© Harold Aspden, 1997

Research Note: 05/97: January 29, 1997


I have decided to compile this Research Note as a diversion from preparing a response to a U.S. Patent Examiner's rejection of my Patent Application Serial No. 08/579,991 which has a U.S. filing date of 12/28/95. The Examiner's comment reads:

'The applicant asserts that the motor receives more electrical power back than was input to the motor. The statement directly violates the Laws of Conservation of Energy. The applicant has provided a mathematical derivation as support for this claim. However, the equation is incomplete. It does not reflect any heat or resistance losses generated by the structure of the motor, nor does it show the reputed magnetocaloric cooling effects. Further proof and explanation is required to support this assertion.'

Now, to be strictly correct, I did not affirm in my specification that 'the motor receives more electrical power back than was input to the motor'. I did, however, explain why it was possible to take more electrical energy from an inductance than had been supplied as electrical power input, but there really is nothing clever or magical in that, because the process I referred to involved pulling a magnet away from a soft iron core member and that involves an input of mechanical work. Every dynamo that incorporates a magnet generates electrical power output in that way.

My real concern arises from the other comments by the Examiner because that word 'reputed' implies that the Examiner is unaware of what I mean by 'magnetocaloric effects'.

I intend now to describe what I might term a 'Magnetocaloric Motor' but stress that this motor is hypothetical and impractical as a commercial proposition. It in no way corresponds with the motor disclosure in my patent, but it does serve to educate those who find it difficult to believe the principles of physics that apply to the motor developments which I am pursuing. I stress that the physics I use is standard physics which in no way involves any breach of the law of energy conservation.

I want you to imagine a solid cylinder of a ferromagnetic substance with a solenoidal winding wrapped around it. It is positioned with its axis vertical and it has, underneath it, a strong magnet held to it by its magnetic attraction. The magnet is heavy and, if it were free from that magnetic hold, it would fall under gravity. It could, for example, be part of a reciprocating engine which supplies a drive torque to a flywheel through a connecting rod and crank system.

Now, a ferromagnetic substance has a 'Curie temperature', which is the temperature of transition at which it loses its ferromagnetism and behaves as a normal metal to which that magnet is not attracted.

My scheme is to supply heat as an energy input to my motor and cause that cylinder to lose its ferromagnetic state. What must then happen is that the magnet will fall from the top-dead centre position to the bottom dead-centre position and turn the flywheel and its shaft through 180 degrees. That delivers mechanical work as output. Note that I could, if I wish, take electrical power from that solenoid as the magnet falls. Indeed, the solenoid has no other purpose.

After this half-cycle of my motor, or heat engine if you wish to regard it as such, I allow the cylinder to cool. It recovers its ferromagnetism and the magnet sees that change and rushes to rejoin the cylinder owing to magnetic attraction. This means that the magnet does work in recharging the gravitational potential shed during its early fall. It also can deliver some more electrical output by electromagnetic induction in that solenoid and we recover heat energy owing to the cooling of the cylinder.

Such an engine will work because there is nothing in what I have said that defies either experience or any law in physics. The energy deployment involves heat energy Q input and, as output, four energy components, the heat energy Q' shed by cooling, the mechanical energy developing the motor drive torque, the electrical output on the down stroke and the electrical output on the up stroke.

Now, obviously, logic governed by the law of conservation of energy says that Q must be greater than Q' by the sum of the electrical and mechanical power outputs. Friction is part of the heat output.

I invite you now to consider a real motor in which there is the usual heating attributable to core magnetization losses and heat generated by current flow in the windings. If I were to deploy some of that heat and convert it into electrical or mechanical power in my motor, would that defy the law of conservation of energy? Surely, it would not, but it might enhance the efficiency of my motor and that was a feature I found in my tests on the motor, the subject of the patent application which the Examiner has rejected.

Now, I am going to go further in developing these ideas for a magnetocaloric motor. First, I will suppose that there is 100% balance between energy input and energy output. After all, just as we cannot get energy from nowhere, so we cannot lose energy into that 'nowhere' world. So, the electrical and mechanical power that my hypothetical motor generates draws with 100% efficiency on the heat input, assuming I have good thermal insulation in my motor construction. I need, therefore, to supply heat at a temperature sufficient to exceed the Curie temperature of that ferromagnetic substance from which the cylinder was constructed.

If it were of iron then that would be quite a high temperature 770 degrees C. If it were of cobalt that means 1127 degrees C. Nickel offers better prospect at 358 degrees C, but, as this is a thought experiment, I prefer to use gadolinium which has a Curie temperature of 16 degrees C. That is virtually normal room temperature, so, if I am working in a cold climatic region, I can contemplate building a normal heat pump, which need not be very efficient, to pump heat up to a temperature above 16 degrees whilst allowing ambient cooling to provide the temperature reduction below 16 degrees. If I were in a hot climate then the environment could provide the heating and I would use the heat pump in reverse as a cooling system.

Now, as every professional engineer knows, heat pumps that change heat through a temperature range of, say, 10 degrees in the ambient temperature range can operate with a figure of merit as high as 10 and more. In other words, for every unit of heat energy fed into my motor at a little above 16 degrees C, I can, using heat shed at a somewhat lower temperature as complemented by environmental heat, supply 10 units of excess heat (Q-Q') to my motor.

