I might just think out aloud here on using microwave oven magnetron and water in car engine or similar.
1. microwave magnetron directly onto cylinder head will reduce losses, 2.45ghz wont be perfect match for given cylinder dimensions but should be ok for experimenting.
2. I'm thinking ultrasonic water fogger as the fuel supply ( water) with no modifications to engine air intake.
3. what timing regime will suit: this is difficult question.
a. water droplets sucked into cylinder via air intake, the microwaves can be used to heat the water mist, provide it with some thermal energy during compression stroke. jWould only need to heat the water a small amount not enough for steam production. During compression the water doplets will become closely packed, I'm guessing they will remain as droplets and not coalesce into bulk water at maximum pressure. Will water in small droplet form absorb microwave energy.....I think yes it should as the dipole structure should react to the e field of the microwaves, in the form of rotation. So I would think the microwave should turn on slightly before tdc to warm the water but not create steam/water vapour.
b. at or near tdc the spark plug should fire in order to start off the plasma phase of the air in the cylinder, perhaps slightly before tdc?
c. during the plasma phase the high temperature plasma in confined space should split a portion of the water into hydrogen and oxygen ( if the temperature is high enough). Only some portions of the plasma might reach sufficient temperature, from memory somewhere in the order of 3,000degrees is required. The will also be production of other products in the plasma, OH radicals, and oxygen should also be influenced. There is one patent by a medical company in the uk, which creates OH radicals by splitting water mist in a microwave air plasma. The OH kills bacteris and virus in the air, and lasts for a number of minutes after being produced.
Its the only patent I've thus far found that uses ultrasonic mist and microwaves ( I will find the link and post later). From patents I've read on microwave plasma ignition systems and air plasma for other purposes it seems the temperature of the plasma is not the same across all the plasma, near the cylinder walls the plasma temperature is lower, and swirling of the air is a method to keep temperature lower on the cylinder walls.
d. lets assume there is high enough temperature for a decent amount of hydrogen and oxygen to be produced, what happens to the H and O when the plasma is turned off?
The plasma is sustained by the microwaves and should cease to exist when the microwave is turned off. What period the microwaves should be produced for across and after tdc I do not know, but in the microwave ignition systems they use on periods in the microsecond range.
When the microwave is turned off the plasma temperature should decrease rapidly, the production of H and O should cease, I'm guessing at some point there will be enough temperature in the cylinder to ignite the H and O, I'm guessing it will self combust. If it does not self combust then another spark might be required at the end of the microwave period to ignite the H and O.
c. will steam be produced? I would think there will be steam present during the microwave plasma phase in the lower temperature areas of the plasma. It will also be produced when the H and O recombines. So I would guess steam expansion should also play a role in the process.
This is just my thoughts on a possible scenario to use microwaves. My guess is there will be an quite complex process going on, which will be related to the temperature profile of the plasma. Ideally maximising amount of H and O produced would give best results ( at present I think temperature of the plasma will be a major factor in that), but there will many different chemical processes occuring. There would likely be significant amounts of OH present. I think there is a possibility of creating an explosive type reaction if there is enough H and O produced and if it ignites at the correct time, also steam production should play some role due to its expansion. It might land up being technically complex to actually get this to work.
Will the water droplets cause quenching of the plasma in some areas of the combustion volume near tdc?
If the system does not work, it might be possible to add small amounts of petrol or other fuel to the water before its atomised so I lean mixture can be used. I'm quite surprised that I cant as yet find anyone who has tried this system purely with water mist.
Charging the water mist before entering the cylinder might also be advantageous, the excess electrons might add the the plasma temperature or stability.
I intend to give this a try on a one cylinder briggs/stratton engine, 4 stroke.