Sciencemadness Discussion Board

Arc melting...

12AX7 - 16-3-2005 at 21:52

I'm not sure if there's already a thread on this, so...*shrug*

I got a pound of gouging rods the other day, and after wasting much time trying to actually gouge with them, I clamped one in the ground clamp, another in the rod handle, set my AC arcwelder for 40-60A and struck an arc. I played it over some firebricks I have and discovered that commercial 2600°F-rated refractory fuses into a very shiny glass quite easily!

I wandered over to some materials I drummed up, fused a mullite concoction to glass (cracking it, since it's a solid ceramic). When I hit my vermiculite brick, which *should* have a rather low melting point (due to the iron in the vermiculite; the composition is vermiculite, sand and clay), it just sat there and only kind of melted on top! WTF! :o :D

Anyways... the real fun with this playing-with-arcs was when I took a broken bit of local dolomite and played the arc over it, fusing it into the surface and whatnot. This is what it looks like now, a beautiful emerald green, probably from iron impurities.



Drool... calcium/magnesium oxide glass...

Tim

Cyrus - 16-3-2005 at 22:21

Yep, very nice. :)

Was the dolomite a solid chunk (if so what was it used for?) or was it bonded with something?

Just curious.

Cyrus (looking, albeit lazily, for some transformer wire for my arc furnace transformer. That project won't be finished for a while.)

12AX7 - 17-3-2005 at 09:08

It was a piece I broke off a chunk of material laying around the basement, which is basically straight out of the ground, minus a hundred years or so. There's a few limestone quarries around here that could get you a few tons of the same thing, though I'm told it's very variable chemically (between limestone per se, dolomite and magnesia).

Mind that calcium and magnesium oxide melt in the range of 2400-2800°C... :D

Tim

uber luminal - 6-4-2005 at 21:17

by request... I'll post some info and pictures later.
My interest and understanding of arc melting has progressed a lot in the last year. A few summers ago I made an imposed arc melter with Cu clad gouging electrodes (graphite), with firebricks and all that. The result was that I was probably using too much power since everything in sight melted. A lower amperage helped, but it was not that spectacular since any metals that you melted would just burn in air. an imposed arc on rocks was kinda fun getting them glassy etc. That and the generation of NOx and COx wasnt really good for the user. I became more and more interested in welding and casting since then and I got a pt job for a Materials Prep Center. my workplace was famous for participating in the Manhattan Project. They developed an efficient process for purifying large quantities of uranium. (little history there). One of my jobs was to measure out all kinds of pure metals (down to .0001 grams). for alloys which researchers had requested, then arc melt them, drop cast them, meltspin them, whatever they wanted. This exposure to practical arc melting caught my interests again, and soon I was building my own closed system, direct arc melters. I have made 3 so far. each better than the last of course. Now as a super nerd, I melt samples on my own time, with my own materials, either looking to find a cool casting alloy or to use for something practical. (or just to do things that the machines at work can't do... like cast pure tungsten)

A closed system has so many advantages over open air. A closed system will run a vacuum, and backfill with an inert gas. I use Argon in my melters. Some use He or Ar. The main reason I use argon is thats what I have for my mig welders. (so I just tap off them). This eliminates your oxygen and nitrogen from the chamber. So when you strike up the arc, you dont have to deal with NOx or O2 screwing up your metals/ metal oxides.

as far as contstuction goes, you need to pick the type of melt you want to make. The traditional 50 year old method is to arc from a water cooled tungsten electrode, to the load which lies on a water cooled copper plate (connected to the ground). Yea, this way works really well. However you must either flip or roll your metal around inside to get a homogmix. Otherwise you get a gradient of heat at the top of the load down to the copper plate which is sucking out the heat.

Another way is to impose the arc, by striking it between 2 electrodes, and dipping it near the material. Then juse guide the meltpool and try to mix the material. again you must flip the sample a bunch to avoide the gradient mix and crystal structures. This works well for non-conductive materials (like ingneous rocks :) ). The other issue is that one of your electrodes will be consumed, or melt. either one. I have considerd using a water cooled copper anode, but sort of defeats the purpose of making heat.
A major problem with this method is you see only about half the heat that you do in a direct method. I will talk more on this later when im not so tired. (*edit* these are not the only 2 ways to arc melt btw. I will add some more after I talk about arc heat a little.)

