Artificial Muscles for a Future Air Conditioner

A prototype device has been developed that is able to transfer heat using ‘muscles’ made from nickel-titanium. Nickel-titanium (aka nitinol) is a shape-memory material that releases heat to its surroundings when it is mechanically loaded in its superelastic state and absorbs heat from its surroundings when it is unloaded.

Nitinol is a ‘smart alloy’ or as ‘muscle wire’. This effect has been exploited by the Saarbrücken researchers who have developed environmentally friendly heating and cooling system that is two to three times more efficient than conventional heating and cooling devices.

They have a patent-pending cam drive whose rotation ensures that bundles of 200 micron-thick nitinol wires are alternately loaded and unloaded in such a way that heat is transferred as efficiently as possible. Air is blown through the fiber bundles in two separate chambers: in one chamber the air is heated and in the other chamber is cooled. The device can be operated either as a heat pump or as a refrigerator.

They used a combination of experimental investigations and numerical modeling they were able to identify how to maximize the efficiency of the underlying mechanism, the wire loading level needed to achieve a specific degree of cooling, the ideal rotational speed and how many nitinol wires need to be included in a bundle. ‘The greater the surface area, the faster the heat transfer, that’s why bundles of wires provide the best cooling capabilities,’ explains Susanne-Marie Kirsch. ‘We use a thermal imaging camera to analyze precisely how the heating and cooling stages proceed.’

They can now adjust and tailor their system to meet different needs. They have a software program to precisely tune the heating and cooling technology for specific applications.

This basic research may well have interesting industrial applications because the novel heating and cooling technology developed in Saarbrücken is highly efficient. Depending on the alloy used, the heating or cooling power of the system is up to thirty times greater than the mechanical power required to load and unload the alloy wire bundles.

The new system is already twice as good as a conventional heat pump and three-times better than a conventional refrigerator. ‘Our new technology is also environmentally friendly and does not harm the climate, as the heat transfer mechanism does not use liquids or vapors. The air in an air-conditioning system can be cooled directly without the need for an intermediate heat exchanger and leak-free, high-pressure piping is not needed.

The team is optimizing heat transfer within the system in order to boost the efficiency of the new technology even more. The goal is to use almost all of the energy from the phase transition for heating or cooling.

18 thoughts on “Artificial Muscles for a Future Air Conditioner”

  1. I know, that Nickel is quite expensive. So I did a cursory check, and I found that Nickel is about $ 13/ kg, and Titanium about $ 60/kg. So maybe I misunderstand something here, but one of us must be wrong. Me – you – Alibaba?

  2. It’s only hot and cold air. Having a bit of leakage and mixing isn’t going to do much harm.

    Just have the duct that the air is supposed to flow through 20 times the (effective) cross-section as the gap the wires pass through, and the leakage only reduces your efficiency by 5%.

  3. Good point, though it’s hard to tell what would be optimum designs in this context. Regular thermal wheels with a thermal medium are pretty slow though. Perhaps not radial wires, but perhaps radial sheets? But the wires would have more turbulent mixing for better heat exchange.

  4. Never fear, the electricity usage will shift towards charging electric cars and industry, so the utilities will get their money. This is one of the things greenies can never understand: you can’t actually “save the environment” by reducing your usage of electricity.

  5. I hate to get all negative. But if people use less electricity, they will just charge more for it. Utilities have to pay shareholders, pensions, executives, lots of workers sitting around or driving around…

    But maybe if you went solar on your own house, you might need less storage batteries.

    And if you used it in your car, your car might be more efficient.

  6. Good point. And to make things worse, the wires would be breaking in stochastic manner. So after 10 million cycles – correct me if I am wrong – half would be broken. But the number of functioning wires (as soon as a wire is broken it is no longer useful for cooling) would be steadily declining during the life of the device. That would mean that the maximum cooling power would be degrading as well.

