LN2/Dry Ice Container Designing

[FishaOfMen]

I've decided that for the winter I'm going to take on the project of LN2 and dry ice cooling. I'm thinking over the design of a container for this stuff and was hoping to get some help from those more knowledgeable in heat transfer.

What's already out there: I've seen some of the containers used by the likes of Macci, sysfailur, and the Japanese guys. The containers end up being simply that: containers. It seems that none of the lessons we've learned in heat transfer have been applied - no heatsinks and no turbulence but that caused by the violent reaction of the dry ice or LN2 encountering heat. Perhaps I'm wrong, but it seems that while these containers definitely get the job done, they can be improved upon.

What I'm thinking about:[list=1][*]Taking my Alpha PAL8045 or Thermalright SLK-800 and welding copper walls around the edges. The only problem here is that with dry ice, it will displace the dry ice chunks a good distance from the CPU core. I'm not sure if there would be any drawbacks with LN2.
[*]Creating my own container out of a solid block of copper. It would stand about 9" tall x 4" wide x 4" deep. There would be a bunch of pins coming from the bottom. I wouldn't have any pins in the center so that I could throw dry ice chunks in the middle. Near the outer edge, I'd have a passage so that a mixing wand could be continually mixing the coolant in a circular motion. I made a really rough bird's-eye sketch to help you visualize this.

[/list=1]
Let me know what you think.

[hydrogen18]

i seriously doubt much performance would be gained. The amount of cooling provided(especially by evaporating liquid nitro) is simply so immense that effeciency is thrown to the wayside. Your ideas are certainly worth a try, but consider this, when using a prometia unit it is so poweful it will not only cool the core to negative temps, but cool surfaces(like the back of the mobo)to condensating tempatures, and it is not even in direct contact like them. So i would focus on somehow concentrating the cooling power available on the core.

[Alchemy]

Originally posted by FishaOfMen
I've decided that for the winter I'm going to take on the project of LN2 and dry ice cooling. I'm thinking over the design of a container for this stuff and was hoping to get some help from those more knowledgeable in heat transfer.

Just to get this out of the way, you understand this is an extremely expensive and extremely short-term cooling solution, yes?

What's already out there: I've seen some of the containers used by the likes of Macci, sysfailur, and the Japanese guys. The containers end up being simply that: containers.

What else would you keep the LN2 in, if not a container?

It seems that none of the lessons we've learned in heat transfer have been applied - no heatsinks and no turbulence but that caused by the violent reaction of the dry ice or LN2 encountering heat.

Your plan is to put the LN2 directly in contact with your board, or on a heatsink directly in contact with your CPU?

If so, the "violent reaction of the LN2 encountering heat" (??) is the LN2 *boiling* as it comes in contact with a solid. A boiling liquid applied to a solid has a heat transfer coefficient several degrees of magnitude above what you can possibly achieve by manually creating turbulence in a one-phase fluid. It's like trying to soup up a Viper by sticking a model plane engine on the front to pull it - the improvement is staggeringly pointless.

What I'm thinking about:[list=1][*]Taking my Alpha PAL8045 or Thermalright SLK-800 and welding copper walls around the edges. The only problem here is that with dry ice, it will displace the dry ice chunks a good distance from the CPU core. I'm not sure if there would be any drawbacks with LN2.

Aside from the fact that you'll probably destroy the core instantly by tempering it improperly when you add the LN2, will most certainly spill it if you're trying to get it into a container that small, and have a better-than-average chance of losing your hands and eyes when the LN2 flash-boils and blasts a 9" column of liquid at your face, and a decent chance of dying an unspeakably painful death when you accidently inhale or ingest the LN2 spraying at you, then no, no drawbacks at all.

Let me know what you think.

I think you don't know what you're doing and would maim or kill yourself if you ever tried this.

Alchemy

[ezlid]

The advantage of LN2 cooling as applied by the guys you refered to is the entire mobo is cooled and all attached components., not just the CPU. And it is not for an operational system. Just a max effort to post some extreme O/C numbers. Now as I understand what you are doing is stick some dry ice chunks in a little container mounted on the CPU and fill it with LN2. Do you know what happens when dry ice and LN2 come in contact with each other? How long will the LN2 last before it boils away and you have to add more? Why use copper for your container? It would be better to build an insulated container so the boil-off is slowed as much as possible. Do you plan to insulate the back side of the mobo? Do you know that most likely this is a one shot thing. The extreme cold shock will cause damage.

Do you have good health insurance to cover the frostbite and other injuries you will most likely recieve.

Finally, do you have any idea what you are doing?

You asked for comments. Please think this over and don't try this at home!