The fact then is that, in theory, meaning standard physics theory and practical engineering experience, I can expect my motor to operate to deliver up to ten times as much electrical and mechanical power output as I supply as electrical input to operate that heat pump. There is no breach of the law of conservation of energy because the heat energy balance accounts for any difference.

So, I say to the U.S. Patent Examiner responsible for my application, that this is what I mean by magnetocaloric effects as a factor adding efficiency to the motor I disclose in my patent specification. It did not incorporate gadolinium, nor did it run on that very slow heat cycle implicit in the thought experiment described, but it did harness physical principles akin to those described.

Whoever may read this may wonder how they can begin to check what I say about the phenomenon of magnetocaloric effects. To assist the reader in that question, I direct attention to a book authored by a leading expert on the physics of magnetism. He was one of the external examiners chosen by Cambridge University to examine my Ph.D. thesis, which was, in fact, concerned with energy anomalies in ferromagnetic materials. The book 'Modern Magnetism', by Professor L. F. Bates, was published in its fourth edition by Cambridge University Press in 1961.

Consult that book or any other of your own choosing and try to establish whether the heat energy supplied to a ferromagnet to cause it to transit through the Curie temperature is the same as the heat energy returned by cooling through the same temperature range. Try to find any hint that the specific heat of the ferromagnet differs in dependence upon the proximity of that magnet in my hypothetical motor. It would have to if it is to satisfy the requirements of the law of energy conservation, but this is one of those questions that may still await an answer.

Certainly, I think the physics community will adhere to the definition of specific heat as the amount of heat needed to raise the temperature of unit mass through a unit degree, even though they normally measure it by what amounts to techniques involving cooling. That shows their faith in the belief that the specific heat is the same for tests involving increase of temperature as it is for decrease of temperature. Note that in my hypothetical motor I supplied heat Q to heat the ferromagnet and received in return heat Q' upon cooling, the difference being used to run my motor. The latter difference can be calculated in terms of the weight of the magnet, range of stroke, and induction effects etc but it has to be substantial. Yet, according to the standard theory used by physicists, as implicit in those specific heat measurements, Q' should equal Q.

So there is an interesting problem. Certainly it causes me to wonder if there is another source of heat energy, perhaps the world of entropy in our underlying environment, but, I am here expressing the views of an engineer, though using my in-depth knowledge of electromagnetic and ferromagnetic phenomena.

I invite anyone to contact me and enlighten me if they can point to an error in the case I put for the hypothetical magnetocaloric motor I describe above. I am very interested in knowing the truths of physics which are at work in these processes, but, please, do not quote a law of thermodynamics as a reason for non-operability. The laws of thermodynamics are based on past experience and I am interested in future technology which has to be well founded and based on something new. All I can say is that I find in my motor experiments reason to believe that heat is being regenerated as electrical power which assists the output of the motor.

I may add that in my Ph.D. research I confronted the problem of why far more heat was generated by eddy currents in transformer steel than should be the case theoretically. I did not at the time even think that it was because the heat generated could itself cause additional EMFs which strengthened the circulation of those eddy currents in the transformer laminations. Looking back at the experiments reported in my thesis I can now see that the excess losses had to be attributable to regeneration of electrical power from heat in a way which tended to escalate the loss, particularly in thin large grain electrical steel.

In the book by Bates, chapter IX between pages 327 and 371 discusses specific heat measurements in ferromagnetic materials and in particular discusses the experimental facts concerning the magnetocaloric effect. There is a transitional increase in specific heat over the transition range near the Curie temperature. There is a discrepancy between theory and experiment which is said to be of 'unknown origin' (page 330). However, there is no suggestion that specific heat values could differ according as to whether the temperature is increased or decreased. I qualify that by observing that it was mentioned on page 338 that, in weak magnetic fields, the relationship between magnetic polarization and temperature did suggest a time delay in adjusting to the measurement conditions, of which it was said: 'In a way, this is merely another mode of stating that hysteresis effects are playing a small part'. There is no breach of the energy conservation law, that is the first law of thermodynamics, whereas the second law of thermodynamics does not even apply because the magnetocaloric motor is not, in a strict sense, a heat engine responding solely to the thermodynamics of an operating fluid. The magnetic factor can, as is well known, serve as a catalyst in redirecting the thermodynamic energy conveyed by electrons in metals so as to generate EMFs. This is a thermoelectric property known as the Nernst Effect and, as with the Hall Effect, it is not governed by the Carnot limitations which apply to the second law of thermodynamics. In saying this I am mindful that the use of magnets as a catalyst to deflect charged ions from a heated gaseous flow so as to generate electrical output (MHD, the magneto-hydrodynamic technology which had its heyday in the 1960s) does comply with the Carnot efficiency limitations. That is because the gas molecules enter as input at a high temperature and all leave as output through an exhaust at lower temperature. These are the conditions demanded by the second law of thermodynamics, but electron flow in metal according to the Nernst Effect has no output current that is fed into an 'exhaust', because no current flows into the cooler heat sink.