Two stroke engines are able to have twice the combustion cycles per RPM (one per every revolution of the crank vs one per every two revolution). This is accomplished by using compressed air from either a positive displacement blower (conventional two-stroke diesels) or from the crankcase (conventional
two-stroke gasoline) to both scavenge the exhaust and force in new air.
While in a good idea in theory, it means that in the case of diesels it is often hard to compensate for the pressure lost from scavenging the exhaust (turbochargers can be added, but they do complicate things and narrow the powerband significantly).
Two-stroke gasoline engines have made numerous improvements in that past few years that have now made their emissions on par with competitive four strokes. One of those improvements is using direct injection and specially designed lubrication so that the crankcase can be used solely as an air pump, with no contamination of fuel and very little contamination of lubricating oils (they have a dedicated pressure lubrication system).
Because of this, my idea is for a two stroke diesel that uses the crankcase as a compressor. My design has a positive displacement (ie: roots) supercharger that both forces air into the intake port to scavenge and replace the air in the cylinder and also forces a limited amount into the crankcase. Because traditional reed valves in the crankcase probably wouldnt work well with the added pressure of the supercharger, a sliding valve can be used (think of a small scale garage door that is camshaft controlled). The pressurized air from the crankcase is forced up a tube that travels to the cylinder head and is channeled into two parallel swirl chambers, as in the old indirect injection diesels. Both channels to the swirl chambers have controlled flow by the same garage-door valves mentioned. In addition, there are two more of these valves that sit flat against the cylinder head that allow for the opening and closing of the swirl chambers relative to the cylinder.
The swirl chambers trade-off combustion events that they are active, meaning while one is actively combusting air/fuel into the engine, the other is having air compressed into it by the crankcase. Due to the fact that the downward stroke in which the chamber needs to be "active" is also the time that the piston is compressing the air into the chamber, two separate chambers must be used. Both chambers have high-pressure injectors attached to a common rail, designed so that there is a pilot injection that begins the swirl process and forces a lean-burning charge into the cylinder, followed by a larger injection once combustion has started.
Basically, these swirl chambers, if designed correctly, would increase the engine's air capacity by compressing and blocking off a small portion of air that will not be wasted to scavenge the exhaust. Also, because the air will be blocked in the chamber for about 180 degrees of crank rotation (the time it takes for the piston to go back up), glow plugs would have more time to heat the air so that cold starting wouldn't be a large problem. Turbocharging could be used in addition to my idea and the necessary blower to provide extra "boost" for the engine.