Our Digitization Manager, Jeremiah, has been working on a local grant to develop a way to capture relief data from otherwise mostly flat materials like embossed paper, wax seals on envelopes, and the surface of coins. The project is currently in the preparation phase, which means Jeremiah is doing some equipment building.

Instead of buying parts and constructing things from scratch, he decided to really begin at square one, designing the main components in Sketchup and exporting them as files that can be used to print the pieces on a 3D printer. That way, when we share the details of the capture technique, we can also share those files.

Unfortunately, trying to make this project scaleable for others means a lot of refinement of the component design and printing process.

The plan

Jeremiah is by no means a novice with 3D printing, but knowing how to create a good design and STL file doesn’t always mean you can predict how the materials and particular printer will behave. So far, he’s now made a couple of attempts at printing a set of roughly rectangular blocks that have two holes running through them, as seen in this image:

Here are aborted attempts at two of those pieces, being created from the bottom up:

So far, these look like the file visualization representing that point in the process:

The process

One layer at a time, the printer provides walls for the object and fills them with a varying crisscrossing pattern you can see up close here, looking down through the layers from the top:

In the middle (in red in the visualization above) is support material that will be removed from the finished object.

You can also see a pattern layer here on the left, in an earlier printing attempt:

And on the right? That’s what you get when things begin to go wrong with the fill.

The problem

As the material cools, it contracts. If that begins happening as the object is still being created, eventually the crisscrossing fill will not quite line up properly with the object it’s filling in. Below, you can see a good print job that went awry:

The printer hit a snag with the piece on the right (see the bottom of the image) and everything began to go off track, as is evident in this closeup:

The solution

Jeremiah knew there might be some trial-and-error involved in perfecting the design, so he was prepared to make changes that will combat the cooling/shrinking problem.

• Use a different material: the lighter, shinier blue material in the above photos, PLA (a bioplastic), is less apt to contract than the darker material, ABS (a polymer you’ll know from its use in Lego bricks).
• Use less of it: reducing the amount of fill crisscrossing the interior of the object (in this case, 20% down to 15%) and thinning its outer walls will lessen the amount of material there to contract.
• Slow the cooling process: the location of this particular printer makes it hard to keep shielded from the room-temperature air of the surrounding environment, so he’s making another attempt on a different machine, one with an enclosure that will keep the piece from cooling as it is still being formed.

Today, he successfully completed one of the smaller pieces out of PLA at 15% fill. We’ll keep you posted as things progress. In the meantime, we hope someone out there can learn from our adventures in 3D printing.