[FishaOfMen]

I’ve been in the midst of a move, so I’ve been away for a few days. Here are the responses to the questions:

Yes, I know that this is a very temporary thing. If I didn’t, I wouldn’t have suggested the mixing wand. I just want to bench some.

Yes, I will thoroughly insulate.

No, I don’t plan on putting LN2 in direct contact with the CPU core.

No, I don’t plan on mixing LN2 and dry ice.

No, I do not plan on using just a heatsink as a container but welding walls to it that would extend about 9” above the CPU core. Anything lower and the water vapor that will go from gas to liquid to solid in the air (from the extreme temperatures) will land on my motherboard and melt.

I don’t appreciate the flames. I was looking for assistance, not thread crap. Alchemy & ezlid, you guys provided no quantitative information on LN2 to copper heat transfer whatsoever.

[N8]

LN is so cold that you will be freezing and condensing water vapor from the air to your motherboard and motherboard components (capacitors, resistors, etc.) all around your socket for up to several inches radius. I do not know how you would effectively insulate the mobo and these components.

I am sitting here right now with only about 4 ounces of LN inside of a stryofoam cup which is stacked inside a second identical sytrofoam cup and ice is forming on the outside of the second cup.

When you apply LN to a solid (for example, metals), the metal will cool enough to shrink in size and dimensions slightly. This is how many parts are press fit together. You need to take this into consideration when making a copper containment device to sit on your cpu. Your mobo will most certainly have to be laying flat so that the weight of the copper container will not stress the mobo. I would NOT attach it to the socket mounts as they will become cold enough to be brittle and the tension of the dimension changes could break things apart.

You would want to attach it directly to the mobo holes and pour in the LN very gradually so the thermal shock will be minimized. One of the reasons people use very large, very thick walled copper containers is to minimize the thermal shock and dimensional changes when the LN is added. I do not think that a heatsink with walls welded to it would work well. The welds may just break or come loose due to the dimensional changes (a possibility). I would test it with LN before actually using it on the processor/mobo to see what happens.

I have access to hundreds of gallons of LN and have been a crazy overclocker for over 5 years and I haven't felt like bothering with all the hassles of LN cooling yet. It is just not worth the effort and troubles. Good luck with your testing.

[TerraMex]

> I do not know how you would effectively insulate the mobo and
> these components.

He can use a lacquer. Use a few layers. You'd just have to cover the areas that have contacts , such as PCI slots, AGP Slots, etc. Works pretty well. I've help a friend of mine do that.

old pic (not my system, i'm not that brave) : frosty.

[N8]

Originally posted by TerraMex
> I do not know how you would effectively insulate the mobo and
> these components.

He can use a lacquer. Use a few layers. You'd just have to cover the areas that have contacts , such as PCI slots, AGP Slots, etc. Works pretty well. I've help a friend of mine do that.

old pic (not my system, i'm not that brave) : frosty.

Yes, I have lacquered/conformally coated computer components before. But lacquering is totally impractical for mosfets, capacitors, resistors, etc. on his mobo for something that he is not going to do full time. Also, this is NOT insulation, it is merely a way to prevent condensation and water from interacting with the circuits. There are still all the areas around the lacquering that would get wet as the water moved around.

[TerraMex]

I've always considered it a fair enough insulation. But you've got lots more experience than i do, so i wont argue with you there .

How about silicone or vaseline ? Combined with insulating foam , plus the lacquer .

[Alchemy]

Originally posted by FishaOfMen
I don’t appreciate the flames. I was looking for assistance, not thread crap. Alchemy & ezlid, you guys provided no quantitative information on LN2 to copper heat transfer whatsoever.

Quantitative information is always secondary to safety concerns, and your plans provoke many of the latter.

Everything you have written so far indicates you know little about heat transfer or handling cryogenic liquids. Nor do you seem to realize how extreme these plans are, or how dangerous.

Making an open container for liquid nitrogen deeper than it is wide is a seriously dangerous thing to do. If you don't find this inherently obvious, there are likely to be a dozen other mistakes you will make.

If you're too arrogant to listen to someone experienced with liquid nitrogen, or to discern the difference between a flame and my attempt to keep you from hurting or killing yourself, then you have no place working with dangerous materials.

Do you think anything in my first post was a lie or exaggeration?

Alchemy

[redleader]

I've decided that for the winter I'm going to take on the project of LN2 and dry ice cooling.

You probably know this, but try to avoid mixing the two.

[hydrogen18]

lmao at this thread. It would seem he has no idea what he is doing...

Anyhow, Alchemy is there a book or something that can be obtained about handling "cryogenic liquids"?

[ezlid]

Originally posted by FishaOfMen
I don’t appreciate the flames. I was looking for assistance, not thread crap. Alchemy & ezlid, you guys provided no quantitative information on LN2 to copper heat transfer whatsoever.

I'm sorry you feel this way. As far as using LN2, there are many safety concerns and we are trying to make you aware.

Also your original post isn't very clear on your intent. You said and I quote " I'm going to take on the project of LN2 and dry ice cooling" and you talked about and I quote again "Creating my own container out of a solid block of copper. It would stand about 9" tall x 4" wide x 4" deep. There would be a bunch of pins coming from the bottom. I wouldn't have any pins in the center so that I could throw dry ice chunks in the middle. Near the outer edge, I'd have a passage so that a mixing wand could be continually mixing the coolant in a circular motion." Since you never said what you were really planning and by the post title, I assumed the "coolant" to be LN2" If you are asking for help, be clear in what you are asking and don't bitch about people demonstrating concern for YOUR welfare. I don't normally do this, I'd rather post a link but I want you to see this:

Safety Notes & Concerns

Liquid nitrogen is a dangerous material. The following is an excerpt from the Air Products Nitrogen Material Safety Data Sheet:

A back of the envelope calculation indicates that the entire contents of a 10 Liter dewar being spilled in a unventilated 274 square foot room with an 8 foot ceiling would reduce oxygen levels below the 19.5% level where Air Products recommends the use of a respirator. Since most classrooms are larger than this, suffocation does not represent a major danger. When transporting the liquid in a car, however, it is probably a good idea to open a window.

The possibility of freeze burns represents a much more serious danger and is therefore our first concern. This does not mean that the demonstration itself is dangerous, but it does mean you must be careful. Dangers include:


Nitrogen can spatter (possibly in eyes) while being poured.

Flying chunks of frozen objects could cause eye injury.

Students (being children) will want to reach out and touch nitrogen or other cold objects. As mentioned above, contact with nitrogen can cause tissue damage, and this must be prevented.

Therefore specific safety precautions should include:


Teachers must stress to their students the importance of not touching frozen objects or nitrogen.


Wear goggles whenever pouring or dumping nitrogen. Nitrogen can spatter into the eyes, and potentially blinding pieces of frozen things can fly around when we drop it.


Use a glove and / or tongs to handle any object going into or out of nitrogen and to carry the nitrogen dewar.

Teachers should familiarize themselves with the following first aid instructions (excerpted from the Air Products Nitrogen Material Safety Data Sheet) for cryogenic freeze burns just in case the worst happens:

If cryogenic liquid or cold boil off contacts a worker's skin or eyes, frozen tissues should be flooded or soaked with tepid water (105-115F, 41-46C). DO NOT USE HOT WATER. Cryogenic burns which result in blistering or deeper tissue freezing should be seen promptly by a physician.

Remember to stress the importance of not touching liquid nitrogen or frozen objects.


I have not seen any posted remarks about ventilation and oxygen displacement, so of course only try this in a well vemtilated room. You know how cold this stuff is.

Which brings me to giving some quantitative analysis for you. LN2 is 77 K, or -320.8 F / -196 C. At those temps the heat transfer quality of the heat sink material, be it Cu, Al, Ag or a brick, is inconsequential. Oh, BTW - the heat transfer qualities of a material doesn't change because you use a different coolant, does it?

And one more analytical fact for you, actually a question for you. How do you induce turbulance into boiling LN2? Maybe there isn't a lesson to be learned on that one.

On second thought, I'm not sorry you feel the way you because it just showed your ignorance and laziness to do some research on your own. At least bring something to the the table before you ask others to do some work for you.

Sorry to all for the long post, folks.

[Alchemy]

Originally posted by hydrogen18
Anyhow, Alchemy is there a book or something that can be obtained about handling "cryogenic liquids"?

I wouldn't know what to recommend. I was taught to use the stuff by a grad student and thereafter supervised for a while whenever I was doing something "high risk" - disconnecting cryogenic lines, moving around tanks, etc. Since I was using safety equipment and using approved containers, I and my department felt a few minutes of training and a few weeks of supervision was enough. Besides, safety in universities is notoriously lax, perhaps on the (often mistaken) assumption that people who can get four-year engineering degrees have a lot of common sense.

Designing containers for cryogenic liquids and using these containers opens up a whole other category of risk, and doing so without significant education and/or experience is foolhardy.

I think ezlid's post is a good primer, but handling liquid nitrogen is one of those activities where you really need someone watching over you for a while to catch you before you make a mistake.

Alone, in one's garage, is not the place to learn about the dangers of liquid nitrogen. People who have that attitude are the same sort of people who get careless (or just stupid) and end up with gastric perforation or other medical horrors.

If I'm good for nothing else on this board, I hope I am at least able to dissuade a few people from doing something very dangerous.

Alchemy