h a l f b a k e r y
0.5 and holding.

meta:

account: browse anonymously, or get an account and write.

 user: pass:
register,

# Ice Engine

Phase-Change Power a'la Slush
 (+7, -1) [vote for, against]

Technical terms sometimes confuse the average person. Consider the phrases "flashlight battery" and "flashlight cell". The average person may use them interchangeably, but technically a battery is a group of cells, and so a cell cannot actually be a battery.

In the present case, there are the words "engine" and "motor" to talk about. Technically, an engine is a device that converts untamed energy into tamed energy (of any form, but usually motion). A motor is merely a device that uses tamed energy to create motion. A windmill qualifies as an "engine", but a gasoline engine should not be called a "motor".

Sometimes motors can be modified to become engines, however. In the present example, see the link about the "bent axis pump". It happens that if fluid is pressured into it, it can act as a motor. And this Idea is about how to turn that motor into an engine.

Once upon a time I saw a short video in which a hollow iron sphere about the size of a baseball (the shell was maybe half a centimeter thick) was filled with water, and then sealed and placed in a container of "dry ice" (frozen carbon dioxide). As you might expect, since dry ice is very cold, the water in the sphere turned to ice. You might not expect the sphere to explode with enough force to embed shards of itself in the walls 3 meters away, but it did.

When water freezes, it expands. Ice occupies about 10% greater volume than the original unfrozen water, and freezing water will exert tremendous force, in expanding that 10%. This is the force that we will tame in the engine described here.

The Ice Engine does all of its power-generating while at the fixed temperature of Zero Degrees Celsius, the freezing/melting point of water. Nevertheless, it qualifies as a "heat engine" because significant amounts of heat need to be added and removed from it, while it operates. Those two statements are NOT a contradiction. The key fact that elminates the contradiction is this: WHILE water freezes or melts, its temperature does not change; all the heat that is either removed or added is directly due to the work involved in the freezing or melting process. This energy is technically known as the "heat of fusion" (having absolutely nothing to do with nuclear fusion), because when individual items (like water molecules) become solidly linked together (as during freezing), it can be said that they have become "fused" together.

So, AT the freezing point of water, a bunch of heat gets added to make ice melt, and a bunch of heat must be removed to make water freeze. Only after all of the substance has completed this "phase change" (from the solid phase to the liquid phase, or vice-versa) can its temperature begin to rise or fall, from the freezing/melting point.

Constructing an Ice Engine is simple enough. We need a set of containers of water, all at the same temperature of Zero Degrees Celsius. Each container holds a different mixture of water and ice (slush, that is); one is all water and one is all ice.

Next, we open up the end of the bent-axis pump, where we can see different distances from that end to the tops of the pistons in the cylinders. In the shortest distance we put pure water, and in the longest distance (it should be 10% longer than the shortest distance, of course) we put solid ice. The mixtures go in the intermediary-distance cylinders, carefully matching the space available with the water/ice ratio. It happens that the exact same WEIGHT (or mass) of water should be placed in each cylinder.

The end of the pump is now completely sealed (we don't want any air at all in those cylinders!); no H2O ever needs to be added or removed from any of the cylinders. At the "swash plate", the tilted disc that the pistons are attached to, we may need to make some modifications. A sharp tilt is required for this type of pump to work as a motor, but here we are starting with about a 10% tilt, thanks to the amounts of H2O that were added to the cylinders. However, we can take advantage of the fact that we know that when water freezes, it will exert enormous force, so we can add some levers in this area, one for each piston, to exchange force to obtain distance, and convert the slight tilt into a sharp tilt for the swash plate.

Finally, we add some heating and some cooling apparatus around the assembly of piston cylinders. Based on the direction we want this engine to rotate, we want to add heat as the piston goes into the cylinder (ice melts and volume decreases), and we want to remove heat to make the piston go out of the cylinder (water freezes and volume increases).

It will probably rotate rather slowly. We may want this engine to be rather large, with more than a hundred cylinders, because of the time it can take for water to melt or freeze. Nevertheless, the Ice Engine will also produce tremendous torque, guaranteed.

 — Vernon, Sep 22 2007

Bent-Axis Pump http://www.animated...mpglos/bentaxis.htm
As mentioned in the main text. [Vernon, Sep 22 2007]

Lethargic power Generator Lethargic_20Power_20Generator
for reference [Ling, Sep 22 2007]

I.C.E. Internal_20crystallization_20Engine
[2 fries shy of a happy meal, Sep 22 2007]

Minto Wheel using fluids and gravity http://en.wikipedia.org/wiki/Minto_wheel
Sorta relevant, I hope, and rather interesting. [baconbrain, Sep 23 2007]

Animation of Gnome Engine http://www.keveney.com/gnome.html
Imagine this half-submerged in very cold saltwater. [baconbrain, Sep 23 2007]

 If the volume of water could be reduced* so that the time taken for freezing was also reduced, perhaps rpm might increase to something useful.

