Alternate Wankel Pocket

Now, I'm not sure if this will work, so it may very well add to my already-impressive collection of bad ideas. But bear with me.

Rotary (or Wankel) engines, as part of their design, have a small pocket in each of the three rotor faces, which acts to give additional volume to the space created by the rotor in its motion. This is a benefit in that it adds to the amount of fuel/air mixture that can be exploded in one rotation, but it also decreases the maximmum possible compression of the gases prior to ignition.

Part of the downside, as I see it, is that the pocket is symmetrical across its long axis (see link for complete description of rotary engines). This creates a simpler design, but also means that an ignition creates a force that is equal forwards and backwards, so the engine must rely on its natural momentum to continue to push the rotor in the correct direction immediately after ignition.

As such, I propose that the pocket in the face be changed to a scoop design (one short deep face on the forward end, tapering to a shallow face on the "back" side), with the same volume as the current pocket in the rotor face. See basic art for cross-sections.

Now, this would alter the dynamics a bit.

First off, this would create a sharp volume increase, and therefore stronger vacuum, in the intake chamber of the rotor chamber at the beginning of the rotation. This would cause the engine to starve itself unless fuel line pressure were increased, but I don't see that being a problem with modern fuel pumps.

Secondly, I hope (*this is the reason for the idea*), it would create a side of the dimple for ignited gases to "push" against, adding to the force already created by the expansion of the gases to power the rotor around. Currently, spark plugs are placed equidistant from the edge of the face and the center, dividing the ignition pocket into thirds...the plugs would have to be shifted forward a bit (downward in howstuffworks's illustrations) to have the same positions relative to the center of the pocket's volume. Alternatively, a third spark plug could be added just past the lower one, although you might have too much spark and not enough fuel.

Thirdly, and this is the tough part, the exhaust chamber would have a vortex in it (caused by the altered pressures of the new dimple shape) that would probably cause exhaust bleeding - some exhaust gases would remain in the pocket for the next cycle, which would interfere with the combustion, and cause the engine to lose power after the first cycle.

So....a rolling teflon "exhaust scoop" mounted on a spring would be added to the system. Just beyond the upper lip of the exhaust vent, the scoop would be spring-loaded to quickly dip into the pocket on the rotor face and act as a method of quickly decreasing the volume of the gas, forcing more of it out and into the exhaust system.

Now, the addition of moving parts does sort of work against the whole idea of rotary engines, even if they are simple plastic rollers and springs. And I'm sure that, given my fairly weak knowledge of fluid dynamics, I've missed something hugely important here...so I'm posting this out of curiosity, partly.