANIFOLD /MANIFOLD groups converge ( as they should… hurray!) the difference is the loss caused by the manifold. About 6-7 CFM.
Note that the Red&Black lines start to diverge at about 5mm and stop diverging at about 7mm ( read a constant loss.. like the one caused by the manifold). Green&Blue start to diverge at 6mm and stop at … 7mm
So if this is true ( and it has to be seen if it holds up to repeated testing) that the crossover point for the port is ~ 7mm lift.. My guess is that the crossover point stays 7mm but the divergence point will shift towards 7mm as the seat area improves.. we will see.
If this is indeed the case this might indicate a tad more specific where you have to look for gains.. even if things are never this straighforward in reality 🙂