[malc9141]

Chapter 3 - cylinder head design, general.

If anyone wants to throw mud at this, go ahead. If I'm an amateur, there's bound to be stuff I get wrong.

First, if we want lots of power, we will burn O2 and have heat to get rid of.

Back in the "old days", English formula 1 engineers built engines with "large pots" and fewer cylinders because it "gave a simple rugged design." The opposition built multi-cylinder engines. Guess which engines burned valves out.

If we think of an individual cylinder, its volume (roughly) predicts the amount of air in it to burn. So for given revs, volume is proportional to heat generated (power). If we halve the volume, we get half the heat (power). But the combined area of the cylinder walls, piston top, and head-face are not halved. Volume is L^3, area L^2. So the heat "seen" by the walls is less, per unit area. Now heat isn't exactly temperature but it's proprtional to it. So I argue that a smaller cylinder is safer than a larger one. But you have half the power, so you need twice as many cylinders.
Remember, for an aero engine, reliabilty is paramount and overheating or burning exhaust valve seats must be designed out at the beginning.
Common experience tells us what a reasonable cylinder size is, but aircraft engineers are historically accepting of big pots, and petrol engine failures are all too common. (It's interesting how such failures always seem to be explained away as "fuel contamination" or "poor maintenance").

So about 450 cc for a pot seems right. For reasons I'll discuss later, for a turbo engine, we want an under-square configuration. In any case, we can't go for high revs.

That means a bore of, let's say, 80mm.
The most robust design is the Heron head, but since a two-stroke has almost total overlap of valve opening, the incoming air goes straight out of the exhaust. Worse, the stream of air actually impedes the residual exhaust gas from exiting.
What we want is: the standard explosive exit of exhaust gas, with if possible, the well known temporary vacuum created behind its momentum. The inlet valve opens at this point (of course, openings are not instantaneous) and the turbo-pressured air comes in. I argue that we want as smooth an entry as possible with minimum turbulence. (Turbulence is for the birds, and carburettors). We want the air to be as slow and smooth as we can get it (given we have about 7 milliseconds to get it in!). We do not want jets of high speed air. Why? Jets cause a local pressure drop and absorb the surrounding gas (exhaust fumes). It looks good, seeing high speed streams of air but it ends as a high volume, low velocity mixture of diluted oxygen. We want the air to come in like water, and expel the residual burnt gas as if it's oil. And then the exhaust valve closes.
To do this, we have to design a particular shape of intake port. This we did, and built a scale model with a glass partition, which showed the flow patterns when we pumped through high velocity air. We used children's fireworks sparklers to show the flow, and b/w photography to confirm the validity of the design.
I can't show the exact layout because it's the subject of a patent application, but it is simple and robust.
The nice thing is, the water cooling can be concentrated on the exhaust ports, and air cooling we believe will take care of the combustion chamber region. But above all, there is a low total surface area, meaning that heat stress in the head is minimised. (This is not at odds with my remark in para 4. That's heat per unit area, this is total heat pick-up).)
So we have a head with a simple surface, the combustion chamber under the cool inlet valves, and an air-flow pattern that scavenges well (perfectly?). We have done away with the usual inlet ports in the cylinder liner (jets, and bad ones at that), and have a smaller piston (no lomger doubling as a sleeve valve), lowering the whole engine, making it stiffer and lighter.

Next: Building the prototype head.

If anyone wants to get involved with is project, let me know.

Malc9141