Revision as of 13:44, 30 June 2024

Physical mechanics of cooling

Passive cooling

Passive cooling tends to mean 'what happens with no moving parts'.

...so whatever amount of conduction, radiation, and/or convection would happen anyway.

Sometimes includes adding a fan, to add to the convection.

You're stirring the air better than just convection would, so heat transfer goes a faster than if warm air just sits around - but the difference is rarely much -- convection always does this at least a little when there is temperature difference (if you're in gravity; this is about density differences).

And you could argue that's technically active cooling (because you're adding work, so using energy), but intuitively it feels like it hardly qualifies.

On the technical side

This tends to mean

conduction - a good conductor spreading heat throughout

if any cooling happens, conduction's spreading brings the whole down

radiation - thermal radiation means movement of charges in materials (anything above 0 K) is radiated as EM at the surface

(black-body radiation can be seen as a "thermal radiation's real-world math becomes easier if we make some assumptions like that it's not really interacting in other ways")

convection - fluid flow, in this context often

air,

flow caused by heat changing temperatures and densities

that flow assisting better heat interchange with that fluid, because warmer air moving up tends to draws in colder air from the sides (which technically is an effect that needs gravity)

In practice there's more than one of these happening, but often one that counts for most exchange.

Evaporative cooling

Refrigeration cycle

Refrigerants

Heat pumps

Free cooling

Dessicant based systems

Peltier effect and related effects

Theory

Peltier device behaviour and uses

Transport

On efficiency

Using the effects - practice and products

When passive cooling works

Windows and windcatchers

Convection in houses

Device cooling

Passive device cooling

Heatsinks

Heat pipes and vapor chambers

Personal cooling

Fans / Why wind helps even when you're not sweating

Spray bottle, misting system, etc.

Some small tricks

Wrapped towel

Ice

Swamp coolers

Mini dehumidifier

"Personal ACs"

Air conditioning

One-hose and two-hose portable air conditioners

Car AC

Is a flipped AC a heater?

"Eco coolers"

Passive effects

Windows

Parts of designs

Heat exchanger

Unsorted

Cold and hot drinks

Ice packs

BTU

COP, EER

Adding some more considerations: SCOP, EER2, SEER, SEER2, CEER

Comparison?

More on fridges and freezers

Super!

On frost

Does a lot of ice make a fridge/freezer less efficient?

My freezer causes freezer burn on food, what do I do?

In defrosting a fridge, where does the water go?

@@ Line 787: / Line 787: @@
 When current is passed through a thermocouple, it sets up a temperature difference between the two sides - roughly the Peltier effect (see above for more context).
-Thermoelectric cooling (TEC) elements, a.k.a. Peltier elements, are many such junctions in series, electrically in series, and physically beside each other, for practical reasons.
+Thermoelectric cooling (TEC) elements, a.k.a. Peltier elements,
+consist of many such junctions electrically wired up in series, and physically beside each other.
-Putting a bunch of voltage will attempt to sustain a specific temperature difference, proportional to the current passed (and it's a mostly [[simple load]]).
+Putting a bunch of voltage will attempt to sustain a specific temperature difference,
+proportional to the current passed (and it's a mostly A [[simple load]]).
-Perhaps the most common use is to try to keep the warm side at room temperature (by adding a chonky heatsink and fan), so that the cold side will be some amount of degrees colder than that. {{comment|...mostly because doing it the other way around, trying to keep the cold side at room temperature, would amount to a an unnecessarily complex heater)}}
+Perhaps the most common use is to try to keep the warm side at room temperature (by adding a chonky heatsink and fan), so that the cold side will be some amount of degrees colder than that.
+...mostly because doing it the other way around, trying to keep the cold side at room temperature, would amount to a... heater.
+An unnecessarily complex one.
-The main thing TECs have on them is size and being solid-state, while still being a heat pump,
+TECs are solid-state, while still being a heat pump,
 which means they are more portable and/or safer than e.g. gas based fridges, or refrigeration cycle fridges.
-It's good enough for beer coolers, mini-fridges, because once those are cold or if they start with cold contents, it's mostly just fighting the little heat that comes through the isolation - and not trying to sustain a 30C difference as fridges may have to.
+They are barely good enough for beer coolers, mini-fridges,
+because once those are cold or if they start with cold contents,
+then it's mostly just fighting the little heat that comes through the isolation.
+But it will not manage to sustain a 30C difference, as fridges may have to.
@@ Line 806: / Line 815: @@
 so there are various more efficient cooling systems both in terms of efficiency and capacity.
-In fact, while the amount of heat moved is proportional to I, the resistive heating is proportional to I<sup>2</sup>, which is part of two issues.
+In fact, while the amount of heat moved is proportional to I (the current),
+the resistive heating is proportional to I<sup>2</sup>, which is part of two issues.
-One, the ability to move heat relies on the ability to move off heat, which at higher current is mostly that resistive heat.
+For one, the higher current you feed in,
+the more of that heat you need to move away comes from the element itself,
+rather than the thing you are trying to cool.
-This is a good part of why the most energy-efficient is somewhere between a tenth (for small temperature differences) and under half (for larger differences) of the max current rating, further depending on other wishes and choices.
-: It lowers at large currents because you probably can't move the heat off fast enough (and it can't exit the element fast enough), so the net effect amounts to more (resistive) ''heating'' than necessary.
+TECs are probably most energy-efficient at somewhere between between
-:: To the point that too small a heatsink can mean [[thermal runaway]] - basically, it needs to be able to sink the heat generated by the resistive part of the whole. You have a sli
+: a tenth (for small temperature differences)
+: and under half (for larger differences)
+... of the max current rating, further depending on other wishes and choices.
+If you're trying to heat something it becomes more of basic a resistive heater.
+If you're trying to cool something, your real problem becomes moving away heat quickly.
+: ...to the point that too small a heatsink on the warm side can lead to [[thermal runaway]]
-: It lowers at tiny currents because the movement of heat is smaller than the heating effect.{{verify}}
@@ Line 822: / Line 839: @@
+In fact, it is easily worth it to run two, each at half current, because in almost any case - the effect will be a bit more than double the heat moved, meaning higher efficiency (though slightly slower movement of said heat).
-In fact, it is easily worth it to run two, each at half current, because in almost any case - the effect will be a bit more than double the heat moved, meaning higher efficiency (though slightly slower movement of said heat).
@@ Line 845: / Line 862: @@
 For 15 degrees it may peak around a CoP of 2, at ~4V
-(this actually depends a lot on models - these things have gotten better over time)
+(this actually depends a lot on models - these things have gotten a little better over time)
@@ Line 851: / Line 868: @@
 : an AC may want a large temperature difference at reasonable CoP.
 :: For CoPs that ACs can do, like 3 or 4 (up to maybe 6, with refrigerants that are being phased because they're nasty{{verify}})
-:: TECs can only do a few degrees. And that's direct surface temperature, not exchanged into the air
+:: TECs can only do a few degrees if you want good CoP. And that's direct surface temperature, not exchanged into the air
-A single TEC running at a high CoP implies running it a low power, so you would need dozens to even approach the overall amount of heat transfer of even the smallest ACs. (And you don't have tiny ones both because a tiny refrigeration cycle is harder to make efficient, and because you probably need a few hundred Watts of power - a human puts out ~100W when idle, and so does various equipment, and solar influx may do a bunch more even with nicer windows)
-And the heatsinking.
-And good heat exchange with the air (not so easy for low temperature differences).
-And the separation between inside and outside of the room you cool (it's not as easy to split, like a split AC), which basically implies a lot of wall space.
+And an idle human puts out maybe 80 watts, so even keeping up with that, a bit of solar influx, and a laptop,
+will take
-So you can do it, you may ''even'' do it at ''vaguely'' comparable CoP, but it's a pain to design. In practice, you can be very happy with a CoP of 2.
+If you want TECs running at high CoP, (more at low power) you probably need at least dozens before you even vaguely approach the capacity and efficiency of ACs.
+So you can do it, you may ''even'' do it at ''vaguely'' comparable CoP, but it's a pain to design.
+In practice, you can be very happy with a CoP of 2.
 And no, you don't want to stack them.
-Yes, that makes for higher temperature differences
+Yes, that makes for higher temperature differences,
+but consider that each one has to forcibly transfer move the heat the previous one(s) puts out - they're being forced to do the same work ''again''. (a more technically correct view is a little more involved)
-can't stack them - intuitively because each one has to forcibly transfer move the heat the previous one(s) puts out - they're being forced to do the same work ''again''. (a more technically correct view is a little more involved)
@@ Line 879: / Line 893: @@
 * TEC1-12704 is has 127 couples, 4 Amp
 * TEC1-03112 is has 31 couples, 12 Amp
-{{comment|(If there's a T''number'' at the end, it' made for a higher-than-usual temperature difference (like 100, 150, or 200 ºC, rather than the usual 60ish ºC, but you probably don't want these)}}
+{{comment|(If there's a T''number'' at the end, it' made for a higher-than-usual temperature difference (like 100, 150, or 200 C, rather than the usual 60ish C, but you probably don't want these)}}
-A lot will function between approx 2V and 18V (see datasheet). Some shops sell them as 12V, but most are pretty inefficient by that point. Look at the curves for the specific model.
+A lot will function between approx 2V and 18V (see datasheet).
+Some shops sell them as 12V, but most are pretty inefficient by that point.
+Look at the curves for the specific model.
 The maximum refrigeration power, in Watts, is generally at maximum power, and again, you probably don't want to use them anywhere near that if you care about efficienct use of electricity.

Cooling things: Difference between revisions