The Design, Construction and Testing of a Thermo-Magnetic Motor

Send to Kindle

By Timothy Raney…Bald Engineer Guy with Glasses

The thermo-magnetic motor converts thermal to kinetic energy by using a ferrous metal’s Curie point. The Curie point is the temperature where a ferromagnetic material loses its intrinsic magnetic properties as a function of temperature. The motor’s operating mechanism depends on this physical property of ferromagnetic substances. This motor essentially functions as a mechanical oscillator and uses nickel metal with its relatively low Curie point. When a heat source raises the nickel armature’s temperature to its Curie point, it becomes nonmagnetic and swings away from the attracting magnet. Once cooled below its Curie point, it swings back to the magnet. I modeled this motor after Nikola Tesla’s 1889 United States Patent for a “Thermo Magnetic Motor”.

This paper documents the design, construction and testing of a thermo-magnetic motor and subsequent experiments. The purpose of this project was to demonstrate a novel means of energy conversion and investigate the Curie temperature of nickel (Ni); a specific property it shares with other ferromagnetic materials. I was interested in making a thermo-magnetic motor since it converts heat energy to kinetic or mechanical energy directly.  Elemental nickel becomes non-magnetic when heated to a sufficiently high temperature or “Curie point”. This point is the temperature where nickel or other ferromagnetic materials lose their intrinsic magnetic properties. In this case, a structure holds a nickel metal armature configured as a pendulum near a permanent magnet. When a heat source raises the armature’s temperature to its Curie point, it becomes nonmagnetic and swings away from the attracting magnet. Once cooled below its Curie point, it swings back to the magnet. The motor operates in self-sustaining mode once heat is applied. I based my design on the thermo-magnetic motor Nikola Tesla patented in 1889[1].

This design uses a pivoted pendulum rod vertically oriented with the ferromagnetic material mounted on its lower end as with Tesla’s patent. I used nickel since it has a relatively low Curie point temperature (358OC)[2]. In principle, the “TM-motor” operates through the magnetic attraction of the nickel and its subsequent demagnetization once heated to its Curie point. As the nickel-tipped pendulum (armature) swings away from the magnet and away from the heat source, it cools.  The magnet attracts the nickel again when the ambient environment cools it below its Curie point. The cycle then repeats itself. Important to this design, a rare earth magnet attracts the nickel, but mechanical stops limits the swing and constrains its motion.

Design and Construction
I essentially adapted modern materials for my version of the TM-motor. The motor consists of a pivoted pendulum-armature and a piece of nickel sheet mounted to the lower end. The pivot constrains armature movement in one plane. This armature is fundamentally identical to a conventional pendulum, except mechanical stops on the motor's frame arrest its swing. Perhaps the most innovative aspect of this design, a rare earth magnet (Nd-Fe-B) supplied the necessary magnetic flux. Gaussmeter measurements showed the magnet’s surface flux was ~4 x 103 gauss or 0.4 tesla. I used a household propane torch for the heat source in all these experiments.

I used machine tools for most of this work to perform conventional metalworking processes, e.g., turning, facing, milling, drilling, reaming and tapping. Materials included free-machining brass (360 alloy), aluminum (6061 alloy), elemental nickel sheet and mild steel (1018 alloy). An experimenter could very well use other materials. I based these selections largely on availability given the different stock shapes involved. The nickel piece dimensions were 2mm thick x 4mm wide and 8mm long. These dimensions were essentially arbitrary, but the resulting low thermal mass decreases cooling cycle time – more about this factor later. The diagram below shows the TM-motor’s general layout and shows one can readily modify the design for hand tools and other materials. This design is just one approach and it is adaptable to a number of other design ideas. For example, one could substitute another ferromagnetic material, e.g., iron (Fe).

However, one must consider a few points before following this approach. For example, iron has a higher Curie point (770oC) – it can take longer to reach the critical temperature (heat source dependent). It could also slow the motor’s oscillation (period) and expose the magnet to temperatures high enough to cause irreversible demagnetization. As a ferromagnetic material, the rare earth magnet alloys exhibit even lower Curie points compared to nickel, ranging from 80 to 300○C. Demagnetization is reversible in other ferromagnetic materials, but temperatures in this range will demagnetize rare earth magnets, depending on the characteristics of the specific alloy.

