Vacuum Sealed Computer

Factual Questions
1

As Cecil said in one of his articles, space is a temperature sink. It seems to be the perfect way to cool a PC.

So why don’t we have them? Or maybe I am just overlooking them.

What obstacles are there? I know hermetically sealing cables coming out might be an issue, but other than that I can’t think of anymore problems.

Seems to me that you could just throw a computer into a vacuum chamber. Maybe the vacuum would have to be a conductive metal, and perhaps outside of that you might want a refrigeration unit to cool off the air around the chamber, but I am still surprised I have no seen a PC mod like this.

It’s like those vacuum freezer bags. At the very least someone could have wrapped their PC in one of those vacuum bags :stuck_out_tongue:

2

Space is cold due to an absence of things to retain heat, a PC has lots of heat and with no air would probably not cool as effectively. Vacuum is not equal to space.

Now you wanna have fun you can submerge a pc in mineral oil and it works just fine.

3

Yeah I’ve seen that.

I did just read while searching some more that having a dense gas would work better, but I guess a liquid would be even more conductive.

You would still have to circulate more and cooler gas/liquid though.

So if it was a chamber of mineral oil then you would need pumps at the top to take out the warmer liquid and pass it through a radiator or refrigeration unit back into the bottom of the chamber.

Also, mineral oil might cause problems later on. Especially for any fans. I suppose you could make custom fans that would survive. I am just iffy about any liquids. Mineral oil isn’t conductive?

4

Fans do just fine, although one could say intuitively the heavier fluid would be a problem but this has not been the case in actual testing. There are examples of these machines that have been running for years with no fan failures.

The thermal load that a 20 gallon tank of mineral oil can absorb is huge compared to air and spot heating is far less likely to develop.

Adding aquarium aerator stones actually acts as a form of cooling of the oil. I will try this someday if only as a curiosity.

5

Nope, it would be the absolute worst way to cool a PC because there would be nowhere for the heat to go (excepting pure radiation).

In an atmosphere a PC can cool itself by radiation and conduction into the environment. I think the radiation component would be the same either way, but conduction into the air is the main method of cooling and that wouldn’t work in a vacuum. Your PC would overheat and become an expensive paperweight very very quickly.

6

Vacum acts as an insulator. That is how Thermos bottles keep your drink hot or cold.

7

Things can cool to a very low temperature in space, but not inside an earthbound vacuum enclosure. As Cecil says:

The thing about deep space (in the absence of bright/intense sunlight) is that there’s really nothing anywhere around you. The only mode of heat transfer is radiation. Everything radiates heat according to its temperature; in deep space, there’s nothing out there radiating heat toward you, so you just keep on radiating heat away until you get damn cold.

If you evacuate the inside of your PC, the walls of your PC case are still at pretty much room temperature, thanks to the air on the outside. Now your CPU is engaged in a radiative battle with those warm case walls, and so it’s going to get much hotter than those room-temperature case walls before its temperature stabilizes.

And it won’t radiate very well, at least not in its current configuration. You’d need a heat sink with huge fins that have a good “view” of the case walls in order to exchange thermal radiation with them. The standard conductive/convective heat sink on your CPU right now is poorly-suited for this.

8

In other words, space cools (almost) absolutely, but very slowly.

9

The other point is that it isn’t the vacuum of space that makes it cold. Not directly anyway.

The logic that goes - space is cold, space is a vacuum, therefore a vacuum is cold - is faulty.

As described above, space is cold because is is almost entirely empty of things radiating at you. The 3 degrees bit is the aspect of space that is radiating at you. (It is an interesting question as to why the sky isn’t totally filled with stars making up a solid visage of stellar surface. One that E.A. Poe solved.)

If you put your computer inside a space simulator you could cool it. Such simulators are used to test satelites and other spacecraft. They are big vacuum chambers with black walls cooled with liquid nitrogen. Not quite 3 degrees K, but close enough. Put a big enough radiator on your computer and thermally couple it to the heat generating bits and it would work. Easier to just push air through it.

10

See Olber’s Paradox.

11

Nope. Mineral oil is used to fill high-voltage transformers. It’s an excellent electrical insulator and a pretty good coolant.

12

I’ve seen pictures and webpages about these tanks, and I’ve wondered: Do the fans actually do anything?

If you had a pump to move the oil around, it seems the fans then become unneeded.

13

If you have a fan to move the air around, then the pump becomes unneeded.

Using liquids versus gasses as heat transfer agents is roughly equally expensive for a given load, though it depends on how you keep score. Gasses do need more room, for example - a duct versus a little pipe.

One huge difference is that if a little air leaks all over your paperwork and furniture, you’re not going to be nearly as upset…

14

The Cray 2 Supercomputer did that. Note the blue ‘waterfall’ in the background; the cooling tower for the machine. (But they didn’t use mineral oil, instead a more inert 3m product called Fluorinert. It was originally developed as a blood plasma replacement for emergency rooms.)

15

But how do you cool the mineral oil without some sort of radiator/refrigerator? I know it would probably naturally radiate heat away to the chamber, and then into the air, but wouldn’t it be more effective the other way.

Or are you saying that pumps are needed if there are fans and vice-versa because they both achieve the goal of liquid circulation. A fan could technically be used to circulate water into a radiator.

–

My original concept for this was refrigeration. I wanted to build a PC into a mini-fridge, and I was trying to think about how to avoid condensation. Then I remembered those vacuum sealed bags.

I was thinking if I hermetically sealed all of my components in bags… possibly using very small rubber boots, and wrapping cables in rubber to make them into rubber tubes that fit very finely into these rubber boots I could create a perfect seal (or nearly perfect)… then my components could sit in the fridge without having to worry about condensation.

I’m not sure if they make any bags that are particularly good at conducting heat, but I guess that would help for heat transfer.

Is my idea crazy? The vacuum would at least prevent condensation inside the bag as there would be no water molecules, right?

16

A vacuum sealed computer would have the benifit of never needing to be cleaned of dust and dirt, also, a controlled environment would be vey useful. I will list a few ideas for brainstorming:

The vacuum chamber: Naturally, a round stainless steel pressure cooker or a scientific vacuum chamber, aproximately 14’’ - 16’’ in diameter. 15’’ - 16’’ in height.With a transparent lid. This really depends on the form factor of the Motherboard.

The valve/gauge assembly: Most pressure cookers/scientific vacuum chambers come with a pressure gauge and valves built either into the side of the chamber or mounted on the lid.

A vacuum pump.

The motherboard: based on the largest stainless steel pressure cooker I found (14’’ in diameter), The X99 Motherboard from EVGA would fit well with extra room for connecting peripherals. the dimensions are:
E-ATX Form Factor
Length: 12in - 304.8mm
Width: 10.375in - 263.5mm

The coolant: Liquid cooling makes most sense to me. This is all open to experimentation, so maybe something like nitrogen gas would be more efficient. For the sake of this post, I will choose mineral oil. However, I do not know how mineral oil will react in a vacuum.

The Heat sink: The stainless steel container itself is technically a heat sink. Fans (or pumps) could convect the oil, pushing hot oil up and away from the cpu while fans at the top push the oil down along the walls of the chamber itself dumping the heat onto the surface of the chamber and recycling the oil over the componants. However, this only transfers the heat, the ambient air outside of the chamber could dissapate it, or a coil of copper tubes could wrap around the outside of the cylindrical container and connect to a passive cooling element or, better yet, you could pass a refrigerant through the coiled copper tubes and just use a series of peltier coolers to run the refrigeration cycle. It would be silent.

The motherboard would have to sit on a rubber pad to protect it from making contact with the chamber floor. As for the wires running out of the chamber, Im quite stumped myself on this one. You’d have to have a hermetically sealed back panel installed into the side of the chamber to connect peripherals or to even give it power. I know they make hermetically sealed USB ports, so youd just need to drill a hole to size, and install it with a seal. Other than that, Im not sure. Wireless keybord and mouse would be an option, they also make wireless 4k streamers for the display feed. wireless internet peripheral?? I dont know how far id be willing to go with that idea. still, how would you get power to the machine bubling away in a its snuggly mineral oil bath chamber?

17

Induction. If I can charge my phone with one of these: charger - surely you could get enough power through the wall of the chamber to run your computer.

18

They do have cooling fins which work by natural convection. They also heat up during the day (heavy use times) and cool over night, so yes they can absorb a lot of heat without heating up to critical levels.

19

As other posters have said, the only way to lose heat in a vacuum is by radiation.

To lose any significant amount of heat the radiating body needs to be very hot- hundreds of degrees C. That is not compatible with semiconductors as we know them. It is difficult to think of any way to pump heat ‘uphill’ to such a higher temperature.

This problem was addressed by folks planning to generate lots of power in space, out where solar cells are no good. A turbine using potassium vapour was developed that rejected heat at around 760 degC.

Here are some details of an extraordinary bit of technology:

http://www.douglas-self.com/MUSEUM/POWER/potassium/potassium.htm

20

What problem are we trying to solve here? If we just want a dust-proof computer, there’s no need for mineral oil or vacuum. Just put a standard PC in an airtight metal box, and cool the whole box. Depending on the power consumption and box, it may be enough to just have a fan blowing air at the box from the outside. Or you can attach some type of refrigeration system.

I’ve worked on one project that called for a PC to work at extremely high altitude, where there isn’t enough air to keep a conventional PC cool. (This was a scientific instrument on a high-altitude balloon.) Their solution was to put the PC in an airtight pressurized box.

Alternatively, you can buy fanless conduction-cooled computers designed for military or industrial use.