There are a lot of improvements possible for internal combustion engines (aka ICE's). It helps to list the areas that are causing losses, to start:
-- The geometry of the physical layout of the piston, connecting rod and the crankshaft is less than ideal. The connecting rod needs to be ~60 degrees past top dead center to get the best leverage on the crankpin; but the pressure from the fuel ignition occurs much earlier than this; when the connecting rod is essentially trying to bend the crankshaft sideways. The motion of the piston is necessarily sinusoidal.
- The power stroke is only 25% of the full cycle, and there is a lot of mass that has to be accelerated, stopped and accelerated again.
- The valvetrain has to physically resist being moved, and it has to work against the air flows.
- The piston tends to scrape the sides of the cylinder, because it would "rather" twist that stay straight. The rings must exert friction on the cylinder.
- The oil must be pumped through little tiny passageways.
- Electricity must be generated.
- An ICE is a self-powered air pump, in essence. Air flow and the pressures generated, and the cyclical nature of them cause resonances, and backpressures, and the gasses become spring-like.
- Small volumes, like the space above the top ring and the top edge of the piston, trap unburned fuel because the flame cannot reach it.
- Everything flexes and springs -- the crankshaft and the camshaft flex torsionally and longitudinally, the piston vibrates and distorts, as do the cylinders. Valves bounce and stretch and distort into potato chip shapes.
The list goes on... The net result is a typical internal combustion engine that uses ~20% of the energy in the fuel for output motion at best, and requires a transmission to keep the torque of the engine relatively close to the speed of the vehicle.
So, knowing all this, how can we make incremental or wholesale improvements?
+ Offsetting the crankshaft center away from the power downstroke gives the connecting rod some better mechanical leverage -- but is the compression stroke adversely affected?
+ Variable valve timing allows the torque to be available over a broader range of RPM's.
+ Valves can be electrically/hydraulically moved in both directions (opened and closed) to avoid fighting the springs. This also makes it easier to use subtle or more abrupt adjustments to the valve timing.
+ Use cams rather than the crankshaft, to gain a lot more mechanical leverage, and to allow the piston motion to be controlled by the designer; like the Revetec:
This particular design also reduces piston scrape (but it introduces some tendency to spin the piston within the cylinder). It also avoid big changes in crankcase pressures (in configurations with even numbers of pistons). This design effectively doubles the efficiency.
+ Use the Atkinson valve timing, like the Prius does, which has a lot of overlap of the exhaust valve with the beginning of the intake downstroke (I think?) so that there is built in exhaust gas recirculation (aka EGR). This also effectively doubles the efficiency.
Hmmm, how well would a 2-cylinder Revetec with Atkinson cycle and electrically activated valves work?
+ Use a rotary design that reduces the reciprocal motion.
+ Use a 2-stroke design to cut the parasitic losses in half.
++ Use a continuous burn design to further reduce the cyclical nature of the engine; or at least reduce the time between power cycles.
+ Figure out how to reduce waste heat from being produced, and then try to use the remaining excess heat to produce output.
What are other ideas to improve ICE's?