Heatpipes work better facing upwards?

[Mercenary Keyboard Warrior]

Just been thinking about heatpipes and the technology.

I didn't really understand that well how they worked so did a bit of basic research and found that they work as I thought. A small amount of a liquid is stored within a metal tube.
The heat of the processor turns this liquid to vapour which then travels along the tube to the main part of the sink where due to the airflow it cools and condenses.
Ok pretty simple stuff.

But the basics just strike me that mavity is surely going to help this situation dramatically no matter what clever strategy has been applied to getting the vapour to travel along the tube.

One scarey bit I found. If you generate an amount of heat that exceeds the amount the pipes can handle they basically stop working as all the liquid can be vapour cooling down. That would mean you would be relying on the passive cooling effect of the big sink rather than the pipes themselves.

So have I got it wrong, or should I be turning my tower onto its side during warmer weather?

 

[PermaBanned]

Pipes aren't used for cooling, only the transportation of heat. They work better upright because hot air rises, and the vapour will rise while the liquid stays at the bottom. You don't want the same amount of vapour and liquid everywhere, otherwise it wouldn't be transporting the heat to the surrounding heatsink, and through that and the fan, to the air.

 

[JonJ678]

Not sure about this. I believe that good heat pipes work at any orientation, but crap ones tend not to. There's some confusion over which are good and which aren't.

There is a wick in the centre of the pipe, which the vapour (and possibly present liquid) are drawn along. I don't think its the change from liquid to gas which makes a difference, as in use it should almost all be gas at low pressure. However when cut, it condenses and water runs out hence the belief that liquid + gas is present.

I'm pretty certain it works independent of orientation, because its full of gas which mavity doesn't have much effect on. One end gets hot, which lets the gas move along the wick more effectively, so you have a net movement from hot to cold. Once it cools down, it's rejected from the wick and moves back normally. Something vaguely like this anyway, I don't have a link on hand and haven't researched them for a year or so.

They do have a saturation temperature above which they don't work so well, and a minimum below which they don't work at all. However the ones in use in a computer will have saturation temperatures over thermal shutdown of the processor and all should be well.

Hopefully someone can do a better job of it than this, I'm quite hazy on the details

 

[shadowscotland]

the 'wick' is on the edge of the pipes interial surface (not like a candle wick)

And yes mavity helps (maybe a 5-10% effeciency gain for just the pipe) - but capillary pressure is a very strong force so they can work in any orintation

wiki linky

 

[Mercenary Keyboard Warrior]

Hmm I saw that diagram and again it shows something that could only be the same or benefit from mavity.

I guess the simplest way to test is to run my PC at 100% load for an hour or two until the temp pretty much stabilises. Then turn it onto its side adn see if it drops with the heat pipes facing upwards.

Its all a little vague to get exact details and of course manufacturers don't like to quote specs.

What I am surprised is that no one seems to test heatsinks with both orientations.
I bet most enthusiasts have a tower case, yet fundametally that could decrease the efficiency of the cooling. If it was 5-10% thats not a small margin.

 

[shadowscotland]

I don't for a minute surgest it's will affect cpu temps 5-10% as there are many other factors.
And I've seen a few reviews that test orintation (but it's mostly north-south vs East-west) see xbitlabs review here

Just like Prolimatech Megahalems, both Ultra-120 eXtreme coolers proved more efficient when they were installed with the heatpipes going across the processor heat-spreader.

AKA hot air rised

 

[JonJ678]

Thanks shadow, I clearly wasn't thinking. Of course the wick is on the inner surface. Should've remembered the term capillary action from all the physiology I did a few years back too.

Good image