New cooling idea?

[Strangelove]

Ok this is an idea that just popped in to my head while reading something really boring and my mind started drifting (anyone who’s studied knows the feeling of reading a whole paragraph and after reading it not remembering any part of it ;) )

Sorry if this got a bit long I guess I got carried away.

First the problems of the current cooling solutions (don’t start screaming, this is the way I see it.)

Water-cooling: First of all you have to pump the water around the system (pumps make noise [well many of them do]). There is always the chance of a leak shorting out your system. You need ekstra space in your case (if you want to keep it all nice and portable) for the pump and radiator. If you use a bong it has to be refilled.

Compressor systems: Expensive. Takes up a lot of space in your case.

Air: Well, it just doesn’t cool enough.

Well here is the idea:

As we all remember from our physical chemistry classes when a fluid boils it remains at the same temp no matter how much you turn up the juice. So if we could find a liquid that boils at say 35 C. Then create a closed system with a reservoir, a condenser and a cobber block to heat the liquid. Basically the fluid would enter the block, which would get heated by the CPU, the heat would boil away the fluid (which gets replaced from the reservoir) the vapour rises in to the condenser, condenses and returns to the reservoir. That way we would have a closed system with no pumps or fans that would keep or precious CPU at 35 C.

Naturally there are a few BUTs:
1. Most fluids that boil around 35 C are flammable if not explosive
2. Many of the fluids are toxic or carcinogenic.
3. It might be difficult to create a good enough condenser.
But hey who said it would be easy??

Once again sorry about the length.

Let me know what you think!

I’ll see if I can make a drawing of the system I’m talking about.

[Tiger]

It's an excellent idea, but unfortunately or fortunately its already been invented and its called a Heat Pipe and is being used in some laptops. There are some excellent designs out there one notably one using a scintered copper base (like a copper sponge). The problem is that they are very expensive to fabricate because most incorporate low pressure systems in them ( a vacuum) in order to get the fluid to boil at those temps and for the kind of thing we are talking about they generally have to be quite large.
The real solution to effective cooling lies in the ability to produce small compressors that work better. They will probably come out with a new multistage type very shortly from what I have read. The problem they have is exactly the same - How do you cool it. So we have come full circle.

[Ridenow]

One little problem. In order to condense the vapor you have to cool it. What would happen is all of your fluid would change to vapor. Then either the vapor would not cool the processor or the vapor would continue to heat until meltdown.
With the heat pipes you still have to cool one end of the pipe to get the fluid to cycle.

[Strangelove]

It's an excellent idea, but unfortunately or fortunately its already been invented and its called a Heat Pipe and is being used in some laptops. There are some excellent designs out there one notably one using a scintered copper base (like a copper sponge). The problem is that they are very expensive to fabricate because most incorporate low pressure systems in them ( a vacuum) in order to get the fluid to boil at those temps and for the kind of thing we are talking about they generally have to be quite large.

Cool!! :D Now I just have to figure out how to make one my self ;) I guess I should dig out my books and see if I can calculate how big the thing would need to be. You might be right that it would need to be too big, but I was expecting a lot more criticism than I have gotten yet so I'm still optimistic ;)

One little problem. In order to condense the vapor you have to cool it. What would happen is all of your fluid would change to vapor. Then either the vapor would not cool the processor or the vapor would continue to heat until meltdown.

Naturally you would have to get the heat away from the condenser in some way, all the system does is move the heat. You would probably have to use a fan for that (which unfortunately makes the whole thing noisier :( ). I would have to build in some kind of system that shuts off the system if the reservoir gets empty, the question is also how big the reservoir will have to be.

[Tiger]

A number of people have made them and I have seen a number of articles on the net. There are several design chaacteristics that make them quite difficult to make.
Firstly the system requires a capillary action. What I mean by this is that at the heating interface there should be small tubes or cavities. That's why the one manufacturer uses scintered copper. The reason for this is that the heating causes convection currents to be established in these small spaces. Hot coolant rises quickyl causing a partial low pressure or vaccum in the tube thereby drawing in cooler coolant.
Next you have to design it in a way such that condensed coolant can get back down to the bottom again and be drawn into this cycle.
The above two are usually accomplished by the capillary area being at the bottom and the sides. When the coolant is condensed at the top it is channelled down the sides. The centre is normally hollow so that vapour can rise quickly to the top.
The next thing is that the whole vessel has to be evacuated so that there is low pressure inside for a liquid with a boiling point higher than 0C or pressurised for a liquid with a boiling point lower than 0C.

[Aesik]

The vapor actually moves from one side of the heat pipe to the other because of a pressure differential and particle diffusion. On the hot side the space is being constantly filled with vapor molecules, on the cold side the molecules are condensing and being removed. This creates the pressure difference and causes the vapor on the hot side to move towards the cold side. This is what makes it possible to use heat sinks in a horizontal position and also in space, and even upside down if designed correctly.

The condensed liquid moves from the cold side to the hot side due to the effects of surface tension (capillary action). Liquid is converted to vapor on the hot side, thus reducing the total volume and mass of liquid on the walls of the pipe at the hot side. This causes on imbalance of surface tension forces and 'sucks' the fluid that has condensed on the cold side down to the hot side in order to preserve equilibrium of surface tension forces.

[PhoenixMDM]

Thinking about your boiling thing, what about a freezing point thing? Cuz if i remember right, the same temp thing happens at the fusion point too. Lol, we did a lab in my science class about 3 weeks ago that used a substance which freezing/melting point was like 53C or something around there. I'll see if i can dig up what the substance was...

Well, if you could have the substance on the cpu, like direct-die, it would heat up the solid until it partialy melted, at which point it wouldn't heat up anymore. So if there was a fan or something cooling the substance's container, it could keep the substance at the freezing point...

If you used a substance with a lower freezing point than the one we used in the lab, you could possibly build a thing similar to Strangelove's idea...

I'll try to make a sketch of the idea after i finish my homework...

*EDIT* Found the substance we used! It's called paradichlorobenzene, and is a common ingredient in moth balls.

I also found a graph i made for my lab report, it shows the temperatures of paradichlorobenzene, naphtalene, and water, in 30-second intervals, cooling at room tempurature, and shows the relation between when one is freezing and the other's cooling rates slow. It may be of intrest to you, i'll post it if anyone wants to see it.

[PhoenixMDM]

Alright, i drew up a sketch in Paint...

Hope it's self-explanatory. If you don't get any part, just tell me and i'll explain. The more i look at this, the more it seems feasible... Maybe i'll try it if i can get my hands on a substance that has the right melting point...

[PhoenixMDM]

Ack, pic didn't post right...

[JFettig]

yea its been done, its actually pretty smart, it flows thru onto the core then evaporates causeing it to swell then pushes it thru he radiator type thingy that blows on it and it condences, 3m makes the liquid and i heard that hp is gonna put em in their puters....and they are suposed to cool up to 3.5ghz

http://www.memagazine.org/contents/current/features/heatout/heatout.html

its cool

[PhoenixMDM]

Originally posted by maskedgeek
yea its been done, its actually pretty smart, it flows thru onto the core then evaporates causeing it to swell then pushes it thru he radiator type thingy that blows on it and it condences, 3m makes the liquid and i heard that hp is gonna put em in their puters....and they are suposed to cool up to 3.5ghz

http://www.memagazine.org/contents/current/features/heatout/heatout.html

its cool

U got the idea, except you're thinking of evaporating and condensing. I'm talking about melting and crystalizing. Close enough though

And that link is really good! I want one, heh...
When is HP gonna use them? I'll buy a HP sys just for that reason:p !!

[JFettig]

when is hp gonna use them? i dont know, i know only whats in that article....