Using 3D Printing for Metal Casting: Part 1

We have many customers ask us about 3D printing in metal and how they can do it. With most printers, the closest you can get to 3D printing metal objects is using a "-fill" material; copperfill, bronzefill, steelfill, etc. But these are all composites and only partially metal, with most of the material being PLA plastic. To truly print 100% metal is not an easy endeavor, and requires a $100,000 printer. However, there are other ways to take a 3D model and end with a metal part in the same shape without 3D printing in metal: casting!

Preparation:

1. Build a foundry. There are many different ways to do this, some will last a long time and work extremely well but cost more to startup, others are cheap and will last only a couple pours, but get you a taste of the process.
   1. I wanted to see how well we could get things going before we commit too much to it, so I went with The King of Random's original mini foundry with the improvements he later made to it.
   2. To sum it up quickly, the foundry itself is plaster, sand, steel wool, all mixed together to insulate the steel bucket to reach a (relatively) high heat, and then a hole cut in the side of the bucket and plaster using a hole saw and aimed at a downward angle.
   3. The lid was made in the same way, with some U-bolts embedded into and a hole drilled into it while it was still curing.
   4. What kicks this up from a campfire to a foundry is a hair dryer and some pipe. This works much like bellows in traditional blacksmithing; the introduction of air dramatically raises the temperature in the foundry, just enough to be able to melt low melting point metals like aluminum.
2. Gather equipment. There are many professional, high-quality tongs, ladles, and gear to make metal casting easy and safe. They're really nice, but for quick and introductory casting, kind of overkill.
   1. Crucible: what you melt metal in. If you're basing this build off the King of Random's build, use a 2kg clay-graphite crucible. It's just small enough to fit inside his foundry but can still fit a soda can. You can also use a used steel fire extinguisher cut in half or even a steel pipe welded watertight to a steel plate and trimmed. Be sure to properly bake the clay graphite crucible, otherwise it can crack, and make sure if you make it out of steel that it's thick enough and won't melt through.
   2. Tongs: These grip the crucible to pull it out of the foundry and to hold it while you pour from it. Like I previously mentioned, you can buy professional, several hundred dollar ones or make your own, but do not buy a couple kitchen tongs. This is 1000 degree molten metal that you're relying on some cheap tongs to carry. I did it, and I will not be doing it again. It's just too risky.
   3. Safety equipment: specifically heat-resistant. A face shield, a set of heat resistant gloves, and common sense and caution. Move slowly, move deliberately, and gently pour your 1,000 degree molten metal. Get a working fire extinguisher just to be safe, not a liquid one either so you don't splash molten metal everywhere.
   4. Ingot mold: as simple as a muffin tin or any sort of steel baking form, the smaller the mold shape the better. Too big and it won't fit in your crucible, so don't go for the jumbo donut tin. On your first pour, the teflon coating will vaporize and catch fire; don't breathe it and keep it on a paver or a bed of sand in case it happens to melt through.
   5. Borax: Just normal laundry borax. By throwing a little bit of this into the crucible, it'll stick to all the gunk known as "dross" that is the impurities in the aluminum, either the ink from soda cans or any coatings to keep the soda from eating the aluminum. Metals are dirty and the borax cleans it out. Fashin a small pouch out of paper and toss it in to prevent a mess of borax all over your crucible.

Printing 3D Models for Casting

There are many many many lessons and techniques used when developing a mold for casting. Things like gates, sprues, risers, and runners. So either pick simple models with flat backs to cast, or take the plunge in learning how to properly "gate." Gates are, essentially, support material for casting. With 3D printing, you need support material for steep overhangs, with casting, you need support (risers) to support the points of the model that are parallel with the bed, or are so steep they are past parallel with the bed. You need gates anywhere on your model that is steep enough, otherwise those spots will capture air bubbles and you'll have a poor pour. Add these gates before your print or add them after with wax designed for gating.

Making a Mold

There is more than one way to skin a cat, and there are dozens of techniques for making molds for metal casting. I'm going to discuss the two that I attempted: Open-Face and Investment Casting.

Open-Face Mold Making

This is the simplest mold you can make, whether your mold material is made of high-grade silicone rubber or beach sand. The idea is your part has a flat back and you pour and pack in the mold material over the top of it. You then remove the part or melt it out and are left with a negative in the shape of your original 3D print (or master). In my attempt, I did the following:

1. I 3D printed the House Lannister Sigil since it was a model with a flat back.
2. I hot glued it to the bottom of a tupperware container to prevent it from floating away once I make the mold. Since the infill lowers its density, it will float away if I don't glue it.
3. Since I knew the volume of the tupperware, and the model wouldn't take up much space, I filled a bucket with enough water to half fill the tupperware.
4. To mix Plaster of Paris into water, and to accuratelly measure it out, you want to gently sift it into the water until you get a milkshake-like consistency. I sifted until the plaster stopped soaking in and started forming a small island in the center of the bucket, then mixed it all together.
5. Once it was thoroughly mixed, I gently poured the Plaster of Paris mix over the top of the print to try and get a coat without air bubbles.
6. With all the plaster poured in, I let the mold sit in the sun for a week to dry out and get as much water as possible out of it.

Investment Casting

This is more difficult to pull off, especially with materials not specifically designed for it. You take your part, gate it, add sprues and runners, then melt out the part. There's a lot of science that goes into making a part that can be molded and cast into that won't have voids and air pockets. In this attempt, I was just looking for something that would work, so I found a model that didn't need support, and therefore, would be an easy pour without any voids. For this method I tried:

1. I 3D printed the Iron Throne since it was a model with a flat bottom and no (big) overhangs.
2. I hot glued it to a piece of carboard, then glued a cup with the bottom cut off around it so I could pour into it.Just like the other mold, I glued it to prevent it from floating away once I make the mold. Since the infill lowers its density, it will float away if I don't glue it.
3. Since I knew the volume of the cup, I filled a bucket with enough water to fill the cup.
4. To mix Plaster of Paris into water, and to accuratelly measure it out, you want to gently sift it into the water until you get a milkshake-like consistency. I sifted until the plaster stopped soaking in and started forming a small island in the center of the bucket, then mixed it all together.
5. Once it was thoroughly mixed, I gently poured the Plaster of Paris mix over the top of the print to try and get a coat without air bubbles.
6. With all the plaster poured in, I let the mold sit in the sun for a week to dry out and get as much water as possible out of it.

So at this point, I had two molds that seemed pretty promising for getting some castings out of them, but first I had to do something to get the prints out of the molds: The Burnout.

After letting the molds sit out for a week, I needed to remove the prints before I could do any casting. To do that, I needed to fire up the foundry.

The Burnout

Normally, with lost wax investment casting, you melt the wax out and let the mold bake over several hours. The wax vaporizes and leaves a negative space of the original design. With PLA though, I found that either A: it doesn't vaporize, or B: the foundry doesn't get hot enough to vaporize it. The PLA just sort of melted out the bottom, or when the heat couldn't travel further up into the mold, it left the base of the 3D print sticking out the bottom soft and pliable, but the top of it (inside the plaster) as rigid as a finished print. The plaster became extremely brittle after heating it too. The Iron Throne still had the back of it stuck in the plaster, then the whole mold shattered. I was lucky enough that I could lean the Lannister Sigil sideways and the PLA could pour out. On the next attempt, maybe ABS would work better since it catches fine instead of just melting, and would burn out better that way, in addition to getting proper investment materials.

The Melt

Now that I had one usable mold, I needed the rest of this project to work. To prepare for casting I needed to do a couple things first:

1. I loaded the foundry with coals, not just charcoal briquettes, but hardwood charcoal. Briquettes are designed to burn slower, and therefore cooler, but we need the heat in order for it to melt. A little lighter fluid and fire and the thing started up.
2. I turned on the blow dryer and taped down the "cool" button after I found that having it on hot was bowing the PVC pipe. Once it was on though, you could hear the fire roaring like a furnace.
3. I continually fed the foundry until the whole thing was getting warm. The foundry itself acts like a sponge for heat, so I needed it to get hot before I could load anything else into it.
4. When it got a little warm to the touch, I put in the crucible and a can, and loaded the sides between the crucible and the foundry with more coals and kept it burning.
5. With some threaded rod we had laying around, I poked the coals around and poked the can to see if it was starting to melt. 
6. After an hour of continually stoking the flame and getting Matt to be my shoveler and constanly feeding the flame, the aluminum stopped being crumbly (weird, right?) and actually started to melt.
7. Once the curcible was full of molten aluminum, I threw in some borax and scooped out the dross.
8. I had to add more cans and more borax a couple times because there is a lot of dross in one can.
9. The crucible was full of completely molten aluminum, so we shut everything down, lifted out the cricible, and poured it into the sigil. At this point, we were aiming for "Attempt 1" and not "Success 1."

Because of the way I poured it and the high surface tension of molten aluminum, I got several voids in the sigil, and an overall terrible casting. But like I said before, now I've got a taste of it and I need more. There are many other designs out there that take weeks to properly cure and assemble the materials of the foundry, and proper "cope" and "drag" molds. Now I stare at the casting on my desk, a rough paperweight, and excitedly plan attempt number two.

If you have any questions about how I might do it differently or what you think might go better in some areas, feel free to leave a comment. 

This was our first attempt at aluminum casting, and it turned out okay at best, but we learned a lot about metal casting. Mostly what not to do, but our mistakes are your learning!