The Giant Magnetocaloric effect is quite novel, and perhaps you will see it in a refrigerator soon. Basically, when an alloy of Gadolinium is subjected to a magnetic field, it gets hotter. When the magnetic field is removed, it returns to the same temperature, provided no heat was lost in the meantime. The working point can be adjusted between about 30K to room temperature, depending on the alloy. So I was thinking, if different alloys are stacked, and a magnetic field swept down the stack, from hot end to cold end, then the system becomes a heat pump from 30K - basically in one operation. So simple, it could be done at home.-- Ling, Aug 29 2006 Basic description of the effect http://www.sciencen...98/3_28_98/fob3.htm [Ling, Aug 29 2006] Tunable alloys http://link.aip.org...k/?APPLAB/70/3299/1 [Ling, Aug 29 2006] Magnetic refrigeration http://en.wikipedia...netic_refrigerationSignificant advances in room temperature applications since the idea was posted. [BunsenHoneydew, Dec 04 2010] That article repeats itself halfway through. Now my head really hurts.
So your idea is a stack that becomes a heat pump. I don't follow how that works. But it is early here. Ow.-- baconbrain, Aug 29 2006 The only way to avoid pumping heat in both directions is to constantly move the elements in and out of contact as you sweep the magnetic field.-- ldischler, Aug 29 2006 Let me see. If you had a stack of Gadoliums, and magnetically-heated up the first one, the next one would get hotter, just from contact. Then, when the field is applied to the next one, it heats up further, and transfers heat by contact to the one further down, but also back to the first one, which is a waste, I think. I can see a pulse of heat travelling along the stack, though. Which may mean something is getting cool.
Hmm, if you added a heat transfer fluid and pushed it along, you'd increase the pump effect . . .
Or if the magnetic field moved each block just enough to make and break heat conduction in the desired direction, or even the expansion of the material from heat did that . . .
So, #1 gets the GME and heats up #2, then switches off and cools and breaks contact, so #1 is colder than it was. #2 gets the GME and heats up #3, then switches off and cools and breaks contact, so #2 is now warmer than #1 is, but cooler than #1 will be when #1 gets GMEed again. So that seems to be pumping heat. Is that legal?-- baconbrain, Aug 29 2006 Please let me explain why I think it will work in more detail: Consider a stack, A, B and C. A is the end which is supposed to radiate heat, so it has a finned attachment. C is the end which is supposed to absorb heat. It also has a finned attachment.
Now, lets apply a magnetic field to A. It gets hotter, and sends heat to the finned attachment, and B. Some heat is lost via the fins, and B gets a little hotter. Now the magnetic field is moved to B. It gets hotter. But now A is colder than C, because we removed the magnetic field at A, which had already lost a little heat to the fins. So more heat will move to A than C, because the temperature differential is greater. Now the magnetic field is moved to C. C gets hotter, but not as hot as B did before (since more heat went to A). It does appear that heat is pumped, but there is a limit associated with "So more heat will move to A than C". The limit is caused by situation where the same heat moves to A and C, because A and C are already at some temperature gradient caused by the operation of the mechanism.-- Ling, Aug 29 2006 Bigsleep, I'm not sure about the waveform. Let's draw an analogy: a 3 phase motor requires a moving field to drag a conductor along. The moving field is important, but the shape of the waveform is normally sinusoidal. A sawtooth waveform wouldn't make much difference, because a sawtooth is made from many sinusoidal waveforms all put together, the main one is the fundamental. I don't know if it is a good analogy, but consider also in your scheme that although there is a big differential between 100% and 0%, the other differentials have longer to act.-- Ling, Aug 29 2006 how fast do these alloys cool back down? i cant seem to get a grasp on the scale of this thing.-- bleh, Aug 29 2006 bleh, that is a very good question. I think it ought to heat and cool as fast as the magnetic field is adjusted. Useful for accelerated testing of devices?-- Ling, Aug 30 2006 Q: If a magnetic field can increase a thing's temperature, can the process be reversed and a change in an object's temperature be made to create a magnetic field?-- zen_tom, Aug 30 2006 Q: Isn't it true that energy/ matter can neither be created nor lost, only converted? If so then where is the trade off? The motion of the magnets alone can not count for all of the heat being generated at the molecular level. Something else in the reaction must be sacrificed to continually produce heat each time the magnets are applied. Otherwise you would be able to create a perpetual reaction and unless I missed that part of my physics class, that isn't possible.-- CNIII, Aug 30 2006 I don't think this will work as described, but I have an idea extension that just might work.
