3 ms·

The absorption refrigerator already existed, but had room for refinement as seen in the patent cited. Einstein's and Szilard's real innovation, maybe not seen

The absorption refrigerator already existed, but had room for refinement as seen in the patent cited. Einstein's and Szilard's real innovation, maybe not seen there, was for a different, electric design with a moving part, sealed inside, that was moved back and forth by a varying magnetic field applied from outside. It is often described incorrectly as having no moving parts. (I used to think ammonia was itself diamagnetic or ferromagnetic, which would have been very cool, until I got a look at their design.) What was innovative was that it had no need for sealed bearings that might leak. Nowadays just the wires for the motor pierce the fluid chamber, less elegantly than a strictly magnetic coupling. Absorption refrigerators are strictly better than the compressor designs used everywhere today, running continuously, silently, and efficiently. The main driver of the latter was that General Electric wanted people to use electric refrigerators, not gas-fired, and wanted them to buy electric motors. Mass production soon gave GE an insuperable economic advantage except in niche applications. Modern institutional refrigeration systems often use ammonia as their working fluid, but nobody likes putting ammonia systems in houses or cars. The modern alternative to freon/"cfc" and "hfc" is liquid CO2 at very high pressure. Hfc that leaks traps heat more than 2000x as well as CO2, so there is a push on to replace it.

a_gnostic 3y ago
Such a linear pump was invented by a Japanese guy, and is commonly used in ENGEL brand RV freezers.
cyberax 3y ago
> Absorption refrigerators are strictly better than the compressor designs That's not true! CoP of ammonia absorption refrigerators struggles to get past 1. You can easily get a CoP of 4-5 with a compression refrigerator. Ammonia absorption fridges are useful when you have a source of waste heat, then it's basically "free" for you.
- angiosperm 3y ago
  I am corrected. Compressor refrigeration benefits from decades of refinement. Just a few decades ago, 3 was considered good.
  - londons_explore 3y ago
    CoP isn't really about engineering refinement to eek out the best efficiency. In fact, we have known how to get really close to the Carnot limit for nearly 100 years. The Carnot limit for a fridge in room temperature is a CoP of ~20. However, getting close to that limit requires huge evaporators and condensers, big pipes, large amounts of working fluid, and multi stage pumps. All of that costs money, and historically people didn't care much about efficiency if it means spending more money.
    - cyberax 3y ago
      The Carnot limit for 26C -> 5C rerigeration is around: 1 / (T_high / T_low - 1) = 1/(300/278-1) = 13. Industrial refrigerators can get to CoP of around 4 for this temperature delta. The overall system performance will typically be much lower, due to the need to run fans and other systems. A couple of interesting technologies to watch in this area: 1. Acoustic refrigerators that use acoustic waves in helium. They theoretically can approach the Carnot limit. 2. Supercritical CO2 cycles. They can be used for larger deltas without losing efficiency.
      - eikoongaepaikei 3y ago
        It's worth noting that the Carnot limit really only applies to systems that return to thermal equilibrium much faster than the characteristic interaction and/or rely on thermal interactions to transmit energy. Electrochemical, magnetohydrodynamic and various other systems are not properly characterized by the Carnot limit and can approach much higher efficiency limits. I think the systems being discussed are appropriately characterized by the Carnot limit, but it is frequently mischaracterized as an absolute limit on systems that it is not relevant to and for example here it should probably not be considered the absolute limit of refrigeration efficiency.
        cyberax 3y ago
        > Electrochemical, magnetohydrodynamic and various other systems are not properly characterized by the Carnot limit and can approach much higher efficiency limits. That usually is a distinction without difference. Other effects limit the efficiency at far lower levels than pure Carnot. This can be intuitively understood like this: Carnot limits apply to gases, that are the simplest interacting systems. You can realistically model them as just a collection of individual independent particles interacting only via simple collisions. Anything more complicated like electrons in semiconductors, and you have way more interactions and way more possibilities for your system to have inefficiencies. Take, for example, solar panels. Sunlight has the optical temperature of around 6000K, so a Carnot engine that uses 6000K hot part and 300K cold part will theoretically have a 95% efficiency. And an infinite stack of solar panels with each layer tuned for a specific wavelength has the theoretically maximum 87% efficiency.
userbinator 3y ago
It's worth noting that CO2 was also used in the early days of refrigeration alongside ammonia, when it was called "carbonic acid", and the relative nontoxicity of CO2 was a feature that made it popular for shipboard systems and a few other applications, but it fell out of use once other nontoxic refrigerants were invented because of the extremely high pressures involved that required far heavier equipment. As an example, the Titanic had CO2 refrigeration.

