Ammonia engine - temporary short description

My previous posts assumed that two-stroke construction based on Lenoir cycle will be used. Charge would be admitted at higher pressure, and admission would take place only at the beginning of downward stroke, then fuel-oxidizer mixture would be ignited by multiple spark gaps, and power extraction would begin. Various combinations of ammonia, hydrogen, air, and oxygen could be used. After researching multiple options, I came to the conclusion, that burning NH3 with air is the best bet for a land vehicle that is expected to have reasonable range. Separating O2 from air (or H2 from NH3) would require bulky counterflow heat exchangers, that would accumulate H2O and CO2 on their surfaces. I was thinking about marketing it as negative-emissions vehicle, but it probably would be a bad idea, nevertheless it could be possible to have heat exchangers connected in parallel, so when one of them clogs, another one acts as a bypass, and can continue to change temperature of passing gases.

Four-stroke engines have significant advantage over two-stroke in precisely actuated intake valves, that can move significantly in long period of time. My previous design had to (de)accelerate valves at very high rates, while travel distance was minimal. By increasing number of revolutions required to complete the cycle in four-stroke, we decrease available power, but make intake process simpler and more precise, with less unburned fuel in the exhaust (direct injection may alleviate some of those problems in two-stroke). Four-stroke reduces percentage of cycle during which hot gases have contact with metal surfaces.

New version of the engine would operate on HCCI (homogeneous charge compression ignition) principle. Mixtures of ammonia and air are hard to ignite, so compression ratio in order of 40:1 would be required to achieve complete combustion in almost all places, at the same time. Regulation of intake valves' timing would be crucial in achieving perfect combustion, introducing into cylinder amount of gas that is not going to ignite prematurely, but also not too late, reducing available power, or missing completely. Geometric compression ratio stays always the same, while amount of introduced charge depends on intake valve timing, just like in some currently produced Atkinson cycle imitating cars, allowing changes of effective compression ratio from cycle to cycle. Expansion ratio always stays constant.

As for geometry of the head, I would place intake valves on opposite sides, as well as exhaust ones, that are rotated by 90°, in reference to intake valves. Connecting manifolds will be harder than in regular engine, but more symmetric charge flow during intake should be less chaotic (turbulent), reducing amount of heat transferred to walls. Ignition of all uniform mixture in almost the same time should equalize rate of heat flow among head, piston crown, and cylinder wall.

This gives opportunity to use steel as main material for parts that have contact with hot gases. Spark plugs produce hotspots of high rate of heat flow, as well as jets of fuel in Diesel engines. Aluminum has high thermal conductivity, and thanks to that, can better withstand existence of hotspots. Connecting rods and bearings will have to experience extreme stresses, as peak pressure inside cylinder will be over 100 bar.

Fortunately, those conditions will be present only for short amount of time. System of crankshaft, connecting rod, and a piston acts as specific type of transmission. During steady operation, one revolution of crankshaft results in piston traveling twice the stroke distance, but big end of connecting rod travels 3.14159... times stroke length. When piston is in the middle position, its velocity is almost the same as tangential velocity of crankpin. But when piston is near its top position, it travels very slowly, while angular velocity of crankshaft is almost constant, and much higher. Taking this "variable gearing" into account, it can be seen that for short period of time, described system acts as very decent lever.

High compression ratio has significant advantage in the form high thermal efficiency. Due to uniformity of HCCI combustion there shouldn't be thermal NOx emissions, but breakdown of NH3 will produce significant amount of unwanted molecules. Selective catalytic reduction (SCR), using ammonia, would seem to be best option when it comes to dealing with those molecules.

Ammonia is so nonflammable, that achieving right pressure rise, that doesn't resemble engine destroying knock, shouldn't be that hard, at least with precise control of intake valves, gaseous ammonia "injectors", and general parameters of charge, such as pressure, and temperature (which would be higher than ambient, so that compression ratio can be lowered, during starting electric heater cold be used to make required compression ratio very low).

To achieve that level of control, I would like to use novel camless valvetrain. To each poppet valve 2 springs would be attached. In neutral position both of them would act on the valve with the same force, making valve stationary. Those springs would be attached to opposite sides of disc that is short section of stem that has increased diameter.

On top off the stem (or near the top) another disc would be present, that is to be enclosed in double acting hydraulic cylinder. On each side of this cylinder connection to low-pressure, or high-pressure, fluid reservoir could be made via variation of voice coil motor that can unblock annular opening under the coil. If this could be done precisely and work reliability, then it might be the best option. Circuital opening requires that plunger lift off 1/4 diameter to fully utilize it's potential. Annular opening at the same lift will provide much greater surfaces through which fluid may flow. Surfaces that bash together should be made of sturdy material. Groves should be present on the sides of "voice coil", and sides of its housing that help with breaking moving mass, by forcing fluid through narrowing, when this is desired. Frequent flow of fluid should be used to remove heat generated by current flowing through conductor. Voice coil valves should be oriented ("polarized" like diodes), in such a way, that when no current is flowing through them, no fluid is passing through voice coil valve.

There are 4 voice coil valves per double acting cylinder of main valve (each side of cylinder can be connected low or high pressure reservoir). Quite complex system of channels will be required to connect everything, so additive manufacturing may come in handy.

