This idea builds on my earlier idea from a few years ago [link], but improves it in a few subtle ways... enough, I think, to warrant this being a separate idea.

Similar features:

* All of the valves of the engine are fully computer controlled, so we can instantly switch from a four stroke cycle to a two stroke cycle, and back, as needed.

* The engine block has no water jacket.

* The areas where hot combustion gasses are produced and exhausted are lined with high tech insulating ceramic, similar to Toyota's failed adiabatic engine. This idea will take advantage of the high rate of heat transfer that caused their engine to be no more efficient than a normal one, and put that heat to good use.

* There is a steam condenser is probably cooled by a water/antifreeze mix, which in turn is cooled by a radiator. This way, in very cold weather, we can control the rate of coolant circulation so the condenser is very-cold-but-not-freezing.

There will also be a (small) vacuum pump, to remove air from the condenser. The air that comes out of the pump will inevitably be very humid, so it would probably be a good idea to route it to the exhaust gas water recovery device.

Naturally, there's a pump to move water from the condenser, to the boiler.

* There is high pressure exhaust powered steam boiler/hot water heater. This component will be producing hot liquid, and hot steam, both of which will be used in the engine.

* Downstream of the boiler, some or all of the exhaust passes through an exhaust gas water recovery system. We only need to recover a small amount of water from the exhaust, and we don't need it to be very pure. The water needs to not have minerals or particulates, and not chemically break down when heated in the boiler.

* A computer selects from three different modes of operation, depending on the inputs of various temperature and pressure sensors.

The default mode of operation is a perfectly normal four stroke internal combustion cycle. Not only does this produce a certain quantity of mechanical power (our main product), but it also heats up the engine block, and produces hot exhaust gas. For the latter two reasons, I'll call this the heating cycle.

Note: There are many different variations of a four stroke cycle, and for the sake of discussion, I do not care which one is used. It could be an Otto cycle, a Diesel Cycle, an Atkinson Cycle, a Miller Cycle, etc.. It doesn't matter very much. I'd prefer that it take advantage of the computer controlled valves to eliminate the need for a throttle, but that's a minor detail.

It *is* important that the exhaust not damage the water boiler, but that can be avoided by using appropriate pollution controls.

When the in-cylinder temperature exceeds some threshold, the computer switches from the heating cycle, to the primary cooling cycle.

We start with the piston at top dead center, with the valves closed, and spray in a small amount of water. This water is pumped from the exhaust heated boiler, using a floating intake which ensures that it is liquid water that comes out of the injector, not steam.

The quantity of water injected is carefully chosen by the computer, to be as much as possible, but within the following additional limitations: (A) The max steam pressure won't exceed the engine's tolerances. (B) The steam pressure at the end of the first (expansion) stroke doesn't exceed the pressure in the steam condenser. (C) The average torque over the two stroke cycle must not be significantly more than the average torque of the most recent four stroke cycle.

The water strikes the piston face and cylinder walls, and is heated by them, changing from liquid to steam. Since this phase chance happens in a confined space, it produces high pressure. As the piston descends, the expanding steam pushes on it, producing torque.

On the second stroke of the cycle, the low pressure steam valve opens to allow steam out of the cylinder, into the steam exhaust manifold, which leads to the condenser.

The computer keeps the engine in the primary cooling cycle until the engine's insides are cool enough to switch back to the heating cycle.

When the steam pressure in the boiler exceeds some threshold, the computer switches to the secondary cooling cycle. This is a simple two stroke steam expansion cycle. It cools the engine a bit, but not as much as the primary cooling cycle.

In the first stroke, the high pressure steam valve opens, steam comes in, then shortly afterward the valve closes. The piston continues to descend, and the steam expands against it, producing mechanical energy.

The quantity of steam admitted is selected by the computer to have the same criteria as was used for quantity of water to inject, in the primary cooling cycle.

The second stroke of the cycle is the same as the second stroke of the primary cooling cycle.

The computer keeps the engine in this cycle until the steam pressure in the boiler drops below some threshold, after which it switches back to the heating cycle.

The reason why the idea is named as it is, is because the the engine will switch between the three modes of operation in a pattern that a human probably wouldn't predict. Also, I fully expect that the exhaust noise would be... well, odd.
-- goldbb, Feb 03 2013

Variable Stroke Count Engine Variable_20Stroke_20Count_20Engine
Like this idea, but only two modes of operation. [goldbb, Feb 03 2013]

What is the block made from? There aren't a lot of materials that can withstand the extreme thermal cycle this system would produce.
-- Alterother, Feb 03 2013


The engine block would probably use some type of zirconia ceramic, since it's been highly studied in the context of engines, and unlike silicon nitride, it's unlikely to degrade due to the hot, humid, environment inside the engine.
-- goldbb, Feb 03 2013


Even if you have a constant power load, a water pump for example, would be hard to keep steady RPMs or torque between cycles. [+]
-- piluso, Feb 03 2013


bigsleep, A Stirling engine takes time to warm up. The more powerful, the longer it takes. This engine idea start with a four stroke cycle when cold, and can start just as quickly as a conventional engine.

Also, it's very difficult to quickly adjust the specific power of a Stirling engine. Sure, done using a continuously variable transmission, but not exactly easily.

piluso, for each of the three modes, controlling torque is not that hard.

In the heating cycle, the computer controls the amount of air and fuel admitted into the cylinder, exactly as in a normal engine. In the primary cooling cycle, the quantity of water injected is controlled by the computer, and this directly controls the torque. In the secondary cooling cycle, the quantity of steam admitted is controlled by the computer, and this controls the torque.

