Home » Uncategorized » Making metal castings from 3D prints

Making metal castings from 3D prints

Why do I listen to my students? It only causes me to have crazy ideas.

So here is the full story.

The student who designed the barman’s bottle opener (see a previous post) was looking at how his design would be manufactured “for real” and decided that it would probably need to be die cast… Of course we couldn’t contemplate doing that but we do have some casting facilities in school… And we have heard of “lost wax casting”, but what about “lost plastic casting” in a traditional sand mould? Errrrr?? This got me thinking, surely at the temperature that aluminium melts the plastic print from our 3D printer would just burn away to nothing. Time for a bit of testing!

With the fume extraction going full blast we placed a sample of ABS in our brazing hearth and played a flame over it… Result a black blob and lots of black smoke. But what about PLA? Tried this too… Result it completely disappeared just leaving a small stain on the fire bricks. Maybe there was some mileage in this, so time for a full test.

We unearthed all the casting boxes, crucibles, leather aprons, face masks, gaiters and other other paraphernalia, made a sand mould with petrobond (oil bound) sand with suitable runners and risers and then fired up the furnace. All very exciting stuff as we don’t get to do this sort of thing very often in school these days.

After a bit of a wait we degassed the molten aluminium and the moment of truth had arrived… Time to pour it into the mould! Everything went smoothly, we now only had to wait for the aluminium to solidify and cool down… The excitement was mounting! And the smell of scorched sand was pervading the whole department much to the disgust of my colleagues.

But now the time had come to break open the mould… Bated breath as the sand was crumbled out of the casting box….. And….. FAILURE!!!…. It hadn’t worked! The 3D printed PLA object was only slightly melted on one side, surely it couldn’t have resisted that sort of heat, but it had.

So, back to the drawing board and a bit more thinking is needed… I’m convinced that it should work but we just have to find out exactly how to do it.

On the other hand, perhaps I shouldn’t keep listening to my students.




  1. Anonymous says:

    Plastic has a much higher specific heat capacity than metals do, so your result does not surprise me much.

    IIRC, the way Vik Olliver did it was to bake out the PLA before pouring the metal.

  2. Dave White says:

    Hi Nophead,Thanks for your comment, you’re probably right. I should probably talk to some of the teachers in our science department and see what they have to say. Anyway, both the students and I learned quite a lot by just “having a go at it”. And it was great to fire up the casting furnace… Something that we don’t do very often! By the way do you know any more details about how Vik did his castings. Did he use plaster moulds? Did he cast with aluminium or some other lower temperature alloys? It would be great to follow this up with some more experimentation…. Safety rules permitting!Sent from my iPad

  3. Anonymous says:

    Here is Vik’s write up: http://blog.reprap.org/2009/12/metal-bits-from-reprap.html. He used pewter, which I guess is lower Tm than AL. He also has some warnings about plaster and AL.

  4. Dave White says:

    Cheers Nophead… And many thanks for the link to Vik’s write up… I think we might well have to settle for a “safe” lead free pewter type alloy and a plaster mould. The resultant metal bottle opener may well be too soft to open many bottles but should be good enough for testing this students design… A shame as I was looking forwards to having to “empty” the bottles that we tested it on, bad luck that the student isn’t over 18 yet and someone will have to do it for him : – ) well it is the weekend after all.Sent from my iPad

  5. Anonymous says:

    Could you skip a step and use a paste extruder to 3D print the sand mould directly?

  6. Dave White says:

    Nice thinking Tom, but sand is very abrasive stuff (would probably destroy any sort of feed mechanism)  and I’m wondering how it could be bound together to make a mould and not just end up with a pile of sand a bit like trying to make a sandcastle out of dry sand above the high tide line on a beach! … And of course engineering something like this would be very time consuming (and we are now running up towards exams with our students and have to focus on that)Some time ago I had the pleasure of dropping into BfB HQ when Unfold was doing some experiments with chocolate paste, mashed potato and clay (there are some videos on YouTube of this) and it was very “exacting” stuff. He had great results with clay making some lovely pots but I somehow doubt that we could apply this to sand…. But now you have got me thinking about some sort of plaster or refractory clay or maybe silicone or…. Damn, listening to my student gave me crazy ideas and now listening to you guys and your comments on this post- maybe something for  another day??!! .. But you have set me thinking!Sent from my iPad