It involves heat pipes, specifically non-capillary heat pipes which act as a thermal diode, letting heat flow only upwards.
Construct a mechanism with an absorber plate at the bottom (just a finned hunk of metal), connected to a chunk of gadolinium in the middle via one set of heatpipes, connected to a heat rejector (just another finned hunk of metal) at the top via another set of heatpipes.
When the field is turned on, it will heat up and reject heat through the upper heat rejector. When the field is turned off, it will draw heat from the absorber plate at the bottom.-- Freefall, Aug 30 2006 //I think it ought to heat and cool as fast as the magnetic field is adjusted//
let me see if i understand this (and i probably dont) but ifi were take a glob of this stuff cooled to 30K and apply a magnetic field to it, it heats up, then as soon as the field is switched off, the thing snaps back to 30k quickly? what kind of temperature differential is created?-- bleh, Aug 30 2006 I'll try to answer all (apologies if I missed anyone). Freefall, I am interested to know why you think it won't work, because at the moment I tend to believe it will. Mechanical systems will, of course, work. The refrigerators, that have already been constructed, consist of a ring of material (torus) that is rotated through the magnetic field. My idea is to try to remove mechanical movement, and don't forget that the magnetic field could be swept rapidly (assuming, of course, that the system actually works). zen_tom, //can the process be reversed and a change in an object's temperature be made to create a magnetic field?// I don't think this is possible, because what is changed is the ability of the material to be magnetised by an external magnetic field. I think it is called paramagnetism? It's a Curie type effect. The same happens with Steel: get it up to about 650C and it cannot be magnetised. By the way, the first discovery of the magnetocaloric effect was with hot Steel. Bleh // what kind of temperature differential is created?// none, unless during the time it was hotter, some heat was lost, and then when the magnetic field is removed, the final result is that the material is a little colder. CNIII, it's not magic: the references, and a little google searching may explain better than I can.-- Ling, Aug 31 2006 my question was how hot do these things get when the field is applied?-- bleh, Aug 31 2006 The wikipedia article on magnetic refrigeration has some diagrams of refrigeration cycles.-- wiml, Sep 01 2006 bleh, thanks. The best I can see is a 27 degrees C change using a permanent magnet. It's in the table in the Wikipedia link: "The less repetitive..."-- Ling, Sep 01 2006 [ling], I don't think it will work as described because you don't have any mechanism in place to control the direction of the heat flow. As one section is activated, the now excess thermal energy will flow away from the activated section into the surrounding material. When the field is moved, the heat will flow from the heated material back into the now cooler material.
If you break the device into segments and add heatpipes between each piece, you can sweep the stack and get the heat to flow.-- Freefall, Sep 01 2006 Freefall, yes, you are right. The overall heat gradient is the same to each end, whatever is done in the middle. The only compensation I can think of, is a small chance that the magnocaloric effect changes the thermal conductivity in different directions, but this is groping at straws. I hereby fish myself.-- Ling, Sep 01 2006 No need to fish yourself, it's a good seed for thought. It simply wasn't complete. Many breakthrough ideas start out as a "what if". Something like this would probably be patentable, except that it's now been placed on a public forum.-- Freefall, Sep 01 2006 When the plate heats up under the field, where does that energy come from? I understand that if you put a fluorescent bulb in a magnetic field under a power line it will light up, but by doing so burdens the electricity in the line so you are stealing power. Ling refers to permanent magnets here which have no power inputs. I have to thing that the heating plate must pull in ambient heat from the air, thus cooling its vicinity. If so one would predict that this effect would not work in a vacuum.
I do like that this could be done with no moving parts, the magnetic sweep being done by electromagnets somewhat in the fashion of a coilgun. You could use it for an air conditioner. It would never wear out.-- bungston, Dec 01 2010 Based on the article, what actually changes first is not the temperature, but the specific heat. That is when the magnet is activated, the material rejects heat to the environment (including itself), so the heating is a secondary effect.
As far as the energy source, since there does have to be one, it appears to be the force involved in creating or moving the magnetic field.-- MechE, Dec 01 2010 But will it make all my fridge magnets fall off?-- BunsenHoneydew, Dec 04 2010 random, halfbakery