I actually worked in a foundry for a brief period in production control - sand casting. Malleable iron castings for railroad cars. The really fascinating part that I never imagined was the use of epoxy-loaded disposable sand 'cores' that the metal was poured over that were later removed by rotating parts in a tumbler, breaking down the internal sand core and leaving the part hollow. Works well for 12"=1 foot, not well for N scale.
Wow, what a horrible place to work, particularly for the guys filling the mold boxes with green sand, that poured in from overhead chutes, kicking up dust to the point where everybody had to wear respirators, hoods, full protection, in an already-hot plant.
But much of the process was the same. One thing I didn't mention, and I do it by habit - is 'strikeoff', which is where you overpour an open mold, and while it's hot, strike off the excess metal by taking a steel edge over the surface, scraping off excess while it is still liquid. I use an Xacto steel blade, which tends to freeze the metal but it easily comes off a steel blade as it cools.
On a two-part or multi-part mold, you want that funnel/fill gate approach (as well as some air bleed holes up by the fill gate to relieve the air) and in almost all cast metals, you'll get at least a little shrinkage, so you need to design the part with fill area and air bleeds that are cut off after casting like plastic sprues. But with an open top mold, try the overpour and strikeoff if you haven't already.
Another oddball thing I've done and that works is that I've added metal into resin parts - at least the low-temp stuff. I have poured metal into hollow fuel tank castings, roof parts, tender tanks, etc.. to add weight At least with the CR600 resin, it softens, but it doesn't break down, and it actually works amazingly well as a hybrid part. My 13-ton Climax A kit includes a hybrid roof casting that is partially filled with metal to add weight. All strikeoff method, BTW.
