HTMAA 25
home about me final projectThis week’s project was to create something using molding and casting processes.
Documentation here.
I thought for a long while about what would make sense to mold and cast - something rigid, probably, with simple enough geometry (not a ton of overhangs) and something that I would want to hypothetically create a lot of copies of.
I thought first, briefly, of mahjong tiles, then of the custom Chinese name chops that I’ve seen sold in touristy places. The thing about a name stamp, though, is that it’s supposed to be unique, so there’s not really a point to making a mold from it?
Still liking the stamp idea, I decided on doing emoji stamps! I’m a very emoji-heavy texter by habit, and there have been times where I’ve been writing notes for people etc. where I’ve felt that, missing the emojis, it’s more difficult to portray the tone I want. My mom also suggested I do a little makeup stamp in the shape of e.g. a butterfly, which I thought seemed pretty cool too.
I found a cool-looking butterfly svg online, and decided on a set of emojis I wanted that I thought were a) useful and b) not very substitutable with an easy-to-draw alternative like T^T for crying or :D for a happy face.
On my phone, I’m used to seeing Apple’s emoji icons, but Apple obviously does not release these in nice svg formats. My first thought was to try and vectorize my chosen emojis from a screenshot. I tried using Illustrator’s image trace tool, trying to tune it by fussing with all of the sliders, and was rewarded with:
Which clearly looks awful. I blamed resolution, and found higher quality emoji pngs to try the same process with. This was marginally more successful:
But I realized that the shapes were still kind of off and they weren’t getting simple enough that I could just take the outlines of them as stamps. My options seemed to be to try and manually simplify the shapes and isolate the important parts to stamp, or to trace the reference image, neither of which I am a good enough Illustrator user to do.
Giving up on that course of action, I decided to give up on Apple emojis, and found this lovely open source emoji library that came with outline-only SVGs, OpenMoji. Wonderful. Here are my emojis, in open-source outline form:
The skull looks ugly and some of the emojis are less expressive than I’d like, so I did a little switcheroo and edited some faces to get:
Now happy with the shapes, I scaled them down (so that the strokes would be thicker) and used Illustrator’s Expand tool to expand the stroke paths into actual shapes that I could later extrude in Fusion. See the difference?
I then arranged all the emojis + the butterfly to the size and spacing I wanted them to be and imported them into Fusion. On the same sketch, I defined the size and shape of the blocks I wanted.
Then added a ~1mm rectangular border ~5mm away from each of the stamp blocks, and extruded everything.
I printed this mold on a Prusa CORE One, sliced for PLA with an 0.1mm layer height. I chose a smaller layer height than usual for a smoother, better quality finish. Since everything is an extrusion from the ground, no supports were needed. The print took just over an hour.
I wanted to use wax to smooth out the details on my print, but given that all my details were already so tiny and that the only details that really mattered for a stamp was the very top layer - which of course was on one layer and hence printed perfectly - Anthony advised me that I could just directly pour on the silicone.
I used MoldStar 15, mixed at a 1:1 ratio and poured very slowly and carefully to avoid bubbles. Since my mold was so tiny, the actual total amount of silicone mix I had didn’t even reach the lowest line on the measuring cup.
Regardless, I poured the silicone to the top of the print and set it to cure under the heat lamp. This took another couple hours to fully firm up.
Once the whole surface seemed relatively firm, I pried the silicone out of the 3D print. This was much more difficult than I expected - it was very hard to get enough leverage to pull the silicone out, especially since it was stuck so deeply in the gaps between the stamps. While I got it out eventually, it may have been better to leave more space between the stamps, and more space between the stamps and the wall.
Some details were lost as the print was just too small, and some bubbles ended up stuck near the surface. When the cast plaster comes out, I will see what I can do to clean up the stamp surface.
I used USG Ultimate Drystone to cast my stamps. After mixing the powder with water until it was the consistency of pancake batter, I poured it (somewhat messily) into my silicone mold.
I periodically checked the remaining plaster not poured in the mold for progress in hardening, aiming for the point where I could no longer scratch the surface of the plaster with the popsicle stick and enough hitting it with enough force makes it crack. After this point, I pried my cast out of the silicone mold.
The results were a little disappointing. The plaster clearly hadn’t gotten into the fine lines/details, and my stamps definitely were not functional as stamps. Anthony suggested I recast the stamps with a slightly thinner mixture, and use an acid brush to tap the plaster into the fine details before filling the main volume of the stamp.
This time, after casting, the results were much better.
I tested out the emoji stamps on paper, first with acrylic paint and then with an ink pad I found randomly.
Some of the gaps between lines were too small, and the more viscous acrylic easily pooled there, making the stamps less clean and nice. Some bits of the outline broke off. The results were significantly nicer with the ink, but I think the pad I was using was a little dry. Besides, some outline pieces had already broken off or chipped away by the time I got to testing.
While this was already quite large compared to my normal handwriting, and so already hard to use in-line, it was clear that the stamps or at least the lines were still too small for plaster as a medium.
The butterfly stamp was slightly larger, and held up better. Testing with sparkly makeup on my hand, it works as expected!
I think something I did well this week was consider a design that was reasonably well-suited to the technique, both in terms of its actual physical form as well as its application. Unfortunately, it was a bit too small, and so resolution was lost when 3D printing, minor defects in pouring became proportionately larger defects in final product, and the fine details in the plaster were liable to breaking off. If I had more time, I would consider using the Othermill or similar for a more precise positive mold, and using epoxy or a more cohesive material for the cast.
