Are heatpipes really that good?

[watdahel]

Wouldn't it be better to have a solid copper tube to conduct heat rather than a hollow tube?

 

[BigCoolJesus]

Originally posted by: erwin1978
Wouldn't it be better to have a solid copper tube to conduct heat rather than a hollow tube?

No, heatpipes allow better dissipation of heat, and cool better, since they are not "hollow" per se, they have liquid in them that transfers the heat, and then when it condenses it helps cool also

A solid copper pipe wouldnt really transfer heat, it would just collect it, and thus cooling area would be diminished (heat wouldnt make it to the fins)

 

[imported_adam1234]

I'm pretty sure the fluid in the heatpipe is able to transfer the heat more quickly than a solid copper tube would. I think that the phase change from liquid to vapor that occurs in a heat pipe is able quickly absorb the heat energy, and then transfer it to the other end of the pipe. It depends on the liquid/gas inside but if it were water it takes something like 2000 kJ/kg which basically means that a lot of energy can be absorbed by the liquid when it is changed to a gas

 

[BigCoolJesus]

Originally posted by: adam1234
I'm pretty sure the fluid in the heatpipe is able to transfer the heat more quickly than a solid copper tube would. I think that the phase change from liquid to vapor that occurs in a heat pipe is able quickly absorb the heat energy, and then transfer it to the other end of the pipe. It depends on the liquid/gas inside but if it were water it takes something like 2000 kJ/kg which basically means that a lot of energy can be absorbed by the liquid when it is changed to a gas

:thumbsup:

he said it better then me

 

[Cerb]

Heatpipes are a kind of passive phase-change cooling. The heating and cooling process cause the change in state. Much like a fridge-type system, it's not so important that you have good thermal transfer on the end as much as it is that the heat is actually being moved away from the source of the heat.

Normal:
CPU->heatsink base->heatsink fin edges, cooled to keep decent delta-T.

HP:
CPU->heatsink base->lower heatpipe->gas rises to upper heatpipe part->fins->gas cools to liquid, falls to lower heatpipe part->repeat from heatsink base again.

Newer heatpipes also seem to have some sort of wick-like stuff in them, so that with the heat changes, capilary action can do a good bit of the work of moving the liquid, so getting the angle vs. gravity right will only mean a few degrees, not a burnt chip (IIRC, a few, I think Thermalright's among them, can actually work upside-down, albeit with a bit less cooling power).

 

[Ike0069]

I believe you guys are incorrect here.

First, heat always goes from hot to cold. Now transferring heat from one object to another is as simple as passing the molecular vibration from one object to another. Conductive heat transfer: Simply place the two objects in contact with each other, and the molecular vibrations from one object will cause the molecules in the other object to begin vibrating.

The thermal conductivity of the objects involved plays a big part in how much heat is transferred. As a general rule, solids have the highest heat conductivity. Liquids have less conductivity. Why? Because in most liquids the molecules are farther apart than in solids. Since the molecules are more spread out, vibration in one of them isn't as likely to spread to nearby molecules. Gasses have the poorest thermal conductivity because their molecules are even more spread out.
(Stolen from Heat transfer book.)

So a solid copper heat pipe should transfer the heat fast than one filled with liquid.
I believe the reason for filling it with liquid is for cost and weight savings.

Now, after having said all that, I could be screwed up here. I'm not a mechanical engineer, I'm a chemist. So heat transfer is not my forte.

 

[Varun]

Originally posted by: Ike0069
I believe you guys are incorrect here.

First, heat always goes from hot to cold. Now transferring heat from one object to another is as simple as passing the molecular vibration from one object to another. Conductive heat transfer: Simply place the two objects in contact with each other, and the molecular vibrations from one object will cause the molecules in the other object to begin vibrating.

The thermal conductivity of the objects involved plays a big part in how much heat is transferred. As a general rule, solids have the highest heat conductivity. Liquids have less conductivity. Why? Because in most liquids the molecules are farther apart than in solids. Since the molecules are more spread out, vibration in one of them isn't as likely to spread to nearby molecules. Gasses have the poorest thermal conductivity because their molecules are even more spread out.
(Stolen from Heat transfer book.)

So a solid copper heat pipe should transfer the heat fast than one filled with liquid.
I believe the reason for filling it with liquid is for cost and weight savings.

Now, after having said all that, I could be screwed up here. I'm not a mechanical engineer, I'm a chemist. So heat transfer is not my forte.

Sorry Ike

You don't understand the way a heatpipe works. This is not transfering heat via conduction, it is phase change. The heat changes the liquid to a gas, the gas carries the excess energy away, gets converted back into a liquid by conduction, and the process starts again. The key is that the heat is being taken away from the hot source, and is able to be dissipated over a larger surface area.

 

[Cerb]

http://www.sti.nasa.gov/tto/spinoff1996/64.html
http://www.process-cooling.com/CDA/Arti...__Features__Item/0,3674,151243,00.html
http://eu.shuttle.com/en/desktopdefault.aspx/tabid-388/633_read-10644/
http://www.thermacore.com/hpt_faqs.htm
(didn't know about the NASA part before)

Also, based on the NASA link, it appears that heatpipes not working upside-down are somewhat abberant designs. Maybe cheaper? I know from rocketting GPU temps that my ZM80C's pipes don't work with even a few degrees of down tilt!