Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
Guitar Hero: 4'33"

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

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



Yet Another Heat Engine

It will work, but how well? Does it matter?
  [vote for,

The world is full of heat engines, gadgets that attempt to convert some of the energy of flowing heat (from hot to cold) like a waterwheel converts some of the energy of flowing water (from high to low). There are steam pistons and steam turbines; gasoline and diesel and Stirling pistons, and Wankel rotaries and gas turbines. There are many others that never saw the success of mass production. All of them have a fundamental limitation with respect to how efficient they can be, due to Laws of Thermodynamics. Just reaching 50% efficiency, of conversion of heat energy into mechanical energy, taxes our materials science to the utmost -- and such efficiencies are only found in the largest of power plants. Simply because we can do better than 50% with other technologies (electric fuel cells start at over 60%), the long-term prospects for heat engines appears to be dim. But in the short term, their heyday is not yet over, so, why not have another to play with? :) If nothing else, it may simply be fun to contemplate....

Every heat engine has a "working fluid" that starts in a hot place, flows through the mechanical innards of the engine, and then exits to a cool place, sometimes getting recycled afterwards. The engine described here is of the recycling variety, having a burner to generate heat, and a radiator to dump it. The engine itself has only 3 or 4 moving parts (not counting subsidiary stuff like ball bearings), and is based on the gear pump.

An ordinary gear pump might be crudely portrayed as (*|*), in which the asterisks represent the gears, and the endpoints of the vertical bar represents the entry and exit points of the flow of fluid (the fluid actually flows AROUND the gears, not between them). This heat engine has 3 gears, and overall might be portrayed something like this:

1. | | 2
.( *|*|* ) ----- engine
3. | | 4

(Ignore the horizontal hyphens.) The vertical pipes connect to both the burner and the radiator, and that's basically all there is to this engine. How and why it works, however, is going to take some more explanation!

Lets start with the pipe that exits the burner and carries the working fluid to the engine (say, #3). AT the engine, the hot fluid encounters two gears, and the fluid splits to flow around the gears. We focus for the moment only on that half of the fluid that flows all the way around the left side of the left gear, heading toward the left side of the radiator (#1). It's pretty obvious that, waterwheel-like, the tendency of heat to flow from hot to cool is being directly exploited.

Next, cool fluid leaves the right side of the radiator, heading toward the right gears of the engine (#2). It splits when it reaches the gears, half of which flows all around the rightmost gear, returning to the burner. The fourth moving part is a one-way valve, to ensure that fluid trying to leave the burner only goes out the left pipe.

The most interesting thing about this engine is actually what happens at the middle gear. Remember, half the hot fluid is flowing from left to right at the portrayed bottom of the middle gear, and half the cool fluid is flowing from right to left at the portrayed top of the middle gear. Thus, at upper-left and at lower-right of the engine itself, hot fluid directly flows toward and meets and mixes with cool fluid, producing warm fluid. It is actually the mixed warm fluid that flows toward the radiator (#1) and burner (#4). ALSO, it is that direct "attraction" of hot and cool at those two points, that is the real motivator behind this engine. I'm not totally certain, but this may be a true rotary analog to the classic Stirling piston engine (which have been known to work at 45% efficiency, by the way!). What fun!

... --->
1. |. .| 2.
.. A < A
( * . * . * )
.. V > V
3. |. .| 4.
... <---

Ignore the dots. Starting at the burner, flow up pipe #3, and split at the V. Half continues on the left side of the leftmost gear to the A, then up pipe #1 to the radiator. From the radiator, flow down pipe #2 and split at A. Half continues on the right side of the rightmost gear to the V, then down pipe #4 to the burner. The OTHER half of the hot from #3 follows the > between the two Vs, and the other half of the cool from #2 follows the < between the two As. OK?

