Nixie - 23-5-2008 at 19:21
Is it OK to melt aluminum in a magnesia crucible? I also have a graphite crucible but I'm just using the crappiest
charcoal-briquettes-in-a-coffee-can-and-a-blower furnace, so there'd be too much oxygen for the graphite.
12AX7 - 23-5-2008 at 20:44
I don't see why not. Seems awfully expensive though. And what about when it breaks? (Around charcoal I feel it's a "not if but when"...)
Graphite doesn't burn very fast at aluminum temperatures, even in pure air.
Tim
Nixie - 23-5-2008 at 21:07
I used mineral wool between the inner and outer coffee can, and the wool melted 
On the other hand, so did the aluminum, though I can't figure out how to cast it. I just poured it onto some sand. I have some clay but it's not
refractory and it's dried out. Soaking it in water trying to restore it.
There are some stains on the crucible inside but pretty minor. Will probably wash away with HCl.
Also black dots on my hands from flying sparks. Need to find leather gloves.
[Edited on 23-5-2008 by Nixie]
leu - 24-5-2008 at 21:58
A simple internet search found these URL's:
http://www.tms.org/pubs/journals/JOM/9611/Binczewski-9611.ht...
http://en.wikipedia.org/wiki/Sand_casting
http://mypeoplepc.com/members/waygat/castingfoundrylathe/id3...
http://www.backyardmetalcasting.com/
US4480681 discusses some fairly advanced methods of creating complex castings, but it's apparently unavailable from esp@net so you'll have to get it
elsewhere 
Mr Binczewski's solution for fueling the aluminum foundry in such a
challenging environment is rather ingenious 
[Edited on 25-5-2008 by leu]
Nixie - 24-5-2008 at 23:22
Unfortunately, I had no tongs so I used pliers to hold the crucible. So it cracked, and when I was washing it broke apart. At least I managed to
cast the aluminum.
Nixie - 25-5-2008 at 17:27
I'll be shopping for new crucibles soon. Does someone know what chemical resistivity differences there are there between the common types, magnesia,
alumina, and zirconia? They all seem to be about the same in terms of temperature handling, but I'm wondering if it might be useful having more than
one type with some being resistant to chemicals that others are not, and the converse. Also, the cheapest places to buy small ones (under 100 mL).
By the way, I've seen sapphire crucibles, and I'm wondering if single-crystal version of the magnesia and zirconia have been used as well. The
sapphire certainly seems to have a good deal higher temperature rating, and I'm guessing corrosion resistance as well. Does that improvement
generalize to the other oxides?
12AX7 - 25-5-2008 at 20:36
Zirconia. It will waste a maximal amount of your money.
Steel cans are sufficient for aluminum.
Tim
Nixie - 25-5-2008 at 21:13
I wasn't talking about aluminum. I finished what I needed with that. I meant what I can use for the broadest range, not waste maximum money.
12AX7 - 26-5-2008 at 00:15
Well clay-graphite suffices for all general foundry use.
If you're melting something like titanium, magnesia would probably be the way to go, as zirconia and alumina probably react to some extent, changing
the alloy.
Tim
Nixie - 26-5-2008 at 00:19
What temp is that good to before the graphite starts oxidizing significantly? If I build an arc furnace, there will be hot spots significantly higher
than the overall crucible temperature I'd be trying to reach.
[Edited on 26-5-2008 by Nixie]
12AX7 - 26-5-2008 at 02:23
The graphite burns off the outer surface as intended. This is why they are suitable for aluminum, bronze, iron, steel and so on.
Arc furnaces typically don't have crucibles, at least not traditional crucibles. Tell more about your intended use.
Tim
Nixie - 26-5-2008 at 02:34
Casting metals is just a part of this. Among various things, I'm interested in trying out alloys that are resistant to high temperature in air or
plasma. I figured I could cover more ground by making them myself than trying to obtain samples without any certificate. I do realize I probably
won't be able to melt tungsten in any crucible I can afford heh.
I've been working a bit more on the plasma speakers. What I find is that the MHCD hole gets bigger as the metal erodes too quickly (I'm using the Pt
plating baked on W instead of Ta based on that patent that was mentioned around here a while back).
[Edited on 26-5-2008 by Nixie]
12AX7 - 26-5-2008 at 04:03
Well that's easy: tungsten, carbides or water-cooled copper, or any combination thereof. Cheap, easy? Of course not. Note cooling only addresses
melting, vaporization and corrosion to some extent, not sputtering.
Tim
Nixie - 26-5-2008 at 16:29
Tungsten burns. The platinum plated version I used lasts better but still. I plated it with platinum black and an acquaintance baked it in an inert
atmosphere, according to patent 4240878 (but the patent uses tantalum as the substrate, which I didn't have). I think some oxygen diffuses through
the platinum anyway during operation, and the plating might keep diffusing through the substrate due to the heat (probably melts at the hottest area
too, hard to make out without proper magnification). I'm guessing need a non-layered material to avoid this. I thought of conductive ceramics, but
it would have to have high heat conductivity. I heatsinked the electrodes by clamping to copper bars, but since the heat is concentrated at the 1 mm
diameter microhollows, doesn't help that much (my setup is the same three electrode one I described on diyaudio a couple of years ago). Maybe I
should use burned natural gas instead of air, but my whole point for using MHCD instead of simple cathodes is to avoid needing a tank of gas, like the
original Plasmatronic speakers which had a helium tank.
Other than this issue, it does work good though--reproduces sound and I didn't hear noise or distortion, but I didn't measure response or anything.
No idea how to design a housing around it that minimizes diffraction and other problems, and whether should make it dipole or not. I wonder if
there's any simulation software that models the wave equation to enough detail to have a good design without building trial and error. I know HRTF
has been found by simulation from laser scan of head, so it should be doable, but I found no place to download the software described in such papers.
Can always write it from scratch but it's a big project.
The problem of UV seems to have no solution; I don't want to use horns which is the one way they could be blocked without cutting off direct openness
to the air. Just not staring at it I guess.
I don't think NOx and O3 are big issues. I didn't smell much, and I'll just add a vent system to outside.
[Edited on 26-5-2008 by Nixie]
not_important - 26-5-2008 at 23:51
I suspect you're looking at somewhat fancy composite structures if you want to use metals. Copper core with built-in heat pipes, clad in several mm
of a resistant alloy, with that protected by diffusion alloying using molten salt electrolysis to form a gradient from base cladding to surface alloy.
Or a cladding of resistant alloy that forms a very tight oxide coating.
I think you'd do best with a conductive ceramic oxide over a metal base; plasma in air just makes for a lot of active species to much on you
electrodes, why try to fight oxidation when you can start out oxidised. Design the electrode so the is a longer path to the exposed metal that can be
used to carry the discharge until the ceramic is warmed up and takes over.
It would be worth investigating the ceramics used in high temperature fuel cells as a starting point. ZrO2 based ceramics are commonly used. For
this application, where actual oxygen transport is not important, lithium doped nickel oxide and Al-Cr mixed oxides both can be made to be somewhat
conductive at room temperature. Of course ceramics are more difficult to fabricate, and can have thermal expansion issues,
As 12AX7 noted, the plasma environment has many ways of attacking the surface of the electrodes, simple oxidation and melting aren't the only problem.
Here's an example of a high temperature oxidation resistant alloy
Attachment: High Temperature Oxidation behavior of a new NiCrMoSi Alloy.pdf (440kB)
This file has been downloaded 766 times
Nixie - 26-5-2008 at 23:53
I can't seem to be able to download the attachment. If I just click it, it loads the file as a bunch of gibberish text in the browser window. If I
right-click and "Save As" instead, Firefox asks me to save viethread.php and stalls before it reaches the expected 440 KB...
[Edited on 26-5-2008 by Nixie]
Nixie - 27-5-2008 at 00:35
Thanks, it worked when I gave it enough time.