Ideally stratified charge is a situation where the fuel/air mixtures in the combustion chamber is not uniform. There are areas where the mixture is stoichiometric and many other areas where the mixture is very lean. The combustion chamber of the Wankel rotary engine is a thin rectangular crescent shape with a bulge in the center. There is no way around this in the Wankel rotary engine. This is not the ideal shape for a combustion chamber. The ideal shape is a sphere. Making the rotor wider as in the 16B and Norton based rotary engines help as a square would be better than a rectangle.
Curtis Wright and John Deere spent millions of dollars on this problem and this is what they came up with. This is from a tech paper called The Stratified Charge Rotary Engine by James W. Walker Robert E. Mount at John Deere.
Here is the HP, torque and BSFC claimed for the SCORE concept engine.
Turbo charging seems essential to achieve these results. Here is what CW and JD say about the turbo charged Wankel engine.
Here is a table of the projected engine series, their size, weight, HP, Torque and number of rotors. Notice the rather poor power to weight ratio and low RPM of these rotary engines burning diesel fuel. A gasoline burning rotary engine would have a power to weight ratio closer to one. This power to weight ratio problem has plagued the diesel aircraft engine from the early 1930's when it was first applied.
We have discovered that the rotary engine is inherently stratified charged at high RPM. This was determined by trail and error leaning the engine under light load and high RPM. Unlike a piston engine it would run at air fuel ratios of 20:1 and 200 degrees F lean of peak. The centrifugal force on the fuel molecules, being heavier than air, tend to congregate out near the spark plugs in a rotating combustion chamber. Therefore the rotary engine runs smoothly at lean mixtures the gas piston engine can only dream about. If you have ever owned a gasoline engine powered boat you know that gasoline vapors congregate in the bottom of the boat called the bilge. Therefor they are heavier than air. If you don't ventilate your bilge properly your boat can blow up.
It is well known the best BSFC of an internal combustion engine occurs around a air fuel mixture ratio of about 18:1. Here is a chart from Heinze Heisler's book Advanced Engine Technology showing that relation ship.
Never the less the fire tends to go out around the the flat recesses around the center bulge. Worse, precious combustion heat is lost to the metal in these areas reducing the overall fuel efficiency. Therefore much research and development has been expended on achieving stratified charge in the Wankel rotary engine. Historically the research has focused on injecting the fuel into the bulge and hoping it would stay there as the combustion chamber rotated around inside of the rotor housing. Stratified charge may also be achieved by injecting fresh air into those parts of the chamber that are unlikely to sustain combustion. This is an idea I had way back in 1975 about injecting fresh air to achieve stratified charge. In a way this has a slight supercharge effect.
More recently I have had some variations on this original idea. This is an attempt to use digital electronics to control the air injection not only as a function of rotor position but also several other parameters such as intake manifold pressure.
After thinking about it some more my air injection idea above may not work well as combustion pressure may leak above the apex seal across the air injection hole. If I move the hole up so the trailing apex seal just passes the hole before the plugs fire there will be no combustion pressure leakage across the hole. Another possibility is to go back to the original 1975 idea and use side ports to inject the air. Never the less we will continue along these lines for the purpose of discussion.
In this design I have done away with the solenoid valve and replaced it with a synchronized piston air compressor. The air injected must be over about 150 psi as the pressure in the intake chamber when the rotor is in this position is approximately 150 psi. Therefore the compression ratio of the piston compressor must be 10:1 or above. Ideally much above 10:1. The piston is phased such that the injected fresh air reaches max velocity and pressure just before the trailing apex seal passes over the air injection hole. The piston compressor is geared to the e-shaft of course. Here is a 3D.
I placed the compressor to the inner surface of the combustion chamber to increase the compression ratio of the compressor and reduce the time for the air puffer to work. There are three air squirting passages but not limited to three. The 3D below shows the piston at TDC. In actual use the piston would be past TDC at this TDC position of the rotor shown in the 3D.
The dual intake valves are in the head of the compressor piston. They are ball valves that work on inertia alone. Much like the Gnome Rhone Monosupape of World War One. (Not strictly true but the word "Monosupape" is a fun word to annunciate :)). Fresh air enters the crankcase of the compressor and then up through the piston head. The last 3D shows details of the piston with the ball intake valves.
The advantages of this system over a conventional high high pressure fuel injection are cost and simplicity. A problem with direct gasoline injection is the fuel does not get a chance to mix with the air thoroughly before some of it burns. Since the fuel is introduced way upstream in this system is has a chance to vaporize. The difference is the air and fuel can mix back in the intake runners and in the compression chamber. No expensive 20,000 psi fuel injectors required.
BTW Mazda ought to love all these tiny parts that work like Swiss watches. Very much in the tradition of the Mazda rotary engine. Still a very simple and cheap stratified charge system compared to high pressure direct injection of fuel. It also has a very benign failure mode.
Also BTW the reciprocating air compressor could be replaced by a very small Wankel rotary type air compressor if one objects to the reciprocation.
The question remains. Will the compressor suck more HP than it will improve the BSFC?