I’m an engine designer currently working on a new two-stroke that promises to be more efficient and cleaner than most four strokes. The answers given thus far mix theoretical and practical perspectives.
In practice, two strokes are usually less efficient and dirtier than four strokes for the three reasons many have mentioned: corruption of the intake charge with exhaust, passage of unburnt fuel to the exhaust, and mixing of lubrication oil with fuel. As several have noted, however, some two strokes are among the most efficient engines in the world, and that’s a clue that two strokes don’t *have* to be inefficient.
The easiest issue to fix in typical two-strokes is to eliminate mixing of oil with the fuel. This can and has been accomplished by injecting oil between the piston rings. This approach requires careful management to prevent oil from passing out the exhaust port in liquid form, but it can and has been done at the cost of increased complexity.
A somewhat more difficult problem with traditional two strokes is the fact that scavenge and intake only operate ideally when exhaust back pressure is ideal. This problem is what gives naked two strokes a narrow powerband that makes them hard to drive. There have been a number of attempts to fix this, some of which others have mentioned. The most common solution in early two stokes involved addition of rotary and reed valves on the ports. The best solution, often used in conjunction with auxilliary valves is use of precise fuel injection timing late in the intake cycle, or better yet, after the ports have closed. One final solution is to end-run the problem by operating over a very narrow speed and tune everything to that speed (this one reason the efficient two strokes some have mentioned operate at such low speeds). This approach can be very effective if mated with a Continuously Variable Transmission (CVT) that can relieve the need for an engine to operate over such a wide band. In ships and aircraft, variable pitch propellers can accomplish a similar end.
So, yes, a two stroke *can* be as efficient as a four stroke, but only by increasing its complexity and/or abandoning the well established paradigms driving a need to have a wide power band. The primary reason designers were interested in two strokes was their simplicity, and the need to add complexity to overcome their challenges eliminated their attractiveness. As for engineering paradigms, well, they’re simply hard to overcome.
In considering the future of internal combustion it’s important to note that we may be entering an era in which engine requirements change dramatically. Hybrids will soon dominate sales as they can meet emission standards without imposing long charge times on drivers or a need to increase clean electrical power generation by a factor of thousands. One particularly interesting form of hybrid is one in which the internal combustion engine serves only as an electrical generator (aka “range extender”). In this application, the engine can operate at fixed speed and load. Thus, there’s no need for all the complexity associated with clean and efficient operation over a wide power band; we might use a simple two stroke tuned for that narrow load range and speed instead.
Wait till my new two-stroke engine is tested. People will be shocked (or amused if it fails testing).