Custom Copper Powder Heatsinks

[pullbangdead]

How much air could really be in the copper powder if the copper powder crystals are only 1.8 µm (microns) maximum, in size?

If you assume perfect spherical particles all of equal size, the maximum packing factor is about .75, which means you'll always have at least 25% empty space.

Having a range of sizes improves packing, but being in the real world means you'll never, ever reach the full theoretical maximum either.

You'll always have empty space when using a powder, always.

Now if you sintered a powder, as they do to make solid parts in powder metallurgy, that's a whole different story, and a topic for a whole other thread.

 

[frostin]

Its just scraping copper... this can be cheaply done. also copper melts at a higher temp than aluminum. maybe what you could do is get the sand/clay mix and use the axis housing as a mold to form the mix around it... but be sure that it is quite thick and you properly heat treat the mix... if not air/water can cause an explosion or popping of the mix when you are getting it up to temp (i think copper is around 1500-1600f). Ive made a forge before that was a cheap make and it cost me like maybe 50 bucks. I had an old forge I made I sold to a friend of mine who has ron reils (forgot how spell it) burners. These burners are cnc machined propane burners and use less propane than a home made burner. If you want to make a heating metal forge its quite easy considering if you have a welder (if you want to run propane). My first one I built was from a 10 gallon old air tank. I cut off the ends to make one of them a door and the other for the exhaust. I might be able to get ahold of this material I used if my friend still has it some where... Anyways heres what I did:
Got a 10 gallon air tank
Cut the air tank with a cut off wheel (a torche or plasma cutter is best) 1 openning approx 6 inch square diamater. The exhaust I cut it to about 5x3".
The next step was to make where the burner would go in
I used a hole saw and drilled a hole in the tank.
Next step.. Applying the insulation and heat coating to the tank.
I first tack welded some nails inside the tank to hold the insulation.
Then I measure and cut the insulation and wrapped it around the tank.
I put about 3 inches of insulation (the more the better but you lose work space).
Then I applied the coating that deflects and extends the life of the insulation.
Let it dry overnight
I already made my home made burner before I did all this but its good to do it now while you are waiting for the coating to cure.
Note: the hole cut earlier was about a 1 3/4" hole in the tank.
Next get some black iron or metal pieces at your local hardware store.
Get a piece of 1 1 /2" about 3-4" inches long. This will be the "flare" from your burner.
Next get a piece of black iron and press it onto the burner (it will have to be 1 1/2" OD so it can slide in the 1 1/2" ID).
This piece will be about 6 inches long.
Next on the top of it drill a whole through the threaded part using a 1/4" drill bit.
After than get a reducer that will go to the pipe and screw on the 1 1/2" OD thread.
The reducer will be about 1 3/4" to the pipe size that the 1 1/2" OD is.
Now take a piece of 1/4" brass pipe (not tubing is wayy to thin) about 3 inches long.
Mark the center of it.
Now you will need to drill a very small hole in it but dont go through all the way!
The drill bit is about .045 or .050 either will work... But If you do a google search on this there is a more in dept of a DIY burner.
Remember the hole you drilled earlier? Slide the 1/4" nipple through it and make it is center inside the pipe. This is where the propane exits the burner.
Screw a 1/4" pipe cap on one end of the nipple.
On the top of the reducer weld a nut on it flat with the top of the nut on the top of ruducer. Now cut you a piece of metal and drill a hole on one side of it thats the same bolt size your nut is. This will be the choke of the burner. Make it round and not square. Bolt the piece of metal to the top of the burner but dont tighten to much as you need this to be flexible enough so you can rotate the flange you have made to adjust the air going into the burner.
Get a pressure regulated propane tank that is adjustable.
Get a 1/4" shut off valve that is regulated for propane and screw it into the other end of the 1/4" nipple. This needs to be a double female valve btw.
Screw the valve to your propane house.
Now crack opening the propane tank slightly with the choke on top of the burner opened up just a smudge. now light the burner but keep enough distance from it so you dont get burnt.
Now crank up the pressure on the propane and if done right you will have a nice scorching flame that sounds like a rocket ship. If your flame is off adjust the 1/4" nipple so it is aiming down center the burner and tighten her up.
Burner is now done.
Remember the hole you drilled in the air tank? Get you a piece of round black iron and make a wedge on one end so that it seats nice on the hole you cut earlier. Now you might need to cut a hole in the insulation with a hobby knife before you weld this piece on. Dont wory if you mess it up a little cause you probably have some of the coating left over and patch it around so it makes a nice hole.
Cut 4 holes in the black iron and tap them (I Used 5/16 - 18) make sure you have enough room to get a wrench on the bolts that will be holding your burner.
Now Slide your burner through the piece and leave about approx 1/2-3/4" of the burner inside the insulation ( I used a tape measure ). You need to leave this amount because you dont wont the burner sticking out of the insulation. Now make sure the burner is straight and tighten up the set bolts to hold it in place.