Equipment and Apparatus
Equipment used for initial experimentation included a propane torch as the heat source and a stop watch to determine the time for heating/cooling cycles. Later experiments included a digital multimeter (DMM) for measuring the output when I added a very crude electric generator consisting of a magnet and solenoid.

After placing the TM-motor on a level surface, I adjusted the left mechanical stop to keep the nickel piece 1-cm from the field magnet. The other stop prevented the pendulum-armature from swinging too far. These adjustments remained constant during the experiments. A laboratory stand held the propane torch with its flame centered and just touching the nickel-tipped armature. Centering the flame allowed the nickel piece to swing away (out of the flame) when it reached its Curie point. The cooling cycle began once the nickel cleared the flame. I also adjusted the torch valve to produce a small flame – a larger flame could potentially demagnetize the rare earth field magnet.

Data and Results
Initial testing determined the TM-motor's heating and cooling cycle characteristics. The cycle is the time required to successively heat and cool the nickel to/below its Curie point. However, the cycle’s speed depends on how fast the nickel reaches and then cools below its Curie point. Ten (10) trials showed an 11 second mean for a full cycle. The heating cycle mean was 2.45 seconds for a given torch valve setting. Higher temperatures reduced the heating cycle time. However, the cooling cycle mean was 8 seconds with passive cooling – ambient air only. I then added a heat sink to reduce the cooling cycle time. The heat sink was a flat copper strip the nickel piece touched when it swung away from the magnet. However, achieving good physical contact between the nickel and heat sink was difficult and only reduced the cooling cycle mean to ~7.3 second.

I also discovered the localized flame from a “pencil tip” torch nozzle does not heat the rare magnet appreciably. I was keeping in mind the rather low temperatures that can irreversible demagnetize rare earth magnets. I did not measure the temperature since a thermocouple type thermometer was not available at the time. As I mentioned above, an increased flame intensity reduced the heating cycle time and the motor ran faster. However, I had not done so before to avoid heating the magnet. These experiments showed the cooling cycle limited the TM-motor’s speed – cooling took ~70% more time than heating. Moreover, one could vary the heating cycle, but not the cooling cycle unless using active cooling, e.g., forced air.

TM-Motor as a Prime Mover for a Generator
In later experiments, I attempted to use the TM-motor as the prime mover for a simple generator. My intent was to demonstrate thermal-mechanical to electrical energy conversion. So, I selected a small iron core electromagnet as the output coil. This coil had 14,700 turns of copper wire (AWG #38). The output potential E (in volts) equals the number of turns treated as a negative number (to account for Lenz's law) multiplied by the magnetic flux difference (Df) divided by the Dt, according to Faraday's law of induction[3]. This relationship simply means more wire [conductor] turns and greater relative movement between them and the flux yields greater output potentials.

The first experiment combined the iron core electromagnet as the output coil and a rare earth disc magnet attached to the pendulum-armature. With the relative motion between the two, the flux would induce a current in the electromagnet. The magnet and electromagnet were spaced 14mm apart. In theory, the generator would produce two pulses per cycle – the cooling cycle governed the pulse repetition rate. By manually operating the TM-motor, the nominal output was 0.08 volts. Reducing the magnet-to-output coil gap increased the flux also raised the output to ~one volt. However, this scheme failed in practice. It only worked when I moved the pendulum-armature manually. The attractive force exerted by the magnet on the output coil’s iron core exceeded the TM-motor’s torque and prevented it the from working at all[4]. What next?

I then tried an air-cored solenoid as the output coil to alleviate this problem. This approach produced a lower output by an order of magnitude since it lacked an iron core as a “flux concentrator”.  Reducing the air gap to 5mm did not appreciably affect this outcome. I had expected this result since the relative magnetic permeability (m)[5] of iron is much greater than air. Importantly though, this experiment did demonstrate thermal-mechanical-electrical energy conversion in principle. I then considered accumulating the output in one farad “super” capacitors. However, the overall experimental results did not warrant further pursuit at that time.

