Sciencemadness Discussion Board

Melting Magnesium

MrHomeScientist - 23-7-2013 at 12:46

I'm a big fan of molten metals (not quite sure why...) and recently cobbled together a very rudimentary furnace capable of melting aluminum. I want to melt several different metals down and cast them into identical size and shape bars, to use as a nice presentation piece to show the physical differences between them.

Anyways, one of these metals is magnesium. It melts at 650 C (lower than Al) so I should in theory be able to cast it. But, as we all know, Mg is quite flammable. Several MSDS's list the autoignition temperature of Mg ribbon at around 500 C. If I were to try to melt several small chunks of Mg metal, broken off of a firestarter block, are they likely to catch fire? I know that ribbon is much more flammable than the bulk metal, so I was wondering if there would be a difference here. Any thoughts are appreciated.

EDIT: Actually, if i recall correctly, the reason that bulk Mg is much harder to ignite is because the metal acts as a heat sink and thus takes much longer to heat to ignition temperature. If I'm melting it, that would certainly take the whole mass above this temperature and may doom this endeavor :( Would I need an inert atmosphere for such an operation? Probably argon - can't use nitrogen for this one!

[Edited on 7-23-2013 by MrHomeScientist]

Dr.Bob - 23-7-2013 at 12:53

It is pretty easy to light Mg, even bulk, and I would guess that trying to melt will almost certainly cause ignition. But if you want to try it, do it outside, away from flammables, and wear cotton clothes. But if it catches, don't bother trying to put it out, just back away. You could put some in a crucible and try in that, but it will not be easy to keep it from burning unless you can shield it from air. Argon or maybe helium, but either is pricey.

bfesser - 23-7-2013 at 13:08

As an alternative to inert atmosphere, have you considered using a flux? I'm sorry to say that I can't think of a suitable flux at the moment, but I will continue searching (thinking along the lines of MgCl/KCl). I had a similar dilemma to yours, some years ago, while trying to <a href="viewthread.php?tid=11946#pid149155">fuse Ca</a> granules.

[Edited on 7/23/13 by bfesser]

12AX7 - 23-7-2013 at 19:04

NaCl-KCl eutectic (around equal parts by weight) is pretty close, and some CaCl2 added to that (not quite 1/3 (of total) by weight, I forget just how much though; alas, the molten salt eutectic finder database has long since disappeared from the internets, so you'll have to search to find the exact composition) is better still.

I suggest pre-melting the flux, as fresh crystals are full of moisture and explode like popcorn, and unless there is intimate contact, the combination won't melt at the eutectic temperature. Melt in a steel vessel, cast bars or plates, break into manageable size pieces, and store away from moisture.

With magnesium, it might help to have a "heel" of molten salt, rather than adding it later.

Beware that steam and hydrogen have excellent solubility in molten salt, which tends to result in poor quality castings. I don't recall if magnesium really has to be degassed, but aluminum certainly has some affinity for hydrogen, and my experience has always been that the melt dissolves quite a bit of hydrogen when an excessive cover flux is used. Calcium hypochlorite or TCCA (pool chlorine) can be used as a degassing agent, for aluminum at least; it must be administered to the bottom of the melt, inside something that helps form small bubbles. Chlorine reacts preferentially with hydrogen over aluminum, removing it from the melt.

Besides inert gas (argon -- not nitrogen or CO2), it's my understanding that SO2 is a suitable cover. SO2 can be provided in-situ by adding a blob of elemental sulfur.

Tim

Gearhead_Shem_Tov - 23-7-2013 at 21:51

You'll definitely need to use argon. It's a minimum three-person job: one to handle the argon wand, one to handle the crucible, and one with a shovel ready to dump pre-dried silica sand on top of all the inevitable Mg flare-ups you'll have. It would probably be best to have two more additional people: one to video the event and another well away from the action ready with a mobile phone to call for an ambulance and a heavy blanket to put over anyone on fire until the ambulance comes.

-Bobby

DrSchnufflez - 24-7-2013 at 00:37

I remember reading about magnalium somewhere that you can use a calcium chloride flux to melt magnesium under before adding the alminium.
I also remember that you can use sulphur powder to stop the magnesium burning as the magnesium sulphide formed doesn't foam up as magnesium oxide and nitride are purported to.

Memory verified Magnalium sulphur however it was for making magnalium.

<!-- bfesser_edit_tag -->[<a href="u2u.php?action=send&username=bfesser">bfesser</a>: merged split posts, sorry]

[Edited on 7/24/13 by bfesser]

[Edited on 24-7-2013 by DrSchnufflez]

bfesser - 24-7-2013 at 05:14

Your suggestion of using S to stop a Mg fire is dangerously incorrect Next time, verify your memory before posting.

[Edited on 7/24/13 by bfesser]

plante1999 - 24-7-2013 at 05:15

powdered carbon does work.

bfesser - 24-7-2013 at 05:28

Could you please explain in a little more detail, plante1999?

[Edited on 7/24/13 by bfesser]

BromicAcid - 24-7-2013 at 05:28

Remember that in industry it's such a pain in the butt to prevent oxidation of molten magnesium and the like that they use precious sulfur hexafluoride as the inert gas in the system instead of argon.

MrHomeScientist - 24-7-2013 at 05:46

Thanks for all the replies! I like the idea of a cover flux - that would be a lot easier to manage than an inert blanket (for me at least).

Quote: Originally posted by 12AX7  
With magnesium, it might help to have a "heel" of molten salt, rather than adding it later.

Beware that steam and hydrogen have excellent solubility in molten salt, which tends to result in poor quality castings.


Could you elaborate on this a bit? What exactly is a "heel" of molten salt? How would one set up such a system? My first thought upon reading was: chunks of pre-melted eutectic mixed with chunks of Mg metal, then upon heating all would melt and the salts would float to the top. Hopefully the eutectic melts first, before the Mg ignites!
Quote: Originally posted by 12AX7  
Calcium hypochlorite or TCCA (pool chlorine) can be used as a degassing agent, for aluminum at least; it must be administered to the bottom of the melt, inside something that helps form small bubbles. Chlorine reacts preferentially with hydrogen over aluminum, removing it from the melt.


Similar question here. Would you simply load the crucible with a layer of calcium hypochlorite at the bottom and aluminum bits on top? What are the consequences of hydrogen infiltration of metals? I seem to recall it makes them much more brittle.

12AX7 - 24-7-2013 at 21:36

Quote: Originally posted by bfesser  
Your suggestion of using S to stop a Mg fire is dangerously incorrect Next time, verify your memory before posting.

[Edited on 7/24/13 by bfesser]


Well, yeah, when you go out of your way to make flash powder, you're gonna get flash powder...

Here's a cite for use of SO2. Seems odd (MgO and MgS are favorable reactions), but apparently it works.
http://www.epa.gov/magnesium-sf6/documents/tms_paper.pdf
Elemental sulfur bubbling away on top of a melt doesn't seem like a very good idea, but I've seen a picture or two of this use. May not be able to find a cite, which might suggest it's a more dangerous method, but I know it's at least possible.
Quote: Originally posted by MrHomeScientist  

Could you elaborate on this a bit? What exactly is a "heel" of molten salt?

Ok... you'll want to familiarize yourself with foundry terminology and practice before doing this. "Heel" just means, liquid in the crucible, in this case, some molten salt, to which you add clean metal. The metal is heated by, and melts under, the cover, so it doesn't have a chance to catch fire.

Melting chunks of both at the same time will probably lead to the salt eutectic melting first and covering the metal, but if the ignition temperature of the alloy is particularly low for some reason, you'll have a bit of a mess.
Quote:
Similar question here. Would you simply load the crucible with a layer of calcium hypochlorite at the bottom and aluminum bits on top?

No, you need a degassing tool, which ensures that, as the compound decomposes, the gasses produced form small bubbles through the melt. The degassing tool is loaded with a charge and thrust into the melt; it decomposes vigorously but not explosively or anything. (At least, when done right -- always be on guard for conditions that may react explosively, like stray moisture on tools or molds. And that includes chemically combined moisture in rust, etc.)
Quote:
What are the consequences of hydrogen infiltration of metals? I seem to recall it makes them much more brittle.

Yes, there is hydrogen embrittlement in many metals, but more immediate is the effect of freezing a gas solution. Consider what happens when a soda can gets frozen... all the CO2 comes out of solution. Very gassy metal will expand on freezing, leading to a "muffin topped" shape, and structurally unsound metal that is full of bubbles. (Interestingly, copper is prone to dissolving molecular oxygen, independent of actual oxidation. A powdered charcoal cover is used to treat this.)

Tim

MrHomeScientist - 25-7-2013 at 07:32

Tim,

That's excellent information, thanks a lot. As with most things, there's a lot more to this than I first thought! I'll have to practice melting the salt eutectic mix alone, then salt plus a different metal (Al or a lower melting point metal), before progressing onto Mg. I'd melt the salt first and then add the metal, as you suggested. Probably needless to say, but when melting the Mg I'd have to do that in the mold rather than a crucible since pouring (and thus exposing to oxygen) would be out of the question.

That's very interesting about hydrogen being trapped in freezing metal. I'll have to cut one of my future castings in half and see what kind of bubbles are present inside. So far all my aluminum blobs/ingots have shrunk after cooling, which is what I expected from thermal contraction.

I'm really going to need a better furnace to get serious with all this. The very rudimentary charcoal chimney I'm using loses a lot of heat, and the fact that the crucible sits atop crumbling coals is no good for doing magnesium (wouldn't want it to tip over and ignite!).

Fantasma4500 - 26-7-2013 at 06:08

im starting to understand that you have more than just ALOT of knowledge on how to treat metals as in melting them
infact i have a large rod of magnesium which will be cut, added to a can (when i someday happen to get a hold of something that doesnt blow up when heated, lol)
as plante1999 said carbon can be used, to my knowledge this is used also when milling MgAl in which is one of the most dangerous parts of making magnalium, aswell as magnesium powder, in order to coat surfaces so they cannot react with oxygen
carbon is suitable as it is easy to get a hold of, and by my understanding just doesnt blend into metals just like that (industrially graphite is used as carbon..?)

so instead of salt (NaCl MP=801*C)

on a sidenot with the whole salt 'heel' thing, i cant seem to imagine magnesium metal not reacting in liquid state with NaCl to form MgCl2? im not keen on reactivity on different compounds, i could be wrong, as it would be an other way of making Na metal, in which i havent heard about yet...
just seems too weird that an such reactive material would not react, not saying that this cannot be done, just a glitch in my mind i suppose

watson.fawkes - 26-7-2013 at 06:32

Quote: Originally posted by 12AX7  
Yes, there is hydrogen embrittlement in many metals, but more immediate is the effect of freezing a gas solution.
Hydrogen embrittlement is of particular importance for ferrous alloys, which is why there's a whole class of high-nickel alloys used in, say, gas desulfurization plants.

Elaborating on what 12AX7 said, I want to point out that, practically, the main source of dissolved hydrogen in melts comes from moisture. Even if the charge looks dry (and won't explode), adsorbed water can be significant, particularly for charges with high surface area (turnings, soda cans). The oxygen will form oxides that become part of the slag, and the hydrogen will go into solution. Dealing with adsorbed water is rather easy: just preheat the charge to drive off water and add it to the crucible hot.

AndersHoveland - 26-7-2013 at 06:54

An aluminum alloy with 57% (by weight) magnesium content melts at 437 ºC. So if making a magnalium alloy is acceptable, melting is much easier if scraps of both aluminum and magnesium are heated together.

http://www.engineeringtoolbox.com/aluminum-alloys-melting-po...

[Edited on 26-7-2013 by AndersHoveland]

Fantasma4500 - 27-7-2013 at 08:39

true, anders.. but then you would need the metal to melt together at first
when you first have a blob of molten metal its very easy to proceed as the surface of contact with the hot stuff is very large
looking into it seemingly flux will remove this effect and remove the oxide layer to make it melt straight away
i suppose cryolite would work for this, as i can get this super cheaply (:

Pok - 27-7-2013 at 14:25

Forget about ideas with eutectic salts. Remember the low density of Mg! It's 1.584 g/cm³ in the liquid state according to wikipedia. Mg will always float on top and not the salt! To avoid reaction of Mg with NaCl (or other salts), MgCl2 could be used. But this is too dense as well.

Charcoal might be good, as plante 1999 said.

A molten 50:50 mix of Al and Mg does not ignite if it's not heated much higher than the eutectic melting point.

[Edited on 27-7-2013 by Pok]

AndersHoveland - 27-7-2013 at 14:33

Quote: Originally posted by Pok  
To avoid reaction of Mg with NaCl (or other salts), MgCl2 could be used. But this is too dense as well.

Magnesium will not react with NaCl.
Mg and Al have more of an affinity for oxygen, they will not reduce the chloride salts of the alkali elements.