Controlling torque obviously controls RPM.
-- goldbb, Feb 04 2013


I'm giving this a bun despite what I'm saying sounding like I'm shooting it down, I'm not. I've considered schemes such as this. In the big picture of things, the ideal vehicle would use electric only for moving the wheels, and the engine would be purely used as a generator. I won't go into all the details of using pure electric torque for the wheels but I am convinced it's superior in every measure. This frees the engine from performance requirements related to large RPM range, broad torque curve, throttle response, etc. Of course, this also makes variable valve timing and all of those features to improve the toque/rpm range irrelevant. To bail on the electronic valves and use single traditional cam would be simpler and probably less valvetrain losses. Permanent 6-cycle and just wait until it's hot to start the water injection. I think you may also be underestimating how much cooler it will run with steam sucking the heat out of the block. Advanced ceramics may not be necessary.

I'm also torn on the issue of water recovery, more specifically the issue of water freezing. It seems like a lot of complication, I really don't mind filling up a water tank if it's going to double how long the gas lasts.
-- AutoMcDonough, Feb 06 2013


You could avoid the problem with the water freezing by injecting antifreeze or chlorofluorocarbons or ammonia.
-- Kansan101, Feb 06 2013


// The engine block would probably use some type of zirconia ceramic, since it's been highly studied in the context of engines, and unlike silicon nitride, it's unlikely to degrade due to the hot, humid, environment inside the engine. //

Okay, I don't know anything at all about zirconia ceramic or silicon nitride. I was sort of assuming that the block would be made of some unobtanium metal alloy. I'll do some reading to try and catch up with you, but first let me get right to the crux of my point: how stretchy is zirconia ceramic? The mechanical process you're describing here is likely to produce a thermal cycle more extreme than any found in conventional combustion engines. The water injection is only part of this issue; the regular build-up and release of pressure not only within the cylinders but also in the boiler and the exhaust gas recovery system will also cause significant thermodynamic exchange (somebody correct me if that's not the proper term).

Basically, this thing is going to be rapidly expanding and contracting at irregular rates in a staggering array of locations and directions. A cast iron block asked to perform like this would crack within a few minutes. Even a highly ductile mega-ultra-high tech alloy like cryo-forged tempered tichrome would probably develop hot spots and microfissures after less than a thousand hours. Would your ceramic compounds be any better at taking this kind of punishment?
-- Alterother, Feb 06 2013


^^That's why there would need to be some Aluminum in the mix to allow for flexing. all very possible.

I like this. What's more, in my job I've been currently researching moisture removing processes; the water recovery in the exhaust is as simple as either passing it through a series of fairly unrestrictive dessicant sillica filters with a sump collector, or a cyclonic separator.
-- acurafan07, Feb 08 2013


AutoMcDonough, the advanced ceramics aren't *just* to retain more heat (although that's an important part of it), but rather to withstand the thermal shock that's involved.

I agree the using electricity to torque the wheels (instead of the motor directly), is the way to go.

And you're probably right, using a fixed six stroke cycle would probably be simpler and lighter, which in a moving vehicle is a very important thing.

This produces a potential problem: if we don't use steam from the boiler, and if liquid water from the boiler isn't taken out at a high enough rate, the boiler could overheat and explode.

There are a handful of solutions.

First, we use steam from the boiler in a small, separate steam engine. Since phase change from water to steam absorbs lots of heat, this keeps the boiler cool enough, and at a low enough pressure, to not go kaboom. We can give this steam engine it's own alternator, to simplify the mechanical portions of the system.

Second, we could use an exhaust gas diversion valve, which stops heating the boiler if it gets too hot. This throws away perfectly good heat though, which could be used to make mechanical power. OTOH, we might want this anyway, as a safety feature.

Third, if we can build the boiler strong enough, then we can use simple high pressure to prevent the water from boiling, no matter how hot it gets. A car's exhaust manifold is very very hot, though. Very high pressures might be needed. Would the water get hot enough to become a supercritical fluid? It probably depends on the mass flow rates of the water and the engine exhaust, but it worries me.

And as for freezing of the water recovery system... whether or not it's an issue would depend on what kind of system was used. For example, suppose we used a spray of concentrated liquid desiccant to remove water from the engine exhaust, used exhaust manifold heat to regenerate the dilute desiccant back into concentrated desiccant, and used the already existing condenser to turn the steam back into water. Common liquid desiccants are very freeze resistant. The only liquid water is in parts of the system (the condenser, etc) where it would be even without the exhaust water recovery system.

Alterother, although I'm not sure how stretchy zirconia ceramic is, I do know that it's highly resistant to thermal shock... which is probably a "good enough" characteristic for it to work.

And I'm not sure what you mean by "thermodynamic exchange." Metal fatigue? Thermal shock?

acurafan, unless the boiler is extraordinarily efficient, the water in the exhaust will be in the form of steam mixed with air (aka humidity), not liquid water droplets suspended in air. So a cyclone would be useless.

Silica gel desiccant is an amazing substance wrt adsorbing water, but... the adsorbed water will stay in place on it's surface, not drip off. Once all of the silica is covered with a thin film of water, the silica will stop adsorbing.

To get the silica to adsorb more water, you need to add heat to cook off the water that's on it's surface, then cool the silica back down.
-- goldbb, Feb 15 2013



random, halfbakery