ALL Done now! If done right you have a forge which you can make knifes, melt aluminum, copper, etc..
If you need any help feal free ask me or visit ron reils webpage has an excellent turorial or just google home made forge.

 

[ZRTMWA]

Wow that's very detailed. Tell me you just cpied and pasted that from somewhere else, becuase I don't have a welder. Sorry.

Would using microscopic copper flakes make less room for air than microscopic round balls?

 

[pwnstar]

A huge portion of it will still be emty space. Flakes dont always lay flat... This seems like a messy alternative to a heatsink.

 

[frostin]

Wow that's very detailed. Tell me you just cpied and pasted that from somewhere else, becuase I don't have a welder. Sorry.

I typed it up. I have a good brain of remembering a 2 day project lol...

 

[Benm]

If you assume perfect spherical particles all of equal size, the maximum packing factor is about .75, which means you'll always have at least 25% empty space.

Having a range of sizes improves packing, but being in the real world means you'll never, ever reach the full theoretical maximum either.

You'll always have empty space when using a powder, always.

Indeed, the packing density does not depend on partical size. If you stack cherries in boxes, or watermelons, you'll be leaving the same ca 25-30% of empty space. This works the same micrometer-sized particles.

If there is some size distribution you could get as good as 90%, but all of that does not matter. Even at 99%, the 1% air will still have a greater thermal resistance than the copper. It needs to be solid to work.

Thermally conductive grease itself is a rather poor conductor compared to any metal, and its only benefit is that it will fill tiny gaps between surfaces.

 

[dr-ebert]

No, that's wrong. Even with 50% air, you could have nearly the same conductivity (but less heat capacity). Heat conductivity depends on the contact area between particles, not on the amount of empty space - that affects only the heat capacity. The air, being essentially a non-conductor of heat (compared to Cu), simply doesn't count. Having spheres touching in a densest packing, for example, would give you high density but still low conductivity due to the low contact area. On the other hand, having a solid cube with lots of channels drilled through it from all 3 directions would result in a low-density sponge-like block which would still have high conductivity.

As someone mentioned above, powder will be less effective than the solid metal (I'd guess 50% to 10%), depending mostly on how tightly it's packed (tight packing makes for good contact) but more effective than the common thermally conductive paste.

So it depends on the use. As a heat sink, it would be less useful than solid metal due to the reduced heat capacity and conductivity. As a heat conductor/heat spreader, it depends - it would be less useful than solid metal but possibly more useful than solid metal with an air gap (or even gap filled with thermal paste) between it and the diode and/or housing.

Powdered copper would probably be much better than the common brass washer-type heat sinks.

 

[Benm]

You're right, its mostly about the contact surfaces. If you had a solid heatsink and drilled holes through it as to remove half the material, it would be 50% air, and 50% as good a thermal conductor as what you started out with.

The actual result will vary on the size of the particals, and also on their shape. If they are round, expect very bad thermal contact. Take 2 apples, place them next to eachother and see how small the area is that actually touches.

When you compact a power, the particals are deformed increasing contact surfaces, but it'll still be pretty much worthless as a whole.

 

[ZRTMWA]

But a copper flake powder heatsink would perform better or at just as well as an aluminum heatsink?

 

[Benm]

Not by a long shot really.

If you need it demonstrated, get a thin glass tube, and fill that with copper powder. Get a copper rod of equal dimensions.

