Lost-PLA Aluminum Casting

Last Sunday, several Hackerspace members came over to check out my David Gingery-style aluminum casting setup.   We tried “lost-PLA” casting, and it worked out pretty well!

Aluminum Yoda

Lost-PLA: Works, It Does

 The test target was the 3D-printer-classic Yoda head.  It was printed on a RepRap in PLA and bedded directly in sand.  Molten aluminum was poured straight into the PLA, vaporizing the plastic as it filled the mold.    Overall, this process makes for a shorter, simpler route from a virtual 3D model to a solid aluminum object than anything else I’ve done or heard of.

Jim printed Yoda on his RepRap Mendel with minimal infill.    In order to put the sprue on the bottom it was cast upside-down.  So I started by setting the plastic model on a molding board, and rammed up the drag (packed in sand) around it.  Rolling over the drag, I set a 1″ sprue pin on the exposed base of the pattern, and rammed up the cope around that.

The actual pour was somewhat exciting (that’s a bad thing), as you can see in this video Seb shot.  On the initial pour, it appeared that the aluminum failed to enter the pattern at all.  (Castings fill wildly faster than I expect, plus I was distracted by the sudden eruption of fire.)   The flames died down, revealing a gas-filled bubble of aluminum rising from the sprue.  This is a bad sign, so I tried pouring in a bit more aluminum.  This is generally pointless and never does any good, but it’s hard to resist trying.  So the extra just filled up the sprue a bit more, expanded the puddle on the surface, and torched the strap holding the flask together.

After the pour, we were all convinced the experiment had been a complete failure.   As we  knocked the casting out of the sand, I fully expected to find a misshapen, formless blob.  So convinced was I of failure, that with the casting still too hot to touch without heavy gloves and covered in burnt sand & plastic residue,  I assumed we were looking at scorched PLA.  I had just begun to speculate about how the plastic could have survived so entirely intact when Jim grabbed a stick and whacked Yoda on the head: Ding!  Surprise victory!

Overall, the reproduction quality is excellent.   In spots you can even make out the fine striations from the 3D printer.  The parts that didn’t work:

  • Yoda’s long, thin ears didn’t fill, which was entirely expected.  This might be resolvable with the addition of gates & vents so that the aluminum can come in from the bottom while vaporized PLA escapes upward.
  • The wide flat-bottomed base of the shoulders didn’t fill.  Upside-down this became a wide flat-roofed space, which trapped a big enough bubble of escaping gas that it couldn’t get out through the  bottleneck at the base of the sprue.  If I’d realized this ahead of time, I would have made a bigger sprue covering the entire base and this wouldn’t be an issue.

In any case, I’m pretty excited about the possibilities with this process.   It eliminates lots of tedious pattern making and at the same time radically expands the variety of things I can cast.


Here is a picture from Jim of the pattern mid-print; showing off the fancy 10% hex infill:



Carbon Fiber Fabrication

Why join a hackerspace?

Everyone has their own reason, but for me it’s mostly building a community to learn from and share with. But, there are practical reasons, too… some of my projects are messy!

A long time ago, I worked at a small satellite company and was good friends with the guy who built all the carbon fiber structures we used.  I watched the process many times and learned the techniques, but never made anything myself.  Fifteen years later, and I decided to try my hand at it — and got some good press for it!  My first experiments were in the basement of a big house I was staying in, but since then I’ve moved in to a much smaller place. Luckily, my office moved in to a much larger place and I could continue my work after hours there… until… that one day.  I was using Bondo to alter the shape of my mold — Bondo is extremely smelly stuff for the 5 minutes between the time you get it out of the can and the time it cures.  That’s the exact time my boss walked in and thought I was killing myself with the fumes. I need a hackerspace so that my boss doesn’t have another heart attack.

My first project used a simple mold.  I’ve been advancing my technique to learn more and more complicated mold-making methods that let me build more customized objects.  Now I’m trying to reproduce my hardshell backpack.  After a bunch of slight mistakes (documented on my flickr page), I made my first shell:

It’s a fully functional part, lighter but much stronger than the ABS plastic that my hardshell backpack originally came with. It also has kevlar embedded in it in case I fall off my motorcycle and this thing skids along the asphalt. But, this part is not quite up to my expectations cosmetically.  I could sand and clearcoat it and it would look great – but because the surface quality is a little poor to begin with and because I’ve got some big epoxy bubbles, this will take a lot of work. So, instead, I’m planning on trying another run – it’ll be quicker & easier.  I’d love to show the processes to anyone who is interested – hopefully in the next week or two.  Anyone up for a 2-hour demo some Saturday afternoon over pizza?  (it has to be an afternoon because I need to run a vacuum pump for several hours afterwards)  It won’t be a class per se, but you’ll learn lots of techniques and find, while there are bunch of steps, it’s not as hard as it looks. And about 1/10th the cost of buying carbon fiber parts commercially.

Suggested reading: West Systems has some great tutorials and free pdfs

— John Maushammer