[Edited on 7-4-2005 by uber luminal]

uber luminal - 12-4-2005 at 00:40

I had planned to upload several pictures of the arc melters I use and have built. However it came to my attention that I could not post pictures on the internet or give the photos to anyone outside of the US. Apparently there are some really stupid exportation rules on US technology.

The equipment photos of lab stuff is obviously forbidden, but I was shocked to learn that the University side, also prevents me from photographing stuff that I built myself! It comes down to me using facilities, tools and information from a source which the technology exportation rules apply. I think its really dumb because other colleges within the same university can issue the same information and photos with no problems. Pfft.

I can however respect that no one wants to deal with shit that might creep up one day. So I will limit the photos to items which do not give away any "technology".

More to come. but for now... These are the W electrodes I use. I threw a quarter in there for scale. These are used ones which need to be machined again. They have rounded off because my old water cooling system did not actualy cool things quickly enough. I added copper connectors such that the W dumps heat into the Cu, and the Cu contacts the water coolant. This works a bit better :).

I took the photo on a march magazine(but I actualy got the photo last week.

electrodes.JPG - 153kB

uber luminal - 12-4-2005 at 00:55

Here is another teaser. (note the lack of "technology" shown). I am starting to melt an AlxCuxSix alloy sample.
I have 4 of them in there, of different coef. Roughly I recall they were Al.9Cu.09Si.01 . Al.8Cu.0xSi.0x etc. @ -15 atm under (Ar).

There is a pit in the center of the cooling plate that I keep a chemical for extracting left over oxygen or indicating that there is O2 in the chamber. (and then either start the evacuation/backfill process over, or just evacuate for another 5 min and back fill to try again)

a welding power supply is used, with a High Freq. start for the arc. Otherwise you practicaly need to strike the arc, and thats no fun when your dex is close to nill.

meltin.JPG - 50kB

12AX7 - 12-4-2005 at 06:57

Neato!

- Why bother arc melting something which is easy to cook up with just charcoal?
What's the hottest metal you've melted?

BTW I predict your alloy to be brittle. Much over 10% Cu (don't give decimal moles, that's uberlame) tends not to exist = too brittle for use...

Tim

uber luminal - 12-4-2005 at 08:12

Quote:
Originally posted by 12AX7
Neato!

- Why bother arc melting something which is easy to cook up with just charcoal?
What's the hottest metal you've melted?

hmm... lets see, temp (surface reaches upwards of 6000 C. depending on the material being hit with the arc), Ease of use (takes 5-15 min to prep a sample, and 2-10 min to melt it. (depending on the materials), and the purity of the melt is so much better. Oh yea, and lets not forget that you can get a product... not a chunk of embrittled mass by Nitrogen, Hydrogen or Oxygen.

Granted you could take an Inconel or Ta tube, place the metals inside and throw it into a "charcoal" fire. But thats a lot of work since you must weld the Ta and the Inconel under an inert environment, and you waste the tubes unless you actualy use Ta as the containment tube (via LME or surface cracking) you will lose the vessel. Don't get me wrong, there are limitations to an arc melting system... Metals with high vapor pressure are difficult to melt in neg pressures. The issue of putting more pressure into the chamber about 5 atm, is you have to start calculating how much time you have the heat on, how much heat can be carried out of the chamber per second and, what the vapor pressure adds, with a variation of the ideal gas law. Oh yea, all that against the max pressure before a seal blows. Sometimes you have to throw together a Ta tube, place that inside another tube (under Argon) and put that into a furnace. you have a lot of loss this way though, and again you have to scrap your Ta and inconel when your done.

the OTHER problem is size. The arc melter in the pictures can take loads up to 100g of a mid density material (like 10). If you go bigger you run into cooling problems. There are ... alternatives to this. I took a picture of our Plasma melter at work, but it falls into the technology restriction. The electrode is 4 inches in diameter :).
It can do bigger pucks.

Quote:
BTW I predict your alloy to be brittle. Much over 10% Cu (don't give decimal moles, that's uberlame) tends not to exist = too brittle for use...
Tim


Your right! it will be some what brittle, but not so much due to the Cu, and not too brittle to use. Si (in small amounts) does interesting things for Cu.

Tends not to exsist huh? happens in nature(roughly), happens in research happens in special *technology removed*. Its not a common alloy, but its more common than you think. Consider that not every alloy is about making a strong metal. For example, Today I am working on a Ge-Si alloy. talk about brittle. Or the last few weeks I have been working with Al-Ge. not made for hardness or strength.

Alloyx-Alloyx IS percent. if you want to convert it to weight percent by looking up atomic weights, go for it.
If you list it as 50%Al 50%Cu, do you mean it has half copper half Al? or 70.1952% Cu and 29.8048% Al? or do you mean something else? AxBxCx is not only exact, but it takes up far less space, far less time writing it.

Fin. The 'hottest' metal I have melted. This goes two ways. you must consider specific heat, thermal conductivity, and melting point. For example, I have melted and cast 99.99% tungsten into the shape of a heart. (for my GF last valentines day). this took quite a bit of current to melt it in the cooling time bracket. (600 amps). Pure Cu however. Not gonna happen. even if you turn off the cooling system, you just can't melt the copper if its near the hearth. If you take a small diameter rod and stand it up, yea molten cu is pretty. but if you are trying to do a sample that is Cu.9, its either not going to happen, or its going to melt onto the hearth.

More stuff

uber luminal - 12-4-2005 at 13:27

The pic for today is the chamber view for one of the smaller melters. It goes with the AlCuSi samples I was making. You can see the Cu and Al pieces in there. Si tends to pop apart while heating, so I tuck metals like that and powders in the back of the Al. Al is cool because you can roll it down and make shapes to hold other materials. The Cu sits on the side of the samples. In this case, I will start melting the Cu 1st, and try to fold the Al shell down ontop of the Si. Melt it good, try to roll it, if it doesnt roll, let it cool and flip it over with the electrode, melt it again. (since you are under a vac, you cant take it out, so you have to learn to use the electrodes.)*edit* by roll I mean heat the sample and try to generate a heat path. (use the plasma to push the sample around, and hopfully it will ball up and roll around.)

Is anyone interested in this stuff? I could ramble on an on. anything specific you want to know maybe?

[Edited on 12-4-2005 by uber luminal]

prep.JPG - 81kB

Quince - 7-5-2005 at 00:40

There's something I wanna know. How the hell do I make a crucible that I can arc-melt alumina in (2072*C)? It must not just withstand the temperature, but also not react with the molten alumina. I was thinking graphite, but I can't find any around here, and blocks are very expensive on eBay. Likewise for tungsten and molybdenum, as I can't find the right shape blocks.

[Edited on 7-5-2005 by Quince]

neutrino - 7-5-2005 at 02:49

I doubt this will work, but I wanted to throw this idea out here. There was some mention of making wood into graphite somewhere on this board. I doubt it is such good graphite, though.

12AX7 - 7-5-2005 at 20:49

Graphite is cheap, I've got some bars in the mail, I'll tell you if I can melt any in them when they arrive. But I'm still telling you melted won't work.

Any form of carbon can turn into graphite, it just needs to be [re]crystallized as such. I would imagine you can burn charcoal at yellow to white heat for a while to do this, though it does nothing for the pores (to fix that, you could grind it to powder and mull with pitch or tar, then fire in an arc furnace. AFAIK, that's essentially artificial graphite in a nutshell.)

Tim

[Edited on 8-5-2005 by 12AX7]

Quince - 8-5-2005 at 03:06

Well, why wouldn't it work? Even if one can't controll the cooling well enough to get a single crystal, a multicrystal sapphire would still be better than the alumina ceramic.

neutrino - 8-5-2005 at 05:59

One would think that the carbon left from the wood would be too porus and weak to take regular use, like plaster after being dehydrated. Tim’s idea sounds good, though.

Mr. Wizard - 8-5-2005 at 07:05

I read an old book somewhere where the porous wod charcoal could be mixed with sugar syrup , formed into rods and reheated to carbonize them, and then after cooling could be re-impregnated in sugar water again and the process repeated until they were non porous. This would be best accomplished under an inert atmosphere in theory, or a non oxidizing one at the least. I would suggest putting the carbonized rods under a sugar solution and pulling a vacuum on the solution to help expel the trapped air and to help the solution enter the rods.

uber luminal's posts are very interesting. Thanks for sharing what you can. Does uber luminal = too bright?

I found that a 50/50 mix of Copper and Aluminum made a very brittle silvery alloy that could be crushed easily, with a hammer, and made a silvery dust that would coat your fingers. I melted the copper wire on a brick with an acetylene torch and then added the aluminum wire. I didn't get a good mix, so I'm not sure of the exact alloy mix. The stuff looked very interesting, but I don't know what good it would be.

12AX7 - 8-5-2005 at 09:19

Only use I have for theta phase is master alloy (to add copper to aluminum alloys). :P
(50/50 Cu/Al by weight is almost exactly the theta (CuAl2) intermetallic; hard, very brittle and melts slightly lower (601°C) than pure aluminum.)

Magnalium is similarly a brittle intermetallic (mostly gamma, with some Mg2Al3 if richer in Al) and quite shiny, with the added bonus that it's much more reactive. :D
30% Sn, 70% Cu is "speculum metal" because it's hard (for the same intermetallic reason), white and takes a high polish.

Back on topic (er, what topic was that anyway?), I wonder how much carbon sugar will leave after each treatment and carburization. Eh, worth a shot, I guess...
Say, anyone know a good solvent for pitch or tar that could get them in there? Anything soluble and rich in carbon...(napthalene maybe?)

Tim

uber luminal - 8-5-2005 at 10:09

fantastic. someone interested :) Prepare for rambling!

is that 50/50 as w% or volume%? I think weight% is easiest to use. So by volume that would have been about 70% Copper and 30% aluminum if not already converted.

but either way, yea, most mixtures of Cu in Al above Al8-Cu11 yield a very interesting luster of silver which is also very brittle. I have never tried to make a Cu9Al1 mix before, but I predict it would also be the same silver luster with tons of embrittlment.

Most people think of alloys as just the mixture of a few metals, but do not account for the more important processing. you will have a whole range of properties with a series of alloys at a certain arrangement of the crystals. Increase crystal sizes and you have a whole new range of properties that could not be seen before. Decrease crystal sizes... same thing... force crystals together, allow them to grow at their own pace for several weeks in an oven... (huge formations), or how about getting rid of crystal formation...

Like the alloys of... er... a certain range, may not be of much use for strength with its current crystal formation, it does something amazing when it is reheated and quickly cooled down, forming an amorphous structure. You would not think that taking a very brittle alloy and making it into glass would make any sense. The elasticity increases dramatically with an increase of Si or Ge, or strength increases with the increase of Al to Cu. Its cool stuff. The glassy Al/Cu alloys also take on excellent thermal properties, as well as something cool to study.

The advantage of trying things is that you sometimes find interesting stuff on accident. Last week I made another series of AlCuGe alloys. I usualy snap them in half, polish one side for a microscope and use the other side for hardness testing and a few other tests if its warrented. 2 of them I could not break. I placed them in a vice and used a hammer to quick shear them. the 1st one broke. the 2nd didnt. I hammered away at the stupid thing with different objects trying to achieve a quick shear, but nothing would break it. I tried to press it and roll it with some success. It acts very much like hafnium does, except its not as dense and has a lower MP. An interesting use for Hf is for Ballistics. I wondered if my AlCuGe alloy could compete with it, since its less costly.
Granted properties of alloys can be predicted and are often documented. What I "discovered" was nothing new since I found it documented on SciFinder. But I certainly would never have known about it otherwise. Sometimes having the keys laying around is better than finding locks and looking for a specific key to fit it.

Cyrus - 8-5-2005 at 14:21

"I read an old book somewhere where the porous wod charcoal could be mixed with sugar syrup , formed into rods and reheated to carbonize them, and then after cooling could be re-impregnated in sugar water again and the process repeated until they were non porous. This would be best accomplished under an inert atmosphere in theory, or a non oxidizing one at the least. "

LOL! ;) Are you sure it wasn't me who said that? (That's pretty much what I was saying in the crucibles and metalcasting thread) I never got it to work really well, though I did get nice smooth sheets of carbon with a porous backing. (I mixed sugar and grog, alumina, etc, with a little water, then made a syrup with it on the rangetop, then poured it into an Al foil mold, then heated it with a propane torch. It burned, foamed, and dripped, but when the Al foil was pulled off, a smooth layer of C was left. (It wasn't graphite, because I didn't heat it hot enough)

12AX7 - 8-5-2005 at 15:13

Quote:
Originally posted by uber luminal
is that 50/50 as w% or volume%? I think weight% is easiest to use. So by volume that would have been about 70% Copper and 30% aluminum if not already converted.


Yes weight (it's standard anyway, unless you weren't talking to me). More like 70%vol Al BTW, looks kind of odd melting a small lump of copper then mushing in this big wad of aluminum... ;)

Quote:
I have never tried to make a Cu9Al1 mix before, but I predict it would also be the same silver luster with tons of embrittlment.


I mixed a pound ingot of C630 which is 5% Ni, 3% Fe (grain refiner I believe; I didn't get it to dissolve though :( ) and 10% Al (er, might be 12, I need to check), balance copper. It's a good light gold color, and quite tough. :D

Quote:
Most people think of alloys as just the mixture of a few metals, but do not account for the more important processing. you will have a whole range of properties with a series of alloys at a certain arrangement of the crystals.


Well, in terms of strength and ductility anyway. Strength tends to go up as grain boundaries go up, so very fine structures (like a good carbon steel; ever broken a file and looked at the cross section?) are very strong, while coarse structures tend to be weak (pennium (US pennies >1982) for instance). The bulk strength of the structure matters of course (aluminum is weaker than steel, period), as does the boundary strength, but in general, within a class, it's true. (Overheating a carbon steel blank, causing grain growth before heat treatment, is suicide for the properties...)

Which reminds me, anyone have suggestions to make a 1% Ti 2% B balance aluminum master alloy? (TiB2 particles precipitate on cooling, providing nucleation sites. Zirconium too.)
I have the Ti and may have the B, but getting them into Al is going to be a bitch.. :D

Quote:
...or how about getting rid of crystal formation...

Like the alloys of... er... a certain range, may not be of much use for strength with its current crystal formation, it does something amazing when it is reheated and quickly cooled down, forming an amorphous structure.


Ah, glass materials. A whole other class unto themselves...

Quote:
The glassy Al/Cu alloys also take on excellent thermal properties, as well as something cool to study.


How fast do you have to cool that, and where would you get Ge anyway (besides a university supply which you no doubt have access to :P )?

Tim

[Edited on 8-5-2005 by 12AX7]

Mr. Wizard - 9-5-2005 at 07:50

Cyrus, I did see your post about the charcoal and sugar, but I saw the method a long time ago somewhere else, without using grog or alumina. I actually wasted a half an hour last night trying to locate where I saw it, with no luck. I used to remember things and where I read them, but now it's just the things. I have too many books in too many places. On the CuAl 50/50 alloy, it was very crystalline. I'm going to try making some more with an argon atmosphere, using some carbon gouging rods. The idea of making a metallic glass by quick cooling sound fun. I need to keep that camera ready.

12AX7 - 9-5-2005 at 08:24

Yeah, my AlCu has long, thin crystals running across the surface :)

No need for argon or carbon arcs, aluminum and copper are more than satisfactorially melted by propane, unless you don't have a burner.

Tim

Quince - 9-5-2005 at 08:35

12AX7, never mind the melting alumina. I got sapphire on eBay, and also the Saint-Gobain Crystals are gonna send me some for free.

Heh, I never managed to get free samples of stuff when I was at DIYaudio. I'd have thought a free opamp would have been easier to get than sapphire...

[Edited on 9-5-2005 by Quince]