 One way to get over the first fill is use water in each piston, and keep both axis in line. Then allow the piston axis side to bend as the water freezes. One minor problem would be uncoordinated freezing of various pistons. Each piston would need to expand in a sinusoidal way.

*- really small
 — Ling, Sep 22 2007

[Ling], if the motor has enough torque, you can obtain decent RPMs just by gearing it up.
 — Vernon, Sep 22 2007

 I love this idea, and I find it to be written out far more concisely than my own. link

; ?
 — 2 fries shy of a happy meal, Sep 22 2007

Damn it Two fries! I read through this, strained my limp brain to understand it, and was all ready to slap it with pastry. Then you snatched it away at the last minute!
 — the dog's breakfast, Sep 23 2007

You can see [Vernon] coming miles away. Hello, [Vernon], fun idea.
 — normzone, Sep 23 2007

 [Ian Tindale], that will only work if the ground is colder than 0 degrees Celsius.

 Isn't this engine just like a Stirling engine, except it uses water instead of air?

 I wonder if there is a substance that would work better with this engine than water.

 I think that most of the time, in most places, heat will either need to be added, or taken away by a power source, but not both. I mean that the ambient temperature will act as either the heat source or the 'cold' source.

 Even if this engine was very well insulated, I don't think it would be very efficient, however, I do think that there might be some use for it.

Nature uses this phenomenon all the time.
 — BJS, Sep 23 2007

 Just some very cursory thoughts: the latent heat of fusion for water is about 144 btu/lbm. According to one source (see link) ice expands about 9% and exerts a pressure of about 100,000 psi (most likely the mean effective pressure would be about 1/2 this peak pressure).

 One pound of liquid water has a volume of about 28 ci.. 9% of this is about 2.5 ci. Expanding a fluid 2.5 ci at a pressure of about 50,000 psi yields about 10,400 ft lbs of work. One btu is the equivalent of 778 ft lbs, so were talking about 13.4 btu of work from 144 btu invested for a net thermal efficiency of about 9% (this is ideal, of course.... actual would be less). Then there's the problem of actually containing this pressure.

I do like the idea, but I doubt it could be practical. Any more info on the pressures that could be exerted by expanding ice? The performance hinges on it. Also, an appreciable temperature difference would be required for any chance of good performance.

[devilishadvocate], there is an assumption here that the pistons will allow the freezing water to expand without the cylinder walls rupturing first. And the only temperature differences needed are the hot-source compared to the frozen/melting ice, and the cold-source compared to the liquid/freezing water. The slush stays at Zero degrees Celsius ALL the time, because there is no significant additional expansion or contraction to be gained by either cooling or heating the slush to any other nearby temperature.
 — Vernon, Sep 23 2007

The freezing point of water lowers as pressure increases, so the more resistance the pistons have, the more the water will need to be cooled to freeze and expand. Once the pressure is removed you will need to add a lot of heat to get it back above freezing.
 — marklar, Sep 23 2007

[marklar], I agree, but I tend to doubt that we will be trying to extract that much force from this engine. Doing so sounds like an invitation to cause the cylinders to rupture, heh.
 — Vernon, Sep 23 2007

 I agree Vernon, the pressure can be controlled (lowered), but this would also necessarily reduce the work done as well, thus lowering the efficiency.

Have you considered that hydraulics is a good way to harness the power? Perhaps ice/water in a hydraulic fluid could even work. The water is more dense and would settle. As the water freezes it would displace the hydraulic fluid and hydraulic pressure could be transferred via high pressure lines to hydraulic motors.

 I don't like the bent-axis part of this for a couple of reasons. First, the angle between the two halves is going to be very slight, what with only a 10% expansion. The rotating parts are going to have incredible amounts of pressure pushing into the bearings, and only a slight amount of pressure causing rotation. (A flywheel might help, if the thing was turning at a higher rate.)

 I think it would work better if the pistons were driving a small-diameter crankshaft, more like a conventional piston engine. That way, all the pressure goes into rotation.

 But not completely like a conventional piston engine, more like--well, exactly like--a rotary-radial aircraft engine, such as a Gnome rotary engine.

 In a rotary radial, the cylinders rotate, while the "crankshaft" sits still. The whole engine goes around and around. (See link.)

 If the ice-engine proposed were splayed out into a rotary radial engine, it would get a flywheel effect and it would be able to use a crankshaft of small diameter or throw.

 The other advantage is that the cylinders would be separated from each other, making temperature management easier. As proposed, the cylinders are going to be heating and cooling each other, unless the diameter is huge (which would make the angle issues even worse).

 A radial layout would definitely have a hot side and a cold side, well separated. It could be run by submerging it halfway into a bath of cold saltwater. The cylinders would dip into the saltwater and freeze, then raise up into the sunshine and melt.

 Come to think of it, I'm sure now that I've seen such a device described before. It didn't use ice, but it did dip in-and-out of hot-and-cold. I'll see if I can go ogle it. (I meant "Google", but my Firefox spell checker prefers "go ogle".)

 Well, I didn't find anything in a quick search, but I recall now that the thing that I was thinking of used phase change to run. What it had, though, was a series of tanks arranged radially, with a working fluid like Freon that traveled through hoses between them. The tanks dipped into hot water, the freon boiled out and condensed in another tank on the high side.

 The problem with that setup was that the heat was at the bottom, and the cold at the top. A rotary-radial ice engine could have a more natural arrangement for the heat and cold.

<more ogling> It's a Minto wheel that I'm thinking of, and it may not be relevant at all, as it uses weight, not mechanical pressure, to turn. (See Wikipedia article.)
 — baconbrain, Sep 23 2007

 //If the ice-engine proposed were splayed out into a rotary radial engine, it would get a flywheel effect//

Huh?
 — Ling, Sep 24 2007

 The splayed-out part takes a takes a bit of three-dimensional visualization. Look at the bent-axis pump animation and make the cylinder block sorta cone-shaped. The cylinders would be splayed out along the cone, and the ball joints would allow the pump to keep working just fine.

 Now, splay the cylinders and pistons out more and more on a bigger, flatter cone. You could say you are taking them from being arranged in a cone with practically parallel sides, all the way to a cone with practically zero height, so the cylinders are radiating straight out.

 See, a bent-axis engine and a radial-rotary engine are extreme examples of the same arrangement. (You've got to fiddle with the middle to make a crankshaft, but it's the same idea.) Splay and squeeze the cylinders on the angle with the center axis of the engine, and you'll see they are related. The bent-axis pump page even points out a comparison to a radial pump.

 A swashplate engine and a non-rotary radial are similarly related/same. One is squished together, one is splayed out.

 I hope that clears up the "splayed-out" part. I worked all this out one feverish night, and forget that not everyone else has had malaria.

 As for the flywheel effect, the further the rotating mass of the cylinders and pistons go out, the greater the flywheel effect. Look at the rotary-radial link and read the advantages of the engine type.

 I hope that helps. I rewrote several times, as ideas were coming to me, but left some confusion in there.

(I went down to the river with my girlfriend just now, and instead of watching the sunset I was figuring how to use the cold water to run a rotary-radial Stirling engine.)
 — baconbrain, Sep 24 2007

Bigsleep, I was thinking the same thing. I've considered various approaches myself in building a practical heat engine using oil as the working fluid. Have you ever checked out how much some oils expand with increasing temperature? Some of them are very impressive, and with very low heat capacities to boot. Plus, hydraulic components are very inexpensive and reliable. Lots of potential there.

baconbrain, Yeah, I got the bit about having the legs splayed out. It was the flywheel effect that tickled me. I think you could probably compare the speed of the motion with the rate that paint dries.
 — Ling, Sep 24 2007

By the way, I've been trying to find what is the limiting ratio for a step up gearbox, but failed completely other than some obscure reference to the multiplication of the friction of the final shafts, which will finally cause lock-up.
 — Ling, Sep 24 2007

 [baconbrain], it seems you missed my solution to the 10% angle problem. I was fully aware it existed, and wrote:

 "A sharp tilt is required for this type of pump to work as a motor, but here we are starting with about a 10% tilt, thanks to the amounts of H2O that were added to the cylinders. However, we can take advantage of the fact that we know that when water freezes, it will exert enormous force, so we can add some levers in this area, one for each piston, to exchange force to obtain distance, and convert the slight tilt into a sharp tilt for the swash plate."

I don't object to your solution, though. Swash plates work but are not the most efficient things, when you want to convert reciprocating motion into rotary motion.
 — Vernon, Sep 24 2007

 Couple of fixes to problems: the cylinders could be cone shaped below the piston travel so that they don't explode.

 You could use hydraulic gearing. The piston pushes a plunger with a large volume end and a small volume, large travel end.

Disclaimer: I still think this engine is completely impractical.
 — marklar, Sep 24 2007

"Sleet engine"
 — Ling, Sep 24 2007

[Vernon], I like the way you think [+]. This might be interesting in some sort of snowmobile for use in cold temperatures.
 — acurafan07, Sep 25 2007

frozen water = frozen engine which will burst....It is much easier to heat water to boiling and use steam...but that's been tried.

[gladlynt], bursting is likely ONLY if the confined freezing water has no room to expand. The essence of this Idea (and a couple others in the links) is to allow it to expand in a controlled way. There are different ways to implement that control, of which one is described in the main text here.
 — Vernon, May 15 2008

I'd like to see the transmission, clutch, and gearshift for this model.
 — normzone, May 15 2008

 [annotate]

back: main index