Vernon, Jun 24 2003

Gas Refrigeration http://www.gasrefri...rs.com/faq.htm#work
Here's a decent explanation [Vernon, Oct 17 2004, last modified Oct 21 2004]

Is this close? http://www.st3f.com...ry/vernonengine.png
No one-way valves and comes unstuck in the last paragraph... [st3f, Oct 17 2004, last modified Jul 05 2015]

Animated Gear Pump .GIF http://www.svce.ac....Unit-V/GearPump.htm
duplicate of link from Variable Transmission idea [Vernon, Oct 17 2004, last modified Oct 21 2004]

Stirling engine operation http://www.stirling...theory-english.html
includes an animated gif [Vernon, Oct 17 2004, last modified Oct 21 2004]

st3f's 2nd attempt. http://www.st3f.com.../vernonengine_b.png
closer? [st3f, Oct 17 2004, last modified Jul 05 2015]

Some relavent rotary steam engine info http://www.svce.ac....5/unit5-module5.htm
Several different contraptions. [RayfordSteele, Oct 17 2004, last modified Oct 21 2004]

NiTinol engine http://bednorzmulle...ngines/demo532.html
Click the image to start the movie [kbecker, Oct 17 2004, last modified Oct 21 2004]

At Last! A fairly accurate image! http://www.nemitz.n...rnon/RSEnoregen.gif
If there were fewer gear teeth, then THIS is the engine described above! [Vernon, Nov 02 2007]

ROVAC http://books.google...%22%20rovac&f=false
As mentioned in an annotation. [Vernon, Jun 16 2011]

Please log in.
If you're not logged in, you can see what this page looks like, but you will not be able to add anything.


       What makes you think that heating a fluid will make it flow away from the heat? Or, are you setting up convection currents using a density change in the fluid? Maybe I'm not understanding this yet.
st3f, Jun 24 2003

       Observation, st3f; people have been observing heat flow from high-intensity to low-intensity ever since cavemen first studied fire. And, before anyone else mentions it, YES, I HAVE thought of combining this idea with the Variable Transmission idea in the CARS section, meaning that the overall gadget would be BOTH the engine and the transmission, all in one!
Vernon, Jun 24 2003

       Heat flow does not imply that the fluid is going to move.
st3f, Jun 24 2003

       Ah. Well, the most effective mechanism in Nature, for carrying quantities of heat, is "convection". That always involves the flow of fluid...for something comparatively tougher to understand (I still have trouble with it), go study the explanation of how flaming natural gas is used to do refrigeration. Anyway, should you understand the gas refrigerator, understanding this will be a snap.
Vernon, Jun 24 2003

       Hence my question about convection currents. I'm not critcising. I'm trying to understand. At the moment, I do not.
st3f, Jun 24 2003

       For an alternate explanation of the engine, not referencing convection per se, consider that hot fluid expands, and the one-way valve will force that expansion to occur in one direction only -- toward the engine. And the radiator cools/shrinks the fluid, so that is all the more reason for the hot fluid to flow through the engine. (Still, that flow of fluid IS "convective", so....)
Vernon, Jun 24 2003

       You've added a link to 'Gas Refigeration'. Is the fluid in your engine changing state? Or is this (to put it crudely) an unusually shaped lava lamp with a couple of turbines in it. (Wow, am I argumentative this morning. I hope that no offence is taken)
st3f, Jun 24 2003

       I deliberately left that aspect undefined. If this engine can qualify as being sufficiently similar in operation to the Stirling cycle, then the working fluid would be gaseous purely and only. But as a variant of the gear pump, mostly used for oils, the working fluid could be liquid purely and only. I DO know that the material expansion that occurs when liquid converts to gas is a Good Thing For Heat Engines, and so I wouldn't want to keep it from being implemented. Take your pick!
Vernon, Jun 24 2003

       st3f, nice try, but nope. You have drawn the equivalent of a multistage turbine, except that it can't work at all, simply because the gear teeth are supposed to MESH, and so the fluid cannot pass between the gears. As I stated in the main text, the fluid goes AROUND the gears. You should see the CARS: Transmission: Variable Transmission idea, and see the links there (especially the one with the animated gif) for gear pumps. Then re-read the main text here more closely.
Vernon, Jun 24 2003

       ...reading about gear pumps...
st3f, Jun 24 2003

       ...still reading about gear pumps...
st3f, Jun 24 2003

       lurch if the hot has expanded and the cool has contracted, then it figures that the excess hot will tend to occupy the space made available by the shrinkage of the cool. That is a kind of attraction....   

