Turning 3D Prints Into Aluminum Castings

[Jeshua] needed a laser head attachment for a 5×10 foot CNC machine he’s working on. Because he has a 3D printer, [Jeshua] could easily print a laser mount and attach it to his CNC gantry, but that wouldn’t look very professional. Instead of decorating his gigantic machine with brightly colored plastic, he decided for a more industrial look by casting a laser head in aluminum using a 3D printed master.

[Jeshua] designed two parts for his laser cutter in OpenSCAD and printed them out on his 3D printer. A few bits of foam insulation were glued on to act as sprues, and an investment mold was made out of 1 part Plaster of Paris and 1 part playground sand.

After the mold had cured, [Jeshua] put is mold in a coffee can furnace to burn out the wax and foam. These hollow molds were placed in sand and the crucible loaded up with aluminum scrap.

The finished laser head fit his CNC machine perfectly – no small feat, considering [Jeshua] needed to take in to account how much the aluminum would contract after cooling. Not bad for one day’s work.

54 thoughts on “Turning 3D Prints Into Aluminum Castings

      1. Too bad I waited so long? Please.

        How many homes had a Z Corp machine 20 years ago? How many even have a Z Corp machine at their home today?

        Sure, blame me that it has taken 20 years for 3D printers to start to become common place in the home.

        Besides, this is an entirely different process. What is your excuse?

    1. I think this proves that there is no need to print wax that I think would be problematic printing. Wax is likely to ooze/drip far more than PLA, and I think it would clog trying to force it into an extruder.
      The point of this experiment is that the PLA burns out just as cleanly as wax.

      If it ain’t broke, don’t fix it!

    2. Check out a SolidScape, they have been around for quite some time now. 2 Types of printed wax, the support wax is dissolved with Kerosene leaving your wax die for investment casting. Common in the dental and jewelry industry.

    1. You can prevent shrinkage through risers. Basically it’s a cavity connected to the mold that holds more liquid metal. If the riser cavity has more volume than the mold cavity, the mold cavity will cool a little faster, shrinking as it does so, which will allow more molten metal from the riser to seep into the mold, making up for the shrinkage.

      1. Actually you cannot prevent shrinkage. The metal will shrink as it cools no matter what.

        What risers do is feed the mold as the part cools. Without adequate risers, you will get defects in the cast such as voids.

      1. Sure, or you can buy a finished part, an entire assembly, or a complete CNC.

        The thing here is that that the part was made in the OP’s shop with his own hands.

        Hackaday would not be as interesting if everyone simply bought something and shared the news.

  1. …could easily print a laser mount and attach it to his CNC gantry, but that wouldn’t look very professional. Instead of decorating his gigantic machine with brightly colored plastic, he decided on an unsightly low resolution aluminum copy of the part instead.

    Not sure how this suddenly becomes professional when you start with a low quality part as a master?

      1. I believe he’s commenting about the extruder lines. Most mold masters are quite smooth both to allow easy removal from the greensand and to minimize post casting work. It’s not so much of an issue since aluminum doesn’t wear too hard on tools but not every material is as forgiving.

    1. I was thinking that too – the master was not smoothed and the casting didn’t seem to get cleaned up either, but above all that I’m fairly sure he could have made the same mount from a bit of steel or ali with a 90 degree bend in it and a few holes drilled…

      But, I still admire the process if not the rationale.

    2. Actually, the reason why I casted the printed parts out of aluminum was not for looks but for function. Plastic parts right above where I am going to be cutting with fire did not seem like a good idea to me!

      As for the looks, I intentionally left the extruder lines because: 1. it did not affect the function. 2. I wanted to see how much original detail I could get in the casting, and 3. I think it is cool that you can see that the original pattern was in-fact a 3D print.

      Note too that in addition to sanding the 3D print, I could still easily polish up the cast print – aluminum is very workable material.

  2. This type of casting is done using blue or pink styrofoam insulation (available at home centers, like lowes, home depot, et al.), hot melt glue, sand, and a wooden box to hold the foam mold and sand.

    THERE IS NO WAX INVOLVED, don’t use it.

    The Lost foam is faster than lost wax, not to mention easier. With a small cnc, you can even mill foam molds. There are a coupel videos on youtube of a person making linear bearings and linear carriages using lost foam casting.

    For anyone interested in doing this, google up

    “lost foam casting”

    or go to YouTube and look for the same thing.

    1. He is actually doing a lost wax type process because he is making a ceramic mold around an item, typically made of wax but in this case PLA, melting out the item and casting into the indentation left behind. With lost foam, you pour the metal directly into the foam item (typically dipped in a ceramic slurry and dried to keep the sand out of the casting and provide a barrier through which the foam gas escapes) and the foam immediately melts and the gasses transfers through the ceramic layer into the sand.