a_gnostic 3y ago

Such a linear pump was invented by a Japanese guy, and is commonly used in ENGEL brand RV freezers.

cyberax 3y ago

> Absorption refrigerators are strictly better than the compressor designs That's not true! CoP of ammonia absorption refrigerators struggles to get past 1. You can easily get a CoP of 4-5 with a compression refrigerator. Ammonia absorption fridges are useful when you have a source of waste heat, then it's basically "free" for you.

angiosperm 3y ago
I am corrected. Compressor refrigeration benefits from decades of refinement. Just a few decades ago, 3 was considered good.
- londons_explore 3y ago
  CoP isn't really about engineering refinement to eek out the best efficiency. In fact, we have known how to get really close to the Carnot limit for nearly 100 years. The Carnot limit for a fridge in room temperature is a CoP of ~20. However, getting close to that limit requires huge evaporators and condensers, big pipes, large amounts of working fluid, and multi stage pumps. All of that costs money, and historically people didn't care much about efficiency if it means spending more money.
  - cyberax 3y ago
    The Carnot limit for 26C -> 5C rerigeration is around: 1 / (T_high / T_low - 1) = 1/(300/278-1) = 13. Industrial refrigerators can get to CoP of around 4 for this temperature delta. The overall system performance will typically be much lower, due to the need to run fans and other systems. A couple of interesting technologies to watch in this area: 1. Acoustic refrigerators that use acoustic waves in helium. They theoretically can approach the Carnot limit. 2. Supercritical CO2 cycles. They can be used for larger deltas without losing efficiency.
    - eikoongaepaikei 3y ago
      It's worth noting that the Carnot limit really only applies to systems that return to thermal equilibrium much faster than the characteristic interaction and/or rely on thermal interactions to transmit energy. Electrochemical, magnetohydrodynamic and various other systems are not properly characterized by the Carnot limit and can approach much higher efficiency limits. I think the systems being discussed are appropriately characterized by the Carnot limit, but it is frequently mischaracterized as an absolute limit on systems that it is not relevant to and for example here it should probably not be considered the absolute limit of refrigeration efficiency.
      - cyberax 3y ago
        > Electrochemical, magnetohydrodynamic and various other systems are not properly characterized by the Carnot limit and can approach much higher efficiency limits. That usually is a distinction without difference. Other effects limit the efficiency at far lower levels than pure Carnot. This can be intuitively understood like this: Carnot limits apply to gases, that are the simplest interacting systems. You can realistically model them as just a collection of individual independent particles interacting only via simple collisions. Anything more complicated like electrons in semiconductors, and you have way more interactions and way more possibilities for your system to have inefficiencies. Take, for example, solar panels. Sunlight has the optical temperature of around 6000K, so a Carnot engine that uses 6000K hot part and 300K cold part will theoretically have a 95% efficiency. And an infinite stack of solar panels with each layer tuned for a specific wavelength has the theoretically maximum 87% efficiency.

userbinator 3y ago

It's worth noting that CO2 was also used in the early days of refrigeration alongside ammonia, when it was called "carbonic acid", and the relative nontoxicity of CO2 was a feature that made it popular for shipboard systems and a few other applications, but it fell out of use once other nontoxic refrigerants were invented because of the extremely high pressures involved that required far heavier equipment. As an example, the Titanic had CO2 refrigeration.