Can a Computer Survive in Outer Space Without Overheating?

[ISX]

Not sure where to put this so general might work..

Anyhow. Forgot the hard drive issues or pretend it's a SSD. Would the computer work if I plugged it into the external outlet of the space station and turned it on... Would it just burn up since the components can't radiate their heat?

Problem 2..I am not that into how the computer actually works, but I think all parts other than the fans and hard drive are purely resistive and therefore turn 100% of the power into heat. With that said, could they make an aluminum computer and if they could, how would it work in the same outer space scenario? This is remembering that aluminum at that super cold temp becomes superconductive, so produces no heat...

 

[jedishrfu]

but they can radiate heat into space they don't need air to do it.

The fan issue is that it won't have any resistance since there's no air to push so you could basically disable it.

The harddrives are hermetically sealed so they should be fine unless the pressure on the inside is too great and causes a leak which may reduce stepping motor drag and cause read/write errors (however the electronics may account for it) but I'm no expert here so this is pure speculation.

The other issue is cosmic rays interfering with the memory and cpu causing bit errors which may be enough to stop it in its tracks. Just like an error in a critical part of memory can cause a computer to abruptly stop (eg bad instruction code, out of bound address...).

 

[Bill Simpson]

It has been way too long since I last did a calculation like this, so several people should check this for accuracy.

If I take my 1000 watt computer and pack it inside a spherical black body with a surface area of 1 meter^2 and I want to calculate the temperature of this at equilibrium in space where the temperature of "space" is zero kelvin and I look at

http://en.wikipedia.org/wiki/Black-body_radiation

down in the Stefan–Boltzmann law section

then I think the temperature will be

1000Watts = 5.68*10^-8 Watts/meter^2/kelvin^4*(temp kelvin)^4

and if I haven't made a mess of that then temp in kelvin is about 364 degrees kelvin and that is only about 90 degrees celsius. That is hot for semiconductors but not infeasible.

Can anyone independently come up with the temperature of a 1 meter^2 black body radiating 1000 watts?

 

[ISX]

Alright so what about the aluminum thing?

 

[AlephZero]

You are ignoring the heat input from the sun.

From http://science.nasa.gov/science-news/science-at-nasa/2001/ast21mar_1/

Without thermal controls, the temperature of the orbiting Space Station's Sun-facing side would soar to 250 degrees F (121 C), while thermometers on the dark side would plunge to minus 250 degrees F (-157 C).

Heat input from the sun would be more than the 1000 w/m^2 you were hoping to radiate into space.

 

[ISX]

I guess I am just trying to see if it would work in an absolute vacuum at absolute 0...forget about the sun and all that I just said space because it has similar properties.

 

[Bill Simpson]

You are ignoring the heat input from the sun.

So send along a little umbrella to let it stay out of the sun? Make that large enough, reflective enough, insulated enough and held far away enough that the sun load is close to zero.

Now on the aluminum computer question, you need something more than just a superconductor with zero resistance. You need semiconductors to make And and Or gates, flip flops, memory cells, all those sorts of things. Eliminating resistance, and thus generated heat, from the wiring connecting all of those to each other would reduce power and thus temperature somewhat, but you still need lots of billions of transistors to build a computer and a quick Google search seems to show that someone was finally able to create ONE super conducting transistor about four years ago. Even with that I don't believe that eliminates all heat.

So to sum up, it seems feasible to operate a specially constructed computer in space that only radiates heat and does not use convection for cooling. Vacuum should not be a problem. Generated heat will keep it far far away from absolute zero.

 

[Integral]

Umbrella? Now you know a reason many satilites have a rotation imposed on them. To provide a more uniform temperature to all components.

 

[Bill Simpson]

Umbrella?

Or block the heat from the sun with the large solar cell array you will likely need to generate the thousand watts of power for the computer.

 

[Enthalpy]

hard drives are hermetically sealed...

Hard drives are vented so they keep in equilibrium with the outer pressure. You see the vents at the case.
This may chage soon as one manufacturer introduces a drive having helium inside instead of air.

 

[Enthalpy]

A normal computer doesn't dissipate 1000W. Desktop power supplies generally claim like 600W and are completely unable of providing them, even if the load is shared optimally among all outputs. Dissipating some 100W is more realistic, or 200W under heavy load.

Evacuate 200W is difficult without air, so space computers tend to dissipate less - an other reason being that power is seriously expensive there. Then, the board is designed accordingly and fastened directly to a dissipator, often of aluminium. Since all extension slots are useless these boards resemble more an embedded computer, simple and compact, which is easier to integrate in the typical satellite boxes.

Ambient conditions can be horrible in certain circumstances, for instance near Jupiter (ionizing rays) or if the temperature varies a lot, but inside the space station (or outside if you keep a reasonable temperature) the conditions are very comfortable for electronics. Radiations are worse than on the ground but nothing bad for electronics on this orbit (that's why we put humans at that low altitude). Electronics' life expectancy won't be reduced at all; you lose on bit from time to time (like on bit in 10 hours) for which memory chips have correction mechanisms but other chipsnot necessarily, so either you choose the components or you have to find the proper error-correction algorithm.

Latchup does not happen with current components. It's a myth of space activity, convenient to explain some failures.