Water cooling in a vacuum

[rwsasser1]

Here is the deal, I have a copper shield in a vacuum that is exposed to q=0.69kw of radiant energy. I need to cool this shield to prevent any warpage. I am going to weld copper tubing to the shield to use for water cooling. So the heat will be removed from the copper shield via convection. So I need a convection q=0.69kw. The problem is I dont know the temp of the shield, I have the temp of the water, the surface area is under my control to be whatever is needed. I dont have flow yet but will after friday. Can anyone lend a helping brain?

 

[ko99]

Please be aware that changing the shield design can change the amount of radiant energy it absorbs. In fact it's not clear how you could know the radiant heat without knowing the shield area. There's a formula to estimate this. Let me know if you are interested.

Also I don't understand how you could know what the water temperature will be without knowing the flow rate. Again, there's another formula for estimating this. Let me know if you are interested.

Now back to your question: shield temperature
If you know the water temperature, that reduces the effort to finding the thermal losses between the water and shield. The formula for this isn't simple, though. The losses depend on the geometry of the assembly (shield thickness, distance to cu tube). If these losses are low, the shield will only be slightly hotter than the water. If the shield is thin compared to the distance to the copper tube, then the shield temperature will not be uniform and warpage could be pretty severe.





ko (

http://www.ecooling.biz)

 

[Montemayor]

rwsasser1:

First of all, if you can obtain convection in a vacuum you will be the first one in the Cosmos. No one, to my knowledge, has ever succeeded in moving molecules where there are none to begin with. What you must be referring to is conduction, not convection. The radiant heat collected by your copper shield will be transmitted (in a vacuum) to your watercooled tubes by conduction.

Is this an industrial project? Or is this a lab experiment or problem? How do you sustain a vacuum under flat glass? The glass thickness must be great. Please give us full details of the basic data and specific scope. Otherwise, I would have to consider this an academic exercise.


Art Montemayor
Spring, TX

 

[ko99]

Convection sometimes refers to any fluid heat transfer, not just air. But I agree with Art, there are too many inconsistencies in the problem discription.

ko (

http://www.ecooling.biz)

 

[rwsasser1]

I am speaking of heat transfer by convection form the copper shield to the water traveling through the copper tubes via the surface area of the inside of the copper tubing.

I know the original temperature of the cooling water. I will be measuring the flow rate when I travel to the site and can measure it with a flow gauge.

I know the area of the shield. it is approximatley 23% of the area under the heat source which is being powered by 3kw, thus the .69kw value I have specified.(assuming black body to be conservative)

Not sure where the glass comment came from, there is no glass involved.

Just a little on the project background. This is a sputtering film process in which elements are sputtered onto film. The element in question is nickel, which is hit with 3kw. The nickle is then sputtered onto the film in a pattern that is produced by the shield, kind of like a stencil. The film rotatates through the vacuum on the surface of a drum. The shield must be placed very close to this drum in order to meet with pattern tolerances. If there is expansion in the shield, it could affect the pattern, or worse damage the drum. Hope this additional info helps.

 

[IRstuff]

How thick is your shield and/or how far apart are the loops on your coolant pipe?

Overall, I'm confused.

A sputtering process is generally NOT a thermal process, but a kinetic process, e.g., bombardment by Argon.

A thermal process, e.g., evaporation, means that there is a crucible that contains molten nickel.

So is it an evaporator or a true sputterer? If the latter, what exactly is 3kW?



Your process is presumably already running. Are they having problems attributed to warping? Why not simply instrument the shield with a few small thermistors? Or better yet, borrow or rent a thermal camera and image the shield to determine the actual temperature and distribution thereof.

TTFN

 

[ko99]

Thanks, rwsasser1. The .69kw sounds like a very rough estimate. In addition to the inlet water temperature, measure the exit water temp. From this, you'll know how much heat the water is actually removing: Q = massflow x Cp (T1 - T2)

If you're unable to get probes or a camera inside the vacuum chamber, then use the water temperature to estimate the shield temp. Water temp will vary along the tube. Conduction loss from the shield to the cooling tubes will depend of the geometry (shield thickness and tube spacing)

ko (

http://www.ecooling.biz)

 

[rwsasser1]

There is no shield in place as of yet. We are trying to determine if this is feasible. The shield is 3/16" plate and right now I have it designed at a distance of 0.0787" away from the drum. It is a sputtering process with argon but the target(nickel) must have a charge in order for the argon elements to be attracted to it which is why it has voltage and current applied to it(3kw), thus causing it to give off heat. So I cant measure the outlet tamp because there is no outlet temp yet. I am trying to figure out how much cooling I need.

 

[IRstuff]

But you could monitor the drum surface temperature?

How far away from the source is the drum? If there were substantial amounts of heat getting to where it shield is supposed to be, the drum would see the entire amount and the film would experience a large heat influx; how is that handled?

TTFN

 

[rwsasser1]

The drum is water cooled also.