I'd say melt the Al, add a flux cover, and then dissolve the Mg into the Al
most safely, in chunks not overflowing, to your rough 50:50. And have dry sand
available at the very least. Preferable ground coal to soak up oxygen should
you experience ignition. It's a delicate balance between heat input, molten
temperature, and time spent dissolving MG into Al.

The more time at idle, the more Zn will fume out of your Al, if there's any.
Zinc chills. Brr. And other compositional elements will boil out, but mostly
ingot Al is pretty stable.

Then, as it cools, you stir it as it freezes to break it up, like making ice
cream. You get a coarse powder that way.

Don't stir it vigorously while it's real hot!

Bah. I have a photo from a friend here showing a pool of shiny metal
inside a crucible inside a furnace, written on the back is "1.5 lbs of
freshly skimmed MagNESIUM (4Al) @ 1250°F" (his capitalization :P ).

And when he poured it, did it go actinic?

The secret commercially is either a gas or flux cover. The latter
would be a salt 'alloy', such as equal parts sodium and potassium
chloride (can be had at the grocery store as "lite salt"; has to be
fused (mmm, salt ingots!) before use), with other select ingredients
to lower the melting point and improve fluxing action.

Love the stuff. Love frozen dinners, too, but what a wallop of sodium.

Maybe try 1/4 your smallest crucible one day, then upgrade....

I tolerance everything and tolerate everyone.
I love: Dona, Jeff, Kim, Kimmie, Mom, Neelix, Tasha, and Teri, alphabetically.
I drive: A double-step Thunderbolt with 657% range.
I fight terrorism by: Using less gasoline.