This paper documented the design, construction and testing of a thermo-magnetic motor, along with selected experiments. This prototype thermo-magnetic motor was suitable for demonstrating direct thermal-to-mechanical energy conversion. This effect depended on heating and cooling a ferromagnetic material alternatively in the presence of a strong magnetic field. Testing included characterizing the motor's heating/cooling cycle. The experiments included attempts at adding a simple generator to produce a measurable electrical output. Though these approaches were not entirely successful, this crude system did produce a measurable output – an interesting experiment in its own right. Future work could certainly refine all aspects of this particular design, use different materials and quantify its energy conversion efficiency compared to other methods or heat engines.

Cited References

1. US Patent Number 396,121, Thermo Magnetic Motor, Nikola Tesla, granted 15 January 1889.

2. R.C. Weast (Ed.), Handbook of Chemistry and Physics (64th Ed.), CRC Press, Inc. Boca Raton, FL, 1984, pg. E-107.

3. T. E. Raney, Experimental Physics Laboratory Notebook #8, November 2002 to June 2003, pp. 12-13

4. L.R. Moskowitz, Permanent Magnet Design & Application Handbook, Cahners Books International, Inc., Boston, MA, 1976, pg. 377.

Recommended References

S.W. Angrist, Direct Energy Conversion (2nd Ed.), Allyn & Bacon, Inc., Boston, MA, 1971.

Department of Energy Fundamentals Handbook Electrical Science (vols. I-IV), Basic Electrical Theory - Magnetic Circuits, 1992.

D. Halliday, R. Resnick & J. Walker, Fundamentals of Physics (6th Ed.), John Wiley & Sons, Inc., Hoboken, NJ 2003.

HyperPhysics (, Department of Physics and Astronomy, Georgia State University, Atlanta, GA.

M. Ali Kettani, Direct Energy Conversion, Addison-Wesley Publishing Company Inc., Reading, MA, 1970.

L.S. Lerner, Physics for Scientists & Engineers, Jones & Bartlett Publishers, Inc., 1996.

J.D. Livingston, Driving Force – The Natural Magic of Magnets, Harvard University Press, Cambridge, MA, 1996.

H.F. Meiners (Ed.), Heat, Electricity and Magnetism, Optics, Atomic and Nuclear Physics, vol. II of Physics Demonstration Experiments, The Ronald Press Company, New York, 1970.

F. Miller & D. Schroeer, College Physics (6th Ed.), Harcourt Brace Jovanovich, Inc., New York, 1987.


[1] US Patent Number 396,121, Thermo Magnetic Motor, Nikola Tesla, issued 15 January 1889.

[2] R.C. Weast (Ed.), Handbook of Chemistry and Physics (64th Ed.), CRC Press, Inc. Boca Raton, FL, 1984, pg. E-107.

[3]See Recommended References or high school / college physics texts for more details on Faraday’s law.

[4]T. E. Raney, Experimental Physics Laboratory Notebook #8, November 2002 to June 2003.

[5]Permeability (m) - Ratio of a material's magnetic flux density to the magnetizing force producing it (Moskowitz, 1976).

This entry was posted in Amateur Science, Electricity, Electronics, Experimentation, History of Science, Invention, Machine Shop, Magnetism, Physics, Projects. Bookmark the permalink.

3 Responses to The Design, Construction and Testing of a Thermo-Magnetic Motor

  1. thomas langer says:

    Thomas Langer
    Hey Timothy Raney…Bald Engineer Guy with Glasses.Have always been interested in this concept ever since I read about it during the 1960’s as a high school student.Hope you have success in developing this concept to a potential useable product. I also have some other ideas also on this process, but have never actually performed any real experiments as you have. Hope you do publish more on your work.

  2. John says:

    I plan to develop a rotating motor of higher speed.

    Can you, or anyone, put me onto a forum where I can discuss, seriously, with others?


  3. John Iovine says:

    I built a thermo-magnetic motor just like the one described here years ago. I called it a Curie Effect Heat engine. My company sells them for $24.95 each. If anyone is interested the link is here:

Leave a Reply