    So how would you compensate for this? By running the cycles quicker, and thereby shorting the remaining life of the device even farther? Or by having so many wires from the start that a *very* slow cycle gives sufficient cooling power, say, in the range of once per minute? Note, however, that going from one cycle per second to one cycle per minutes necessitates 60 times more mass of nitinol to get the same cooling power…

  7. Hm.. 10 million cycles is not that much. If you use one cycle per second (my guessed cycle), you would get about 3000 hours which equates to about 125 days of continuous operation. Not much.

    Conversely, if we assume 10 years of operation, we need 87600 hours of operation, which in turn gives us a cycle time of about 30 seconds. Does that sound realistic? The “wheel” would be spinning awfully slowly, and you would need a lot of nitinol to produce the thermal effect.

  8. Hm.. So the wires are like spokes on a wheel, where the center is displaced to produce the tension? OK, that would explain how you get the tensioning and relaxing of the wires. It don’t understand, however, how you get the wires from one chamber to the next without air leakage. If you just rotate the “spokes” from one chamber to the next chamber, there needs to be a free space for the wires to go from one chamber to the next. You cannot rotate the “spokes” through a solid wall, right?

    And if this is the case, what is preventing the air from taking the same route? The air could move along the tangent of the circle instead of perpendicular to the circular surface.

    Now, if you didn’t rotate the spokes but only the center to produce the tensioner in the wires and combined this with switching the airflow into the two chambers, I would understand. But this would also produce noise as well as increasing the complexity/cost of the device…

  9. They basically built a rotary heat exchanger, AKA thermal wheel, where the thermal medium is not a pure passive thermal exchange, but a mechanically induced change in the medium. So, pretty simple AC (just need a motor+thermal wheel, then 2 motors+fans for circulation) so simple motors are OK, no refrigerant involved nor a compressor, and a simple bimetallic strip sensor thermostat could manage the thing.

    GoatGuy in another article pointed out that synthetic chloroprene rubber (neoprene) apparently has a similar property as well, which may be suitable for this configuration.

  10. Regular nitinol is on the order of 1 million cycles for fatigue, but Exergyn is claiming an alloy variant that goes to 10 million cycles, which was necessary for the core of their waste heat power generator.

  11. Seems pretty straighforward. You put the wires in tension between a hub and a ring, which rotate together, but have centers that are displaced from each other. So as it rotates the wires are stretched and relaxed at predictable points around the rotation.

    Then you route air through the wire bundles in two places, chosen for maximum efficiency at taking advantage of the wires stretching and relaxing.

  12. What about fatigue failure? Nitinol’s resistance to fatigue is far beyond any other metal’s, but is it good enough in day-after-day/year-after-year thermal transfer?

  13. They say the following:
    “Air is blown through the fiber bundles in two separate chambers: in one chamber the air is heated and in the other chamber is cooled.”
    So that seems to be how it operates. Perhaps this contractile cycle could be seen as a form of peristalsis. There are peristaltic pumps for fluids, so this could be something analogous for heat.

  14. If correct, this could be huge.

    Just want to point out that the technology could be used for heating buildings as well. In Europe alone, 770 000 heat pumps were installed in 2014 [1], so this could be huge is the technology pans out. Increasing the COP number from 4-5 to, say, 10 would make heating much cheaper in many countries. (I don’t really care about “saving electricity” for the sake of the “environment”, but I do care about making peoples lives better and living costs lower).


  15. Interesting. Seems like nitinol is not that expensive. On Alibaba it costs about a USD per kg, so this indicates that this *could* become a viable technology.

    This article, however, could be greatly improved if there were some numbers in it. “Increased efficiency”. OK, what does that mean? A run of the mill – correct me if I am wrong – airconditioner system can remove about 4-5 times the energy from the air as is put into the compressor motors. COP 4-5. What is the corresponding number here?

    Another question. How does it work? We know that nitinol wires are stressed and relaxed, but how are they alternatively put in contact with hot/cold air? Are they placed on a rotating drum, two different chambers…? If so, what is the geometrical arrangement? And why did it take the researchers 5 years from first principle to have a working prototype? What is the technical catch?

Comments are closed.