# Water Cooled Stage

Locrian
I have a simple structure I need to design. It is a simple cylinder, roughly 6" in diameter and 8" tall that holds and cools our samples.

The one we are currently using is nothing more than a hollow shell into which two 1/4" tubes are connected at the bottom - one a water inlet and the other an outlet. I have been told that the inside is simple; one of the tubes travels up near the top and the other is a little closer to the bottom. As we pump water through it the water fills and cools the entire stage. The surface would easily reach 1500 degrees C or higher were it not cooled, but is instead operating at between 600 and 1200 degrees C depending on our conditions.

The one we have is pretty mediocre. Firstly, 1/4" connections are simply inadequate to allow the kind of water flow we need. Secondly, the center of the stage is cooled more thoroughly than the edges, which is a no-no. We need to increase its cooling capacity significantly and hopefully make it more uniform.

One idea is just to stick bigger tubes into it. The problem is that temperature gradients on the surface severely limit us. Devising an internal system that would cool more evenly that is also cost effective is something I'm not that familiar with.

Homework Helper
Gold Member
I have this picture of a coffee can with a tube that runs in, then out of the inside of the can and something inside the can is creating heat. Or is the something that's creating heat on the outside? Please be more specific or provide a picture. These kinds of detailed questions seem to pop up here every so often, and the best feedback is obtained when you can provide pictures and better detail as to how the thing works and is used, etc...

That said, have you considered putting a water jacket around the can?

Mentor
Capacity is simply a matter of more flow, obviously - a smoother temperature gradient is far more difficult. However, if the tubes are simply made larger (while the size of the entire device stays the same), you wouldn't just get more flow, you'd also get faster flow and faster flow means a lower delta-T between the inlet and outlet and thus a smoother temperature gradient.

The calculations for this, fortunately, are relatively simple, if this is a steady-state device (constant heat dissipation). Since all of the heat produced inside the device is going to have to be absorbed by the water, doubling the flow would mean cutting the delta-T in half. If you use this as an opportunity to increase the heat dissipation, delta-T is proportional to the heat dissipation of whatever is inside: ie, double the heat dissipation and quadruple the water flow and you get half the delta-T.

Last edited:
Mentor
Q_Goest said:
That said, have you considered putting a water jacket around the can?
I believe that is exactly what is being described - a coffee can inside a coffee can, with water between them.

Homework Helper
Gold Member
Cylinder? Or cylindrical shell? What temperature gradients can you tolerate? What's the actual sample size (presumably centered) inside the cylinder? Power throughput? How are you handling end effects? 6x8 open ended is going to be like driving around in Death Valley in high summer with the windows open and wondering why the A/C is only keeping the vents cool.

Without knowing the deails like what the other guys have asked for, the first few ideas that come to mind are:

-Change the plumbing around so that you have perhaps 3, possibly .375 (-6) inlet lines. Two on the sides and one on the top. Have your scavenge line that is a larger size, say a .75 (-12) at the bottom. The multiple larger lines will give you a larger flow capacity and should help even out the temperature gradients by introducing the cooling water at different locations. I mostly work with tubing and 37° flared fittings, so this kind of stuff would be a piece of cake. I don't know what kind of piping/tubing you guys have for your plumbing right now. However, if you have someone available that can do a bit of welding and plumbing, I would think the changes would cost a minimal amount.

Locrian
Sorry about not including the picture. I want to assure you that no picture I can draw or generate is worth a thousand words, unless they be words of disgust and bewilderment.

On the other hand, I do have a camera:

http://www.pitofbabel.org/Vista/Stage1.jpg [Broken]
http://www.pitofbabel.org/Vista/Stage2.jpg [Broken]
http://www.pitofbabel.org/Vista/30Torr-1.jpg [Broken]

The last is a picture of a sample ON the top (flat part) of the stage (scroll down to the bottom; the sample is a standard SPG-422 tool). The plasma above it is the source of the heat. The water comes into the bottom of the stage with the goal of cooling the top. We would prefer not to have a gradient of more than 30C between each sample. Unfortunately the cooling is not the only issue affecting this, but it is important.

I appreciate your willingness to deal with my poor description. This is what I've gathered so far:

-Higher flow is not only essential to total cooling, but will decrease the temperature gradient due to the water traveling more quickly through the stage
-Multiple inlets would reduce the temperature gradient on the surface
-I should be able to calculate the flow rate necessary to achieve a particular temperature if I fill in the appropriate variables

One reason I posted this question is because I wanted to make sure there isn't a stock solution out there somewhere; I wanted to see if someone said "no dummy, everyone does it this way." Of course, someone still might, but it seems to me that my primary goal isn't the geometric placement of the tubes inside the stage, but the final flow rate I achieve.

Thanks again for your help on this elementary problem.

Edit: I still neglected to say what was actually inside our current cooling stage in the picture. I haven't because I have no idea what is; I've never taken one apart. Based on how heavy it gets when water is pumped through, I believe that there are two tubes connected to it (inlet and outlet) which fill it up like a resevoir. What I don't know is how far into it they reach and their geometry inside. I keep asking if I can cut the old one open to find out but no one here will let me

Last edited by a moderator:
Homework Helper
Gold Member
30 degrees between samples? Or runs? Still going to depend on total power being dissipated to the stage. If you're having to hook a FD pumper truck to keep things cool, just a slow trickle from the tap in the lab sink or what? Do you have any instrumentation on the inlet and outlet line for ΔT, flow rate? "Tap water in, sort of steams in sink from the exit, takes ten minutes to fill" type numbers/descriptions at a minimum. Or, "sounds kinda like a giant espresso machine when it's running --- steams the lab up real bad."

Looking at the discoloration, you've got enough energy throughput that you ain't coming close to a 30 degree gradient between center and edge of the stage for a single run. Might be a big ol' copper slug ironing out gradients inside that stainless jacket, but someone would have given you at least that much information.

Addendum: Just looked at the pictures again, and I'll swear there are roll marks on the end plate; this was just torch cut and welded? No machining for a plane surface?

Last edited:
Homework Helper
Gold Member
I did just a little bit of research on this, just enough to be dangerous! LOL

I'm guessing that the can you show is inside a chamber and heated. There are small square things you call a sample, being subjected to intense temperature that are placed on the circular end (top?) of the can. Perhaps these samples are getting a diamond coating or something, not sure. But the reason for putting these samples in a chamber is to heat them or apply a coating perhaps. So the top of the sample, the side of the sample away from the can, is being heated, and the bottom of the sample which is attached to the top of the can is being cooled. Maybe this is done to prevent the material/sample from overheating. So you want the various samples which are distributed across the face of this can to be uniformly cooled, but right now there must be some problem with them not being cooled uniformly. I mention all this just to verify I'm understanding what's being done.

My gut reaction is I wonder how these samples are attached to the top of the can. They have to be attached in a way that is very uniform, otherwise there is a thermal resistance across that contact which possibly dwarfs the variation in temperature from one side of the can to the other. So how are the samples being thermally attached to the can?

If the samples are well anchored thermally to the top of the can, then perhaps as you say, there is a thermal gradiant across the can. Is there any way to measure this? Have you tried for example, optical methods of measuring it? Alternatively you might attach thermocouples to the underside of the can top surface in various places. In any event, to understand how much of a thermal gradiant there is, there must be some way of measuring it.

It could be right now that the water is boiling and pockets of water vapor form in stagnant areas under the can top. Perhaps you can comment on if that could be a possibility. If that's true, then what you need is to have something more like a heat exchanger that has a flow of water through a block so there can't be any stagnant areas. I might suggest using copper to better distribute the heat, but obviously the copper will melt if exposed to the high temperatures. The copper then has to be attached to the stainless steel (is it stainless?) can from the inside. You would then flow water through passages inside the copper heat exchanger. Does that help or am I missing something?