Use of high mass for rapid conduction cooling 1

[slavelabor]

We would like to use a high mass object brought in contact with a heated chuck for the purposes of rapid cooling. The idea is pretty common: use a low mass heated plate for a fast ramp to temperature and then bring that plate into contact with a relatively cool, high mass object in order to quickly conduct heat away. The problem as we see it: what material and structure can we use for the high mass object that will not warp due temp differences from top to bottom? Temperature to cycle between 100 and 600C, in vacuum, forced cooling of mass, and the heated chuck is a high stiffness ceramic. The interface must stay flat within .050mm over the area of the 200mm diameter chuck. It seems like the high nickel alloys would be the best for this application but their thermal conductivities are so poor that we suspect the temp difference will cause warping. Is there anything out there that can be machined and welded that may work here?

 

[zekeman]

How thick is the ceramic plate and what is it's specific heat? Why do you have to keep the interface within .05mm?
And finally, how fast do you have to cool down the chuck?

 

[slavelabor]

The ceramic plate (heater) will mostly likely be 6-8mm thick with a cp of 500 J/kgK. Unfortunately we have to then mount the same size piece of silicon carbide to it for use as tooling, so the overall mass becomes somewhat substantial. The planarity is required in order to keep everything nice and flat for the press operation that follows heating. We would like to cool somewhere around 100K/min. I was thinking of using a cast material such as tungsten due to high thermal conductivity, stiffness, and lack of stress- any experience?

 

[chicopee]

about copper for your high mass object? temp gradient should not be subtantial to cause wraping and you can silver solder the copper instead of welding it.

 

[slavelabor]

That would be the nice feature related to using such a high conductivity material. Does copper corrode readily at high temps? Is there an appropriate coating for something like this? It also has a weak E-modulus (relative to the pressing operation). I did find that MarketTech is selling sintered tungsten that they infiltrate with copper in order to improve thermal conductivity, which seems pretty interesting.

It would seem that in order to remain flat I need
1. high E-mod
2. high thermal conductivity
3. low cte

Perhaps we could use copper if it was bolted to something very stiff outside of the thermal envelope. I guess experimentation is in order. Thanks chicopee.

 

[btrueblood]

Have you considered cast iron?

 

[slavelabor]

No... the thermal conductivity of most alloys is poor, and then of course there's the rusting problem which is made worse by our high temperatures.

 

[btrueblood]

Rusting in a vacuum?

If you have contact between tungsten and air at 600 C, you have an even greater oxidation problem.

 

[slavelabor]

Well, the exterior of this piece will also see the chamber environment which can be anything from air to high vacuum to over pressure. Simply opening the chamber for product insertion will introduce enough moisture (in a lab environment) to wet all surfaces.

What more can you tell me about tungsten oxidation? This is info that I cannot seem to find published anywhere.

 

[btrueblood]

slavelabor,

I can't remember the source, but remember that at red heat (more than 800 F) it oxidizes pretty rapidly. I know that when we had a vendor forming coiled-wire heaters, they would heat the tungsten with a reducing flame to red heat. Another vendor used resistance heating, but you could see the "smoke" of oxidizing tungsten curling up. Both processes would heat the tungsten in air for only a few seconds at most.

We would always try to keep the heater wire temperature below 300 F before backfilling the vacuum chamber. May have been overkill, but I've seen tungsten foil degrade even with this precaution (we surmised that residual oxygen in the test cell at temperature (>2000 F) was causing the degradation). However, we were worried about even a few microns of wire evaporation....

Tungsten oxide forms a blackish "smut" on the surface, and is easily rubbed off. Our other worry is that tungsten oxide has a fairly low vapor pressure (my old CRC handbook of Chemistry doesn't cite it, but newer versions should have it), and the oxide would outgas and "plate" onto adjacent surfaces when the wire temperature rose. But, we were using the wire at incandescent temperatures, so check the CRC for the vapor pressure, you might or might not need to worry about that. We also worried about the oxide forming a hydrate (absorbing water from the atmosphere) which would then have to be outgassed as the heater warmed up. We never proved whether it did or didn't, since there were a lot of other refractory metals used in the heater and adjacent heat exchanger which did form oxide-hydrates (molybdenum, rhenium).

 

[sailoday28]

k*rho*c product of two materials in contact
From the following reference ANALYTICAL METHODS IN CONDUCTION HEAT TRANSFER, GLEN E. MEYERS McgRAW-HILL 1971
pgs 200-202
"One of the interesting applications of the semi-infinite solid solution is the estimation of contact temperature. The contact temperature indicates how ""hot"" or how ""cold"" an object will feel when touching it......."

Myers gives an example of a finger touching a wall and explains that for very short times both behave as semi-infinte solds with the contact temperature given by

A=product of thermal conductivity, density and specific heat
subscript w, wall f, finger
initial wall temp Twi initial finger temp Tfi

numerator= Twi*sqrt(Aw) +Tfi *sqrt(Af)

Denominator = sqrt(Aw) + sqrt(Af)

Contatct temp = Numerator/denominator.

Per Myers
"Thus it is seen that the contact temp is dependent upon the k*rho*c product of lthe two solids"

 

[slavelabor]

btrueblood: thanks for your tungsten experience- this is very valuable for me.

sailoday28: I presume this equation is meant to aid in determining what temperature the thermal mass piece will see while in contact with the 600C heater? This is info as well, particularly if oxidation at the interface is a concern.

This project may have just become a bit easier since we're now thinking of going straight to water cooling rather than air.