    2. Lost foam is a great process, but it has its own share of problem too.

      For one, you can print shapes that would be impossible to mill out of foam.

      Two, burning out bio-plastic is more environmentally conscious.

      Three, you have to be very very careful how sand is loaded onto the foam part to avoid distortion. This problem is practically eliminated with the lost PLA process.

      In short, do the process that works best for the part being made.

    1. He probably doesn’t have a single billet large enough to do it. Home casting isn’t very hard to do. The hard part is to just get started. It’s not something that brings instant gratification. Expect to spend a couple weeks learning how to build a furnace and a week or two actually building it.

      If you’re interested in home metal working, I suggest David Gingery’s set of books. He covers everything.

      You might want to start right here:
      http://www.backyardmetalcasting.com/

      1. J.Eriksson says: If I were to cast a large billet; why not just cast the final parts needed like I did instead?

        Besides, this was a good test to prove that lost PLA is viable. It is possible to print parts that would be practically impossible to mill.

        Further more, milling has its own drawbacks too. For one you have to wear out expensive end mills. Maching aluminum is rather slow going on my CNC machine as well, and you have to take other things into account such as coolant.

        I think this process is far more ideal than milling even for these parts which could have been milled.

        1. Now that someone has done the real world experiment, others will follow. I can cottage industry growing out of this.

          Plenty of printed items look great, but can’t stand up to actual use. Personally I hate lost wax. Doing this in PLA damned cool.

      2. 3dtopo:
        Talking precision or consumer products, there is no such thing as casting final parts. Probably the only exception is Mineral Casting/Polymer Concrete which does not require any post processing.

        But I agree this is good practice and proof of concept testing. It’s even good enough for quick fixes to get your laser running. But please don’t say your part is final when its missing an entire corner with elongated slot. Its bad practice, think like the poem “there is goal and meaning in our path, but it’s the journey that is the labour’s worth.”

        I will forgive you if you use your laser to cut out improved parts though. :)

      3. J.Eriksson:

        Actually, these cast parts required zero post-processing other than just cutting off the sprues and vents.

        The elongated slots are part of the design (see the OpenSCAD screenshots and pics of the 3D printed parts). The elongated slots makes it so the parts are adjustable which was one of the primary design criteria.

        The missing corner is actually my fault. When I was cutting off the sprue I should have been more careful, so I actually broke it off. Because of the slots there wasn’t very much material there and the weight of the sprue won. That said, I actually do not need it and it will function perfectly with the two working screws slots that it has.

        So to reiterate, these castings are completely usable in their current state that required zero post processing. I guess you can add lostPLA to your list.

  3. Did he add 6% overall in his OPENSCAD design to account for shrinking aluminum? I’d like to get into sand casting to replicate some obsolete motorcycle parts in aluminum, but am trouble by how to accounf for shrinkage.

    1. He says he printed a test piece at actual size to test the fit, then scaled it by 2% during the next print. I assume that means he changed the scaling at the slicer, though he could have done it in OpenSCAD too.

      Somewhat OT: anyone mildly annoyed by the required Quicktime plugins for the videos? I didn’t think I’d ever see that plugin request again…

    1. I’ve never heard of anyone using silicone for a form before. IMHO, it would probably leave too much residue. The PLA burns up so completely that all it leaves behind is some carbonized ash flakes.

      Why would you want to try it with silicone? Just curious. If you already have a mold that you could squirt silicone into, you might as well just use wax.

      1. Well I was thinking of doing custom silicone things, that are usually produced in mass (e.g. Custom cases). I don’t necessarily want to meet it like it’s done with metal, but rather use silicone gel or something to cast it.

        1. Hmmm, I think those are injection molded. You’d need a really smooth mold, too. They aren’t quite the same material as silicone caulk. I don’t know much more than that about silicone molding. I’d be interested in learning about, it though.

          It’s become an exciting time for hackers.

      2. Yeah, me too. I did a bit of Googling a while back, and came up with almost nothing.

        I mean, people have figured out the “3D Design -> Mould” part, however the “Mould -> Silicone product” part isn’t quite there yet. Oh and I also meant melt, and not meet…

  4. Just last week, there was a discussion in my college class about copyrights and such. when I asked the question about 3d printing your own car. At first I didn’t know how to do this, but this changes all the rules, again.

  5. Aluminium is most often formed through casting. It is used to make industrial and automotive components like gearboxes and engines from A380 and A383 alloys. Aluminium casting is done in many ways. Aluminium die casting is good for high-volume manufacturing owing to speed and surface quality; permanent mould casting is suitable for high-strength aluminium components, and sand casting is suitable for small numbers and complicated geometries. Aluminum casting produces robust, lightweight pieces that are very flexible.

Leave a Reply to Hack ManCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.