  7. Anonymous says:

    The major step you skipped in the lost wax process is burning out the wax. Generally lost wax process involves making an investment mold around the wax positive. Your positive would include all of the vents to allow the gas to escape during wax burnout and casting. BTW, an investment mold usually contains plaster and refractory which is usually sand. After the mold sets and dries it is placed upside down in a burnout kiln.

    This is where things get interesting. In my undergrad we had a small aluminum and bronze casting area. The burnout kiln we had was an old leftover kiln we received from the ceramics department. Generally it would not heat to high enough temperature if not constantly attended. Even then, most times large wax positives would not burn out completely.

    So, one year the digital art department purchased a Dimension 3D printer. The plastic they were printing with was ABS. So, we had them print out a test part and made an investment mold for it. After the burn out, we had the cleanest casting we had ever made. The plastic burned out MUCH CLEANER than the wax ever did on our shoddy kiln.

    Give it a try again but burn out the plastic first. You will be pleasantly surprised.

  8. Dave White says:

    Thanks for the comment reprapperMC 
    Some very useful advice. 
    Perhaps in the main post I didn’t describe what we were doing properly… I mentioned “lost wax” casting as this is probably familiar to most people and fairly similar to what we were trying to achieve. I suppose I should really have said it was similar to a “lost polystyrene” process. Embedding an expanded polystyrene positive into a sand mould and then pouring in molten aluminium totally burns away the polystyrene which leaves a cavity in the sand that is then filled by the molten metal… Out testing by heating some PLA resulted in the polymer “disappearing” and our hope was that it would therefore behave a bit like the polystyrene. I now suspect that in our test the PLA didn’t burn away but was liquid enough to soak into the fire bricks that it was resting on. In the sand mould I suspect that the surface in contact with the molten metal simply melted and the petrobond (oil bound) sand wasn’t porous enough to wick it away. The PLA only melted on this plastic/metal interface and plugged up the gap preventing the metal running any further… Hence what can be seen in the photo, just a little melting or scorching where the runner (inlet) was located.
    I think our best plan if we are to continue with this will be to do as you suggested and look at something more akin to “lost wax” methods.. Ie plaster moulds, lower melting point alloys and melting/burning out the PLA model.
    Anyway, thanks again for your observations and advice reprapperMC (and also everyone else taking the time to read through our adventures in casting). It has been a great learning journey for us!

  9. Anonymous says:

    I have personally turned PLA objects into aluminium and/or pewter objects using “lost PLA” , which, as others have said *does* require you to melt the PLA out of the mould first, before pouring metal in. 🙂

    For Pewter, I recommend any sort of plaster of paris mix as the mould compound, and once it’s set and airdried overnight someplease warm, bake it in an old oven at a nice high temperature for a few hours to remoce the PLA. Removing the aluminium object from the plaster is just a matter of smashing the plaster, and attacking it with sharp pointy objects. I have not tried this method with Aluminium.

    For Aluminium, I have tried a high-temp epoxy based( 250C, 500F ) kneadable putty as the mould agent, and had limited success, but it degraded the resins, stinks bad, and is not a very good solution. I did however end up with an aluminium copy of the PLA original, so in that manner, I guess it was a success, although not one I’d do again the same way.


  10. Anonymous says:

    You may remember from the RapMan forums that my son works in an art Foundry and took my printed ABS torso, painted it with wax and sanded (to remove the imperfections) and used the lost wax casting method to burn out the ABS from the mould and filled with bronze. They did have to strengthen the torso with bronze bar to ensure that it maintened it’s integrity, but other than a basic description of what they did, I have little more info.

  11. Anonymous says:

    I have had issues with ash and burnout when casting softer wax pieces (higher oil content). Solution is to burn them until they stop smoking… takes a loooong time. And then canned air to blow out ash. Canned air is crucial.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: