Alright, I'm about to indulge on a pure nerdism. I don't think I'm gonna get any replies to this one and probably shame myself from posting here ever again.

So here's the question. Would it at all be beneficial to have supersonic intake flow velocities?

We are well aware that ram-air intakes are usally work pretty well at cramming more air into the cylinders, but I'm not sure if anyone has ever considered the speed which this air enters. Under normal conditions the air would just enter the cylinders at a rate proportional to the pistons speeds. What if we used a converging-diverging nozzle pipe to increase the intake velocities? That would be a pretty smiple way of getting the sonic flow, assuming the conditions make it possible.

Now, if the intake velocity was much faster than the piston itself, then couldn't more air could be drawn into the cylinder therefore producing more power?

The problem that I see right now is that when the intake valves close, then the air/fuel behind it just sits there with essentially zero velocity. This would negate any benefits that supersonic speeds would provide. Maybe if the engine was designed correctly with a very small amount of space behind the valves (i.e. the ej25 with short intake runners) then perhaps such a design could work.

 

I think the answer to the zero-velocity question resides in head design. If you position the intake port and angle it enters properly, you get plenty of swirl in the combustion chamber to keep the velocity up and prolong atomization. But I have no idea how good or bad EJ20s and EJ25s are about this.