This project was an experiment on aluminum casting using a 3D print to create the mould, following these general steps:

3D model
3D print
Sand mould
Melt away the plastic from inside the mould
Aluminum casting
Final adjustments

The final outcome is not complete, due to some issues during the process. However, it is worth to explain each step I took during this exploration, keeping in mind that this was the first time I made something with metal casting.

This process is surprisingly fun and interesting, it doesn't require a lot of equipment and is great for experimental projects.

Step 1: 3D Model and 3D Printing

As a starting point for this experimentation I took a 3D model of a (Mammoth) vertebra from Sketchfab (also downloadable here in the files):

https://sketchfab.com/3d-models/mammoth-lumbar-ver...

The same process can be applied to different objects, but some decisions I made during the process were depended on the shape of this particular object (i.e. leaving the 3D print inside of the sand and melting it).

This 3D model was created with a 3D scanner, therefore a bitmap texture of the bone is also available. However, it was not needed for the process.

After downloading the model, I imported it into Slic3r and started setting up the machine. I used the Original Prusa i3 MK2 as a 3D printer, with normal PLA white filament. I left the settings almost to default, reducing the grid infill to 10/15% (since I didn't need strong mechanical structure for the object), and the layer height to ~0.15mm. I added the external support structure to the model, since it couldn't be printed without it.

G.10770_obj.obj
Download View in 3D

Step 2: Sand Mould

Once I had the 3D printed model cleaned up from the support structures, I started creating the mould for the aluminum casting.

One important note to consider is that the shape of the vertebra is very complicated, and is not very suitable for moulds and casting (it has a lot of undercuts and asymmetrical features). Therefore, this entire experimentation was about giving a try to cast this object without having to split the mould (how you would normally do) in different parts.

To do this I placed the vertebra inside a metal frame, with the three elongated extrusions facing downwards. Then I started filling up the frame with casting sand (sand mixed with oil), hammering it and pressing it really hard to make it become more rigid and stiff. The initial idea was to keep the biggest round-ish part of the vertebra facing upwards to let the metal flow inside more easily.

Once the 3d print was fully covered up in sand, I drilled several holes reaching the bottom part of the object, to allow the air and the metal to flow freely inside the mould. On the upper part, I created a bigger hole with a funnel, and this was the point from which the metal had to be poured.

I then tried to melt away the PLA with a torch. However, since I did not pour the plastic out completely, some of it stayed inside of the mould on the bottom part, and the final casting turned out incomplete because of this.

Step 3: Aluminum Casting

I took as raw material some scrap aluminum I found in the workshop, and melted it in a pot at around 990-1000°C. Even though the melting temperature of aluminum is around 650°C, I had to always keep it higher due to impurities/other materials combined with it.

Once melted, the aluminum shows a layer of patina that has to be removed. This is the part where the impurities combine. Also, I increased the temperature of the pot a bit more using the torch (~2000°C), to obtain a very liquid state.

When the aluminum seemed liquid enough, I poured it through the funnel of the mould, until I saw that it also passed through the secondary holes I created.

Step 4: Final Object

Due to the mistake I made previously (not melting away the PLA completely from the mould), the final object turned out incomplete.

Since the plastic that was still inside accumulated in the lower part of the shape (the three extrusions), the aluminum didn't reach that part, and only stayed in the upper bigger part of the mould.

After cutting away all the unwanted parts, I smoothed out and slightly polished by hand my now incomplete vertebra.

Without the three spines a lot of imagination is needed to see a vertebra in it, but it was nonetheless a worth experimentation to understand what could work and what couldn't during the aluminum casting process.

Here are some things I will definitely keep in mind for the next time: