For those who don't know about 2-stroke diesels, they're quite efficient machines. The way they work is that they have intake ports in the sides of the cylinders that are uncovered by the piston as it goes down from combustion. When the ports are uncovered the exhaust valves open and a supercharger (2-stroke diesels require a supercharger) scavenges the cylinder while providing an intake charge.

Now while this saves two whole cycles, since the engine doesn't have a dedicated exhaust stroke, it requires a supercharger to scavenge the cylinder. Because this chews up the intake stroke's time, the supercharger's boost never actually translates into much higher cylinder pressure. And since the ports are fairly low (they can't be too high or else they'd interfere with combustion efficiency), they only have so long to do their thing anyways.

Now I present to you the 2-Stroke Diesel Miller Cycle. It has a 3-valve, SOHC head (2 exhaust 1 intake) [edit: intake valve would actually be a sleeve valve for improved overall efficiency]. The exhaust valves opperate as normal. The intake valve, however, opens after the ports are covered by the piston yet still early in the process. The intake valve is also connected to the supercharger, so when the valve opens the engine acheives the Miller Effect (increasing efficiency dramatically) but also higher cylinder pressure and lower required compression ratio (remember that diesels are compression ignition), meaning you get more out of your supercharger than was ever possible.

I know some people have experimented with Miller Cycle 4-stroke diesels, but I've not heard of this before.
-- acurafan07, Jul 31 2007

Good idea. But: how about sleeve instead of poppet intake valves? Single sleeves a la Burt-McCollum are a lot simpler, more durable, and generally better-baked than the more common Knight variety. In a CI 2-stroke their movement can be further simplified, as all they have to do is induce asymmetric intake timing. That is, they open the intake ports somewhere before bdc and close them somewhere up the compression stroke. It is good for the ports to sit low in the stroke, in the interest of scavenging airflow. With sleeve valves they may occupy the entire bottom third of the stroke.

(This is easily achieved with rotary-only or axial-only reciprocating action, or a combination, depending on the sleeve drive you choose. The advantage of a part-circumferential reciprocating rotary action that springs to mind is that a conventional head-block interface may be used.)

The exhaust valves would remain poppets, as per the usual "uniflow" arrangement; and as usual, you've got room for three or four of them. With the possibility of timing the intake ports allowing them to be a lot bigger than would otherwise be possible, this thing should outflow anything short of a gas turbine.

It strikes me that the difference between the Otto and Atkinson/Miller cycles is one of degree, and that there is a valve timing regime (cam grind, if you wish) where you wouldn't be able to decide which cycle applies. That ambiguous range might be worth exploring.
-- Ned_Ludd, Jul 31 2007


The sleeve valves do sound better. I'll make an edit.
-- acurafan07, Aug 01 2007



random, halfbakery