Going back to main intake and exhaust valves, system proposed by me works on relatively simple principle. If no friction would be present, valve could act as simple harmonic oscillator. When it is pulled out of neutral position, springs store potential energy. When valve is released, it starts to gain kinetic energy, and is speeding up. When reaching middle position velocity is at the maximum, and valve starts to ram/pull springs until it stops, and farthest possible position is reached. Double acting cylinder with voice coil valves is used to hold main valve in place, or restore energy lost due to friction.

Symmetrical arrangement of main valves would allow placing "dummy valves" around the center of symmetry, that accelerate in opposite direction, allowing for balancing of vavlvetrain. Even if main valves are not symmetrically arranged, there are possible configurations of dummy valves, that can take care of vibrations.

Electronic control of valves should allow turning of this fourstroke engine in reverse. Normally it is impossible because power stroke would have to come after exhaust. In camless engine order of strokes is not controlled by gears and belts, so fast changes in modes of operation are possible. It should be beneficial when used in clutchless hybrid, as engine can quickly enter a mode that only "puffs" small amounts of gas, without any fuel burning.

I would like to use a few tricks that should be useful in increasing engine efficiency (and making it bulkier).

First of them would be inclusion of Rankine cycle coengine (word I just invented, to describe small engine, that is part of the bigger engine), running on waste heat entering main engine's head. Liquid ammonia would enter head at 4 points that lie between valves, on the outer edges of head (I am assuming that each cylinder has separate head), then flow into central chamber, that is mostly filled with a "heatsink", that transfers heat from the bottom-center of the head to pressurized boiling NH3. Generated gas would be then exiting via well insulated tube on the top of the assembly, later entering separate expander cylinder, which piston is running at half the main's engine speed. Regular speed would make this cylinder too tiny. Amount of ammonia that would pass through this coengine should be equal to the amount that will be burned in main engine, thanks to that we can get additional kick out of pressurized form of NH3 storage. Any NH3 leakage should be directed into main engine intake.

Hot exhaust gases from main engine still contain useful energy. I propose that two Brayton cycle coengines could be used to recover parts of energy from exhaust. Counterflow plate-fin heat exchangers would be used as interfaces between exhaust and compressed air. Use of air as working fluid has huge benefit of not having to condense or refill water, like in typical stream engine. First Brayton coengine would work at higher temperatures and pressures, an at the end of a cycle would just dump warm air overboard. Second Brayton would be used mostly to warm air that is to be supplied main engine intake, as amount of recovered energy would be minimal.

For those Braytons I would like to use pairs compressor and expander cylinders, with previously mentioned camless valves, only weaker version. Assuming two main cylinders and a pair of compressor-expander Brayton cylinders, we can imagine specific inline-6 engine (I am not counting ammonia expander as proper cylinder), where 2 main piston are in straight-twin configuration, and the rest of lightweight pistons are rotated by 90°, making sure that not all of the compressions happen at the same time. Separating reciprocal masses by 90° means that when one of the masses has maximum kinetic energy, another one has minimum, sum of the two in other situations remains relatively constant. Reciprocal masses act as weak flywheel.

I assume that pairs of expander-compressor pistons move in unison. This means that there is no overlap between compressor expelling air into heat exchanger, and expander absorbing new portion of gas. Fortunately volumes of heat exchangers should be big enough, as to not cause any major issues.

Typical piston connected to crankshaft produces vibrations due to 2 sinusoidal forces acting together. First of them is larger and has the same frequency as number of crankshaft's revolutions per second. The other force has frequency that is two times higher, and has much lower amplitude. Inline-6 engine that I described could be perfectly balanced by placing pairs of balancing shafts on each side of the engine.

On the crankshaft large gear would be located, that is meshed with another large gear located on balance shaft. Another large gear is located on the mentioned balance shaft, but it meshed with another large gear on other balance shaft, located on the same level, but on opposite side of engine. Now we have two counter-rotating balance shafts, which when equipped with right eccentric weights can destructively interfere witch sinusoidal forces created by pistons.

Upper level of balance shafts would have gears with half the diameter of gears on lower level and on the crankshaft. Each of gears on higher level would be meshed with lower level gears. Function of eccentric weights on counter-rotating upper level will be transforming vibrations of the piston into pure sine wave, with which lower level of balance shafts can deal.

NH3 expander cylinder that uses separate minicrankshaft should also be balanced, by separate set of 4 balance shafts for example.

Inline-6 configuration that I described above might be well suited to power small tandem two-seater, with small trunks on both sides, and RR layout (or AWD where rear axle is configured as parallel hybrid, while front one is a series hybrid).

Originally I started developing this engine for a cabin motorcycle. In this configuration only single main vertical cylinder would be used. Brayton coengines would be horizontal, their crankshafts located higher than main crankshaft. Those coengine crankshafts would be separated by main cylinder, but still connected by 3 balance shafts, that envelop main cylinder. Main sinusoidal vibration of the first order could be taken care of in similar fashion as in V-twin engine, by placing eccentric weights on coengines crankshafts, that produce circular vibration, that matches two linear sinusoidal (or cosinusoidal) vibrations that are at right angles to each other.

This text was spontaneously written on mobile device. I'm currently imprisoned in capitalist psikhushka, so I can't share spreadsheets with calculations, or make decent drawings. I will try to update this post in the future, if it will be possible.