Vacuum chamber heat transfer

[Woody1]

I have a heat transfer problem. I use an electron beam vacuum deposition system to coat lithium on a nickel foil substrate in a continous web operation. Yes this is for making batteries!
The problem:
The foil collects heat by radiation from the e-beam guns,sufficient to melt the lithium back off the substrate.
I have a cooling drum cooled by a chiller at 20deg C , the foil is wrapped around approx. half the drum , drum is 20" in dia. 10" wide. contact with the drum is by tension.
Drum is 6061 aluminum with a polished surface.
I need a way to get maximum heat transfer from the foil to the drum. Remember this is a high vacuum operation.
Any kind of coating?? surface prep?? different material??
All ideas accepted.
Thanks

 

[gunsmith]

I do not know the the heat transfer coefficient of aluminium. The material with the highest coefficient is pure graphite 168 W/(m*K) and brass 116 W/(m*K) when having 63% of copper (short time heat transfer to a gun´s cartridge). I have no idea if these material are useful for Your process.

Andreas

 

[IRstuff]

Does it have to see the source? While that produces the best uniformity, there are locations that are not in direct line of sight of the crucible, which would reduce the radiant heat load.

TTFN

 

[prex]

I'm afraid that with polished aluminum you are already in an optimum condition. Supposing the nickel foil is either sufficiently thin or flattened to insure good contact, there is not much more to do.
As suggested by IRstuff, can't you provide radiant screens on the e-beam gun to shade the area around the melting zone? prex
motori@xcalcsREMOVE.com

http://www.xcalcs.com

Online tools for structural design

 

[Woody1]

Thanks for the input. The process is working but I was looking for a way to increase yield.
The foil I am using is very thin .0007 and at least a 32 finish as it comes from the supplier. Contact is probably at it's max right now.I was hoping that someone had a new type of coating that could be put on the drum that would improve conduction of the heat.Unfortunatly that doesn't exist yet.
Positioning of the guns IS helpfull, a slight off-set from direct line of sight did help, (tried that) I believe I am going to try some additional cold-plates near the drum to block more of the radiant heat.
This is best described as a thick film process as I am putting about .0005 thick deposit on the foil.
I guess I'll have to be satisfied with what I have.
Woody Ebersold
Consulting Engineer
"AN EXPERT IS ALWAYS SOMEONE FROM OUT OF TOWN"

 

[gunsmith]

I don´t what You are really looking for but I saw a table in an older book (Hütte 18th edition, 1902, page 276 Band I) where Al has an heat transfer coefficient of 12.4 to 13 (m*st*°C) [I do not know what physical unit it is] and that for pure Cu is 330.0 (unit is the same as for Al). So, if Your drum is made of pure copper it would have a heat transfer coefficient which is 24 times higher than that of Al. What coating You are looking for that will make this job better than a pure copper drum?

Andreas

 

[Woody1]

Copper was considered BUT copper is reactive with Lithium and cannot be used without very big problems.
Due to the properties of the material I am using, I am limited in choice of metals.
Thanks Woody Ebersold
Consulting Engineer
"AN EXPERT IS ALWAYS SOMEONE FROM OUT OF TOWN"

 

[gunsmith]

Here are coefficients [W/(m*K)]of the materials used for CPU cooling devices:

aluminium = 234
copper = 398
argentum = 429

That shows that the coefficient of "Hütte" for Al is wrong or the computer indutry uses an Al-alloy. Perhaps there it is possible to cover an copper drum with argentum getting rid of the reaction between copper and lithium.

Andreas

 

[IRstuff]

possibly it was alumina, which has compositional variations with thermal conductivity in the 10 to 30 W/m-K range. There is an 85% alumina with 13-17 W/m-K conductivity:

http://www.matweb.com/search/SpecificMaterial.asp?bassnum=BA2C

TTFN

 

[waerf]

and why dont you coat a copper drum with anything that isnt reactive?

 

[kpfam]

Heat transfer increases with difference in temperature. Is there any reason for not dropping the temperature of the drum?

Another angle to look at is whether the problem may be because the heat is not removed from the drum, then it will be a pumping or chiller capacity issue.

 

[butelja]

Is there a way to increase the contact angle of the foil with the drum? If you do, you will increase the residence time and improve the heat transfer.

 

[Astroboy]

Hi
The heat xfer by conduction to the drum should overwhelm the xfer by radiation unless the guns are really hot, 1300F and very close to the foil, say, 5 inches. What are your parameters? Without air your contact with the drum should be quite good with a thin flexible foil. In general it sounds like what you have should work quite well. What are you experiencing?

 

[Woody1]

Well, the cooling drum is held at 38 deg F by a chiller,the foil wraps half way around the drum,drum is 20" dia. drum is polished 6061 aluminum, foil is 32 finish nickel .0007 thick X 5" wide.
The heat problem.
e-beam crucible temp is 550deg C and distance is 8" from foil. Deposition thickness must be .0004 to .0007,yes REAL thick film. Foil transport speed is 9 to 15 ipm and only sees the heat for about 40 seconds. need that long for deposition build-up.
Problem is trying to increase yield by raising crucible temp to 600deg C but foil gets hot enough to remelt the deposition coating at about 180deg C creating blisters and generally making a mess.
After all things considered, I may be at the wall with 550C
System works well but more is better.

I saw a trick a few years ago that I think I will try.
Nickel is magnetic so I will have some (about 40) rare earth magnets machined into the drum surface to get a good mechanical bond with the drum. Expensive way to go but!!! Woody Ebersold
Consulting Engineer
"AN EXPERT IS ALWAYS SOMEONE FROM OUT OF TOWN"

 

[rjw57]

I would definitely look at the heat transfer of the cooling media on the inside of the drum as the source of problems. Could be that you are creating a vapor layer and seriously affecting your ability to conduct heat away from the aluminum surface (Leidensfrost effect - dancing water on the frying pan). Also, I would think about an chemical treatment on the aluminum drum to improve emmissivity since any gaps you do have between the foil and the drum rely on radiation heat transfer (vacuum problem) and not on gas conduction. Are you certain about the "flatness" of the circular drum? Perhaps a mild crown would help distribute the foil (not sure about permanent strain issues)? Can you move the crucible and mess with the geometry to get it further away from the foil without affecting the deposition rate.

 

[gunsmith]

Hi Woody1,

don´t look at the drum - the really problem is the nickel foil. Let´s have a look to the energy content of Lithium layer - at 600°C it is 2.03 J/mm foil when Li has specific heat of 3.58 kJ/(kg*K) and a density of 530 kg/m^3.

The heat transfer coefficient of Ni is 89.99 W/(m*K), the Ni layer is 1.7*10^-5 m thick. Putting in the temperature difference of 580°C, the contact time of 1 mm length of the foil is 126.67 s (speed of = 381 mm/s, contact length to drum = 781 mm). We get the energy transport of the Ni-foil with 1.23 J but the energy content of the Lithium layer is 2.03 J/mm (38.5% remain in the Lithium-layer).

The simple calculation shows that the Ni-foil is unable to transport the energy content of Li-layer with that temperature difference (even suggesting that we neglected the energy content of Ni-foil itself in this ratio).

The conclusion is that you have only two parameters to play with - contact time to the drum and temperature difference between the drum and the Ni-foil.

Andreas