Hacker News Reader

Comment by klyrs | Hacker News Reader

This sounds pretty neat, but their description of cold things sucking thermal radiation doesn't jive with my freshman-physics understanding of thermodynamics. AFAICT they've got cold water circulating in a moisture-repellent membrane, so they're able to absorb heat by chilling the water in the pipes instead of fussing with the humidity in the air like most AC does. So it sounds like trading inefficiency of standard AC units for an inefficient thermal interface -- you need to stand next to it to feel cooled?

View on HN

It's the cooling equivalent of a radiant heater, and you need line-of-sight to the walls to feel cooled. The thermal interface is deliberately inefficient for convection (to avoid wasting power chilling the air), but efficient for radiation.

Radiation is proportional to absolute temperature T^4. So compared to a glowing radiant patio heater, I'd expect to need maybe (1000^4 - 300^4)/(300^4 - 270^4) ~ 350x the area for the same heat transfer in the opposite direction, thus their big tunnel instead of a little filament. Even with their cold source at absolute zero, they'd still need >100x.

And you still have to pump the heat out to somewhere else.

The whole premise with this system is that you can keep the temperature higher so you have to pump less heat, which I'm extremely skeptical of, at least for indoor environments.

The biggest gain is that you don’t have to pump out all of the heat of condensation of all of the moisture in the air. Condensing out a pound of water is 200x more energy-consuming than lowering the temperature of a kg of air by 10 degrees C.

Thanks for explaining. I am also surprised by this possibility given my limited physics knowledge.

Do clothes substantially reduce the effectiveness? Is true line of sight required - visible or near visible spectrum?

Secondly, can you expand a bit about the humidity/condensation aspect? I get the impression that condensation represents inefficiency, and that this somehow avoids having to cool air as a middle layer to cooling a person.

A cubic meter of air around room temperature and pressure has mass around 1000 g, and isobaric specific heat capacity around 1 J/(g*K). So cooling that air from 30 C to 25 C requires us to suck out 5 kJ. But at 30 C and 100% RH, air holds roughly 30 g/kg of water, while at 25 C it can hold only 20 g/kg. So as we cool the air, 10 g of water will condense, releasing about 20 kJ that we also have to suck out. There's no way to cool the air without condensing the water. In my example, the "latent" heat from the water condensing was much bigger than the "sensible" heat from air cooling. In practice it seems the latent heat is typically around 25% of the total[1], since we're probably at <100% RH, and since heat probably leaks in faster than moisture does (since air leaks let in both heat and moisture, but conduction through the walls/windows lets in only heat). The latent heat would be a bigger contributor in the poorly-sealed or outdoor cooling applications that these researchers seem to be targeting.

The idea of a radiant cooling tunnel appears to be old, but not very popular because it's not very effective. The cold plates leak cold into the air, just becoming a more cumbersome version of conventional air conditioning. The cold plate also can't run colder than the dew point, since the condensed water would drip and make a mess. The innovation here seems to be that they've placed a thermally non-conductive but transparent (to the thermal radiation) membrane between the cold plate and the user. There's a thin layer of cold, dry air between the cold plates and the membrane, but the membrane stops the cold from leaking out into the room air.

It's analogous to the sun shining through a well-insulated window on a winter day--you still feel the radiative heating from the sun, even as the window keeps the warm and cool air separated. Normal window glass wouldn't work for the cooling case, since the heat source isn't the sun (~6000 K) but the human (~300 K), so the wavelengths are much longer, around 10 um, and window glass is opaque there. Plastics with transmission around there are known, though, like the ones used to make the Fresnel lenses for PIR motion detectors. In any case, that membrane lets them keep the cold plate colder than the dew point without condensing water onto it, and also decreases loss of cooling by convection into the air.

It would work best on exposed skin. It should still work with clothes, as long as the clothes aren't too thermally insulating.

1. https://www.energy.gov/sites/prod/files/2013/12/f5/issue7_se...

It might be better to think of it as creating an environment where you are not surrounded by warm objects radiating at you. So it should not matter where you are in the space or how large it is. The effect should be the same.

About 60% of human body heat is lost through radiation. The cooling you would feel would be proportional to the percentage of the area around you that is covered by these cooling walls. Distance wouldn't cause that much loss because the air doesn't absorb much radiation in blackbody range of humans over distances less than 50 meters.(Infrared range with intensity peaking at 93 thousand angstroms)

But when I lose that heat through radiation it would presumably bounce around the room and not much reflects back onto my body.

So I’m not seeing how much this really helps. My radiated heat shouldn’t be returning to me anyway.

Also not sure how effective this system really is, but it helped me to imagine my walls reflecting visible light as if I were in a room of mirrors and thinking about how many higher order reflections of myself would be able to reach just my pupil, multiplied by the surface area of my body.

You lose heat through radiation, but the materials surrounding you are also radiating heat.

You emit thermal radiation at a black-body temperature of 37C or so to the environment, but the average surface nearby (in a non-air-conditioned environment) will be radiating back at 30C or so, for very little net heat loss.

The innovation described here is to replace the local environment with a chilled one that interacts only by radiation. The air with very little mass contributes negligible radiation to the system, so you radiate heat outwards but receive little in return.

Gaining or losing heat via radiation isn't as easy as conduction, but it does work.

It's possible to make ice at night in temperate climates by taking the equivalent of a solar concentrator, putting water at the focus and pointing it at a dark patch of clear sky.

You know, probably standing in the sun vs shade is probably the opposite analogy. In the sun you get radiative heating from the sun, in the shade you just get heating from the nearby air.

> but their description of cold things sucking thermal radiation doesn't jive with my freshman-physics understanding of thermodynamics.

Pseudoscience alarms started to sound in my head, until I read the comments.

If I understood this correctly - you will still radiate heat as normal. But, since the thermal radiation you emit will reach these things (assuming line of sight), you won't have much reflected back at you (or at nearby objects, which would heat them up). And so, you feel colder, even though the air is not being chilled.

That would make sense, similar to how we can make ice during the night even in a desert. You can radiate heat into space(as not all thermal radiation will be absorbed by the air)

https://tecped.com/process-of-freezing-ice-in-the-desert-acc...

From a freshman physics point of view, “sucking” is also not a pull-from-low process but a push-from-high process. There is some nuance there but simply, a straw doesn’t work right in a vacuum.

> fussing with the humidity in the air like most AC does

Humidity will have to be dealt with no matter what, either by the AC system, or though condensation, as cold air holds less water vapour as warm air.

Reply on news.ycombinator.com