Fire up your gas stove, hold one rod in each hand, and the other end in the flame - You'd drop the rod much quicker. In fact, even if that rod were a hollow tube such as used in plumbing, its like to still be a better thermal conductor than the powder, no matter how hard you try to compact it*.

* sintering excluded here

 

[dr-ebert]

I've discussed this a bit here. Copper is about "twice as good" as aluminium, taking heat capacity and conductivity into account (take this as just a very qualitative statement). You get the most out of both copper and aluminium heatsinks only if you press the diode directly into them. Otherwise the Aixiz brass case forms a sort of bottleneck for the heat transport due to its lower conductivity, so if you're using the standard type of heatsinks (which you put around the Aixiz), it doesn't really matter except for the "cool factor" of copper.

 

[Benm]

I'm affraid it will be mostly guess work: As far as i know there is no specification for the thermal resistance to the aixiz module, nor for the module to ambient.

Especially the resistance from the diode to the module is tricky, since it will depend on how you insert the laser diode. I've heard that there is a fair deal of tolerance in the size of the mounting spot, and sometimes the diodes fit in loosely. This would be a concern.

The ones i've used took a fair amount of force to insert though, and i'm sure the resistance from diode to aixiz is not more than a few K/w. The module itself to ambient has a great thermal resistance, i'd speculate in the 20 to 50 K/w range, so adding heatsinking to it will help.

If i have the time ill try to do some measurements with and without a heatsinking star around the module.

 

[dr-ebert]

Hm, from what I've heard, you need a considerable amount of force to insert the diode - from engineering drawings I've come upon it seems that the diameter of the mounting hole in a module is 0.1mm less than the diode rim diameter (which has several indentations around the rim so that it can give). So we should have an intimate metal-metal contact with negligible added resistance. The resistance from anything to air (or another fluid) depends heavily on the fluids flow, so there won't be a fixed number.

I currently run my lasers with just an Aixiz module, no further heatsink. Once that becomes hot, I stop. That means maybe up to one minute (I've never timed it) with an LPC@420mA. With a PHR@120mA, it heats up much slower. I guess with another heatsink around it to absorb heat and spread it to the host casing, it could run indefinitely without thermal problems (which does not mean infinite lifetime as there are other limiting factors). Unfortunately I have no means for metalworking, so further plans are on hold.

 

[Flaminpyro]

Dr. E if you need some metel worked drop me a line I would be glad to help..

Hm, from what I've heard, you need a considerable amount of force to insert the diode - from engineering drawings I've come upon it seems that the diameter of the mounting hole in a module is 0.1mm less than the diode rim diameter (which has several indentations around the rim so that it can give). So we should have an intimate metal-metal contact with negligible added resistance. The resistance from anything to air (or another fluid) depends heavily on the fluids flow, so there won't be a fixed number.

I currently run my lasers with just an Aixiz module, no further heatsink. Once that becomes hot, I stop. That means maybe up to one minute (I've never timed it) with an LPC@420mA. With a PHR@120mA, it heats up much slower. I guess with another heatsink around it to absorb heat and spread it to the host casing, it could run indefinitely without thermal problems (which does not mean infinite lifetime as there are other limiting factors). Unfortunately I have no means for metalworking, so further plans are on hold.

 

[ZRTMWA]

Not by a long shot really.

If you need it demonstrated, get a thin glass tube, and fill that with copper powder. Get a copper rod of equal dimensions.

Fire up your gas stove, hold one rod in each hand, and the other end in the flame - You'd drop the rod much quicker. In fact, even if that rod were a hollow tube such as used in plumbing, its like to still be a better thermal conductor than the powder, no matter how hard you try to compact it*.

* sintering excluded here

This doesn't apply; I said "would a copper powder heatsink perform as well as an aluminum heatsink?".

I didn't know about the negative effects of using an Aixiz case. If I ever get a diode in an Aixiz I'll make sure just to use an aluminum heatsink.

 

[Benm]

@dr-ebert: I'm considering continous operation in a stable situation here. If you run something only until it gets too warm, you're not heatsinking in the general sense of the term, you're buffering with thermal capacity.

If you'd put a laser diode in a kilogram copper block, and then in a styrofoam box, it would take very long for the diode to overheat, but it eventually will. This is a situation with large thermal capacity but near-zero heatsinking.