Author Options:

expanded graphite heat-sink? Answered

Okay , I've searched and searched and cannot find what I'm looking for. Maybe the other diy'ers can help me.

1.What I'm looking for is a heat-sink made of expanded graphite or (carbon, graphite (∥) as it might be known as well) that is actually in the shape of a traditional aluminum heat-sink with the fins and NOT in the shape of thin film as it is very commonly available. I don't even know if it exists or is manufactured by any company. I've tried contacting some companies about it but they have yet to get back to me. The reason I want this is for the tremendous increase in heat conductivity as in this chart shown on this webpage http://physics.info/conduction/ . The application I am trying to use it for needs a raised, elongated surface area to displace heat semi-uniformly, hence why I need fins like a traditional heat-sink instead of the commonly available film expanded graphite online.

2. If it does exist or is able to be manufactured, would it be a ridiculously priced part or would it be a reasonable priced part? (Just doing small scale tests for now so it would be the size of a FET finned sink).

3. From what I've read it seems possible by maybe layering the films into a stacked lattice, keeping heat transfer perpendicular to the lattice, but what are your thoughts all? (I question bonding the layers myself but I don't know if you could hard press the layers together without a bonding agent since any bonding agent being used I believe would reduce the thermal conductivity quite a bit, but I'm not incredibly familiar with expanded graphite).

Any and all help would be appreciated from this wonderful community
Best regards,


2000F ? what are you making ?

Graphite will only work at 2000F in the complete absence of oxygen. I am actually working on an instrument right now, insulated with graphite, and running at 3300 F.

Tungsten oxidises and embrittles.

Niobium will handle it, if you can't exclude oxygen, graphite WOULD be your best bet if you can. Then consider water cooling.

Tungsten is quite ductile at room temperatures - you can make wire with it. When its heated in oxygen, it become very brittle. The process of becoming brittle is called embrittling.

Can it be housed in an inert gas, or vacuum.

Well i could change from water to an inert gas. that is an idea and actually combines another project i was working on. So instead i'd have a helium stirling engine (which im working on anyway for a reduced friction design) rather than a boiler . Hmm thanks for the idea

From your comments below, it sounds like you want a custom-machined tungsten heat sink. I hope you have a chunk of change to throw at it.

I might just use nichrome since it's in stove heating elements (though im not sure if the elements are solid nichrome or coated) and they are already available in coil form

DO you actually want to cool to 2000 f or simple transport the heat?

Transport thermal radiation and disperse through a medium via convection

Graphite is very brittle, not ideal for a heatsink at all. Expanded graphite has a much lower thermal conductivity than solid graphite. Also, you'll note that the conductivity is anisotrophic - it conducts very well along the grain, but not across the grain.

Hmm the way that article read i thought it conducted perpendicular to the lattice but i read it on like 2 days no sleep so i misread parallel for perpendicular. Well what other material would you suggest that can withstand over 2k f and conduct the heat well, tunsten or nichrome maybe? It doesnt have to be in a heat sint shape it also could be a coil i just figured the fins would give a good surface area to dissapate heat. Any suggestions kind sir?

Nichrome is a rotten conductor, its used for heating elements for a reason.

There have been some experiments with foam graphite heat sinks but as you can see at the bottom of that article aluminum and copper outperformed the graphite by a wide margin.

It is used for some laptop heat sinks because of the weight savings but it is at the cost of efficiency. Laptop CPU's also tend to be low power variants of their desktop brethren and thus generate less heat. Generally the layers are combined some how and then machined.

The problem i run into is the application im working on heats up the heatsink up so much that copper or aluminum would melt. Im looking for something that will withstand well over 2000 f at the point of contact.
its not an electronic component that will be hooked up to the heatsink thats why i was thinking graphite heatsink or even rod but the heatsink shape seamed better for even heat dissapation.