       UnaBubba, I can draw, but not here in the HalfBakery. If I could just use a fixed font here, I might manage a better ASCII sketch. Sorry.
Vernon, Jun 24 2003

       doing some work... real work, that is... will doodle again later...
st3f, Jun 24 2003

       Alright. HVAC Engineer stepping into the fray.   

       This is simply another stirling engine, missing a few details. All told, there needs to be compressibility, an engine that extracts power from the compression / heat, a heater / compressor, and some form of heat sink.   

       The standard gear-pump scheme includes two meshing gears with fluid and is used to raise pressure levels, I've never seen it reversed to extract power, and I doubt it would be efficient, if it would spin at all, since there is no preferred directionality built-in to a gear pump system. Thus, the pumps would simply evolve into a standard multi-stage turbine anyway.   

       The reason you cannot get more efficient than a stirling engine is because you cannot extract more heat than the surrounding ambient temperature of the system. Only in absolute 0 would you raise the theoretical thermodynamic efficiency to 100%, ie. the amount of available, convertible energy is directly related to the maximum and minimum enthalpies (internal heat energies) of the fluids.   

       Yeah, a picture would be really really nice. There are a million places to scan copies, and a million ways to post 'em online. Heck, I even figured it out once, and I'm not exactly net-savvy.
RayfordSteele, Jun 24 2003

       RayfordSteele, I'm well aware that the temperature differential between hot and cold is a major factor limiting the efficiency of a heat engine. Do note that you could APPROACH 100% efficiency if the hot temperature was hot enough (we can do 100,000,000 degrees in an exploding A-bomb), and the cold-sink was COMPARITIVELY cold enough (room temperature). I shall maintain that limitations on materials are actually a more important factor, with respect to extracting mechanical energy from any such extreme temperature differential.   

       Next, gear pumps do indeed work backwards as motors if desired, and not significantly less efficiently than electric motors (90%+). The ordinary gear pump has one inlet and one outlet, so any compressed fluid applied to the inlet WILL cause the gears to move. Yes, the three-gear thing presented here has no naturally preferred inlet, so I force the issue with a one-way valve.   

       [Generic definitions for ALL: engines and motors are frequently two different things. Engines convert raw/wild energy into a 'tamed' form; motors convert some tamed form (electricity, air pressure, etc.) into mechanical motion.]   

       Part of the reason I compared this thing to a Stirling engine is because neither does a lot of compression/expansion. Yes, there is some, but much of what a Stirling does involves just shoving air (that device's traditional working fluid), between hot and cold regions. I've added a link.   

       In this gear-pump engine, the best working fluid would probably be some liquid with a high temperature/expansion coefficient, and a high breakdown temperature. Mercury, perhaps (there's a reason why they use it in thermometers!), although of course that would not be ecologically acceptable. I wonder if some fairly simple organic compound (only the simple ones can withstand lots of heat) might have some "springiness" in its molecular structure, so that EACH MOLECULE physically expands when hotter....
Vernon, Jun 24 2003

       2nd diagram added. Hopefully closer to Vernon's imagination. I really can't see any advantage this has over any other method of capturing convective motion. Then again, I don;t understand the phrase, it is that direct "attraction" of hot and cool at those two points, that is the real motivator behind this engine.'   

       As regards efficiency, the two additional gears needed to make a gear pump run in two directions will have much more fiction than a single 'waterwheel' type of approach. Plus, I find the claim that gear pumps run well as motors somewhat surprising, since the main use of a gear pump is to pump against a high back-pressure.
st3f, Jun 24 2003

       hang on - are you extracting energy from the cold->hot section or is the gear there to input the energy to move against the temperature gradient?
chud, Jun 24 2003

       The fatal flaw with this design is that your gear pumps move the same volume of gas in both directions. This will produce no power. As you said, your system has no preference for direction of rotation. Adding a one-way valve wont cause it to start rotating in the right direction, it will simply prevent it from rotating in the other direction.   

       You need to change the design so that a smaller volume is pumped from the cold to the hot than travels from hot to cold for each revolution. The smaller volume of cold gas expands when heated, and exits as a large volume of gas, allowing it to run the other pump as well as produce excess power. This is similar to the principle of how a jet or turboprob engine works. (replacing grear pumps with turbines, using ambient air as the working fluid so no radiator is necessary, and heating the fluid by internal combustion)   

       With this change, I believe that this would be a working heat engine. I'm not prepared to guess how its efficency or power/weight ratio would compare with a stirling engine.
scad mientist, Jun 24 2003

       I don't doubt that gear pumps can run well as motors. Consider that on one side of the motor, there is a higher pressure than the other. This pressure is applied to the surface area of the gear teeth. The gear teeth around the outside have a pressure differential causing torque on the gears. Of course the meshing gears also have unbalanced pressure, but only half as much because the meshing causes only the area of one tooth to ever be exposed at one time. The net torque is equal to the pressure differencial times the surface area of once side of one tooth times the radius of the gear.   

       For extracting energy from a high pressure uncompressible liquid flowing at a rate where not a lot of energy is lost to stiring the liquid, and assuming the seals don't leak (and also have little friction), a gear pump would be close to 100% efficient. For this application, I'm not convinced that a gear pump is the _best_ device, but that could depend a lot on the design of the system.
scad mientist, Jun 24 2003

       st3f, if you take off the eight arrows at the four corners of the diagram, (and the four associated loops) you got it! Think about it for a minute: you have drawn four pipes from burner to engine, and four from radiator to engine, but I've only described two each in the main text here. Thanks! (Naturally, once the connecting pipes reach radiator or burner, the plumbing within those units can be as spaghetti-like as needed.) --Oh, and the close-fitting HOUSING for the gear-pump/engine seems to be missing.   

       chud, in any heat engine, cold fluid always flows toward the hot zone, one way or another. I'm claiming that at the lower-right engine-output pipe, AND at the upper-left output-pipe, hot fluid mixing with cool causes a lowering of pressure of the hot to the extent that more hot will want to flow toward those places.   

       scad mientist, because of pressure changes such as just described, this engine will indeed work. I did not claim that it would have terrific efficiency, however. I have also previously indicated/agreed that some modified variant, allowing for more expansion/shrinkage of the working fluid would probably be a worthwhile improvement. One possibility that springs to mind, along those lines, is a real screamer of an air conditioning compressor that was touted some years ago in Popular Science (well before the Internet). It was called ROVAC (ROtary Vane Air Conditioner). Modified for use here, the vanes would take the place of the gear teeth of one gear, and some room for expansion could be made available at that side of the engine.
Vernon, Jun 24 2003

       I still dont see how the fluid does work on the hot->cold _and_ cold->hot temperature gradients <checks textbook> says here: heat engines (hot->cold) do work - heat pumps (cold->hot) need work done (I may be confusing the movements of fluid and energy)
chud, Jun 24 2003

       chud, it is the linked gears that is doing the work of pulling the cool fluid toward the places where it can interact with the hot. I deliberately used quotes on that word "attraction" in the main text because I know that it is really only a one-way kind of attraction, hot toward cool, as you stated. Yet there remain two places in this design where hot can be attracted to cool. It is possible that some variant of the three-gears-in-a-straight-row may be an improvement, with respect to minimizing the transport of cool fluid. Certainly worthy of thought.... Thanks.
Vernon, Jun 24 2003

       Sorry, I was thinking Carnot engine, not Stirling.
RayfordSteele, Jun 24 2003

       RayfordSteele, thanks for the nice steam-engine link. I note the text there stressing the need to allow for expansion/cooling of steam, and certainly that is quite important for any such engine. I've known for years about the 1600-fold expansion in volume that happens when water is boiled into steam. There certainly isn't room for that in this design! Which is why I have used the word "fluid" so much, without being specific. I should mention that I thought of this idea quite a number of years ago, when I was contemplating the Stirling engine, and wondering if a purely rotary form of it could be designed. Stirlings really are quite efficient (40%+) for heat engines, and I was trying to more thorougly understand why. The average auto engine tops out at about 30% efficiency, last I heard, and rotaries are inherently smoother (and often have fewer parts) than reciprocating engines, so there seemed good reason to try to imagine a rotary Stirling. I did not claim to have succeeded here, but I think there is some resemblance in the principle of operation.
Vernon, Jun 25 2003

       reensure: not hours. Probably no more than 10 minutes, as is true of old-fashioned steam-engine cars. Not to mention that modern steam cars have flash-boilers able to generate steam in only a couple minutes. And, remember, I've not actually tried to present this as a steam engine. If its working fluid merely expands a modest amount (say 1%) upon application of heat, its axles will probably start turning quite early, even if slowly. I'm tending to think that this is not an inherently high-speed engine, anyway, but if it generates enough torque at low RPM, then obtaining speed merely depends upon the transmission gearing.
Vernon, Jun 26 2003

       If you want it simple try a memory wire engine (link). As I understand the description of the gear setup the concept of bringing together hot and cold by some motion that's generated by the engine itself is the same.
kbecker, Jun 27 2003

       Is nice your idea. Please contact me at dinamic@email.ro. Thank you
danm, Mar 21 2010

       I do believe that there is no reason for the fluid to circulate.
WcW, Mar 22 2010

       [WcW], try "temperature difference" as a reason.
Vernon, Mar 22 2010

       Missed this one originally. Temperature difference is not a reason for fluid to circulate, it is a reason for heat to circulate, not the same thing.   

       Contrary to the claimed similarity an absorption chiller has a two key steps that this lacks. The first is a phase change, that is the boiling amonia providing a pressure diferential. This is what causes the amonia to move up and into the condenser.   

       The second, and this is the critical bit is the absorption portion. The low temperature gaeous ammonia wants to mix into the water. This results in it being pulled out of the air, providing a net pressure drop, and the impetus to keep pressure from building up and halting the system.   

       This idea has nothing on the cold side causing the working fluid to return to the heater.   

       It is possible this could produce a convective current, but the forces involved in that are minimal (in fact, it becomes a gravity engine, with the difference in density being the driving force, not a heat engine as such). Even then, for that to work, you would want both the heater and chiller to be on the vertical tubes opposite each other so the gravitational forces wouldn't balance.
MechE, Jun 27 2011

       Not so fast, [MechE]. I'm fairly sure that there is a direct tendency for a liquid to flow down a temperature gradient, independent of changes in phase or density. I don't yet understand the reasons in detail, but temperature contributes to a liquid's potential, as do pressure, gravity, and solute content (which causes osmosis in the special case of water). I think it may be a simple statistical tendency for hotter (faster) molecules to end up further from their starting point than colder ones, which is observed on the macro- scale as a flow of the liquid, and results in an increase in the entropy of the system. Or it may have something to do with surface tension.   

       For example, when a thin film of water is placed in a pan, and the pan is heated in one spot, the water can be observed to move away from the hot spot, leaving the bottom of the pot dry, well before boiling occurs. (Expansion and convection would make the water film slightly thicker where the temperature is highest).   

       Whether useful work can be extracted from this is a nother question.   

       All of your other points, I endorse.
spidermother, Jun 28 2011

       The "water moving away" is due to evaporation from the heated spot, and possibly thrust generated by the water vapor (This is also why water skitters over a hot surface). Remember, water evaporates before it reaches boiling, just not as fast.   

       I'm not saying there is no motion at all, even in the worst case you will have brownian motion, which does increase with temperature. But expecting that to produce any useable output is like throwing a bouncing ball into a room, and expecting it to empty the trash.
MechE, Jun 28 2011

       //evaporation// Plausible, but unlikely. I have wondered whether it's a distillation process, where the water evaporates and re-condenses in its new location; but it seems to happen much too quickly for it to be just evaporation. It takes a few seconds for a clear patch to appear, but minutes for the water to evaporate completely at the same (low) heat input.   

       I don't think it's thrust due to evaporation either, since it happens at much lower temperatures. It also happens more readily if the water is cold; it therefore seems like a different process from the skittering, which occurs at higher temperatures.   

       The increase in water potential with temperature is a real thing; it has to be taken into account when calculating osmotic pressures etc. However, since the height difference in the thin layer of water is only a couple of millimetres, the potential difference need only be a couple of Pascals.   

       I admit that I'm not sure either way about this.
spidermother, Jun 28 2011


back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle