14. Moulding & Casting

Group Assignment

See Reykjavik Site

Hero Shot

The Idea

Making a thing. Using that to make a thing which makes more things. Fairly intuitive.

I was sitting at the “On-Call” desk in the lab during open hours racking my brain trying to decide what to make when I ran into one of the things I was reading up on the day before. Normal maps. They’re this really clever way to encode 3D lighting information onto a 2D surface for computer graphics and they look really cool:

So with that idea in mind, I decided I´d try to re-create that effect in real life. The plan is to CNC out the positive on the Shopbot and cast silicone into that. Then use silicone mold to cast Jesmonite. I’ll then spray paint the Jesmonite from different directions to recreate the effect.

Design

Starting off in Autodesk Fusion I designed this arrangement of different shapes:

Not a lot to say here. It’s just some shapes on a flat plane. The trickiest part was definitely the gemstone shape, Getting the facets lined up was surprisingly hard.

CAM

I’m making the toolpaths for this in Fusion.
Starting off with an adaptive clearing path to remove most of the material:

I decided to use PU600 instead of wax since we’ve had a stack of blocks in a cupboard for almost four years now but still haven’t tested it. I couldn’t find any definitive source on machining settings so I did some guesstimations based off prior experience on the shopbot. I used the Fablab Speeds and Feeds Calculator which has a reference table of materials for speed and chip load. Here’s what I ended up with:

In retrospect I probably could’ve doubled the RPM/Surface speed without issue. But this material is fantastic.

I also added a second path after the Adaptive Clearing to clear out the edges for the upper lip of the negative mold.

After that I moved on to the finishing passes. I ended up doing two finishing passes.

One with a 3mm flat end mill to clean out the remaining material from the 6mm Adaptive Clear pass:

And a final Scallop pass to smooth out edges and clean things up.

I had to do a lot of custom tool definitions for this. Measuring the bit with calipers, Defining everything in Fusion and using the Speeds and Feeds calculator.

Here’s the completed simulation:

Post-Processing is relatively easy. I’m using the ShopBot OpenSPB Post-Processor. I make sure to export them individually by selecting each in the operations tab and naming each clearly.

.

Routing

Starting off with the adaptive clear:

Then moving on to the Quarter Inch

I noticed the Z-Axis Zero position wasn’t quite correct and I had a tiny amount of material remaining. The zero plate was probably resting on a piece of debris. I re-zeroed the Z-Axis and ran the job again. This gave me a flat finish but some minor tool marks on the surface.

And here’s after running the final pass with the 1/8” ball nose.

And here’s the piece after sanding. In retrospect I should’ve sanded it more, Especially on the gemstone and triangle shapes. But it was good enough. The 1/8th” bit also didn’t remove as much material as I expected. The toolpath was long enough that I didn’t want to re-run it as it was already getting very late.

After sanding everything nicely I went on to pouring the silicone.

Silicone Casting

I started off by using Fusion to calculate the exact volume in the mold. I created a new component, Copied the original body into that then created another body and combined the two leaving only the negative inside the positive. This tells me the exact volume required for the silicone.

I’m using Zhermack ZA13

Here are the safety sheets:

• ZA13 Catalyst
• ZA13 Base

Main takeaways from these are:

• Wear gloves
• Wear goggles
• Wear overalls(Best I had was an apron)
• Use respiratory protection where ventilation is insufficient or exposure is prolonged. (The chemicals workspace is very well ventilated with constant negative pressure.)

Knowing this I mixed the materials in the Zhermack mixing bowl using the calculated volume as a reference. I Neil mentioned stirring at least twice as much as you thought was necessary. I stirred until my arm hurt. I think that was enough. I didn’t manage to stir cleanly enough to avoid bubbles but if poured carefully they shouldn’t matter much.

After pouring I left it for 12h and came back the next day.

Getting the silicone out of the mold was surprisingly easy. I added draft angles to all the smaller parts but forgot to add them on the periphery.

Here’s the completed mold:

I’m sure the aggressive mixing caused a lot of bubbles. Most of them were not on the surface so I wasn’t particularly worried about them affecting the surface finish. I did however notice that despite what I thought was sufficient sanding the cut marks still transferred onto the silicone.

Jesmonite

Andri suggested using Jesmonite AC100 as a casting material. It’s a surprisingly versatile material with an incredibly short set time.

The stuff we have here in the lab is Jesmonite AC100

Reading through the Technical Data Sheet I was genuinely surprised to see an initial set time of 15-20 minutes.

The mixing ratio is 2.5 units of base(powder) to 1 unit of liquid.

The Safety sheet specifies similar rules as the Zhermack Silicone.

• Gloves
• Goggles
• Jumpsuit
• Respirator if in poorly ventilated area

Mixing was surprisingly tricky as the scale I used for the Silicone was out of batteries. I managed to borrow a very accurate three decimal scale from a coworker but that had a maximum weight of 200 grams and the mixture I was making was slightly over that. Luckily I could weigh the powder in a separate container and then add it to the liquid after measuring both.

Mixing this stuff was easier than the silicone. Although I did have some difficulty getting completely rid of some clumps. It’s less of a liquid and more of a paste. The guide recommended brushing it into the mold before pouring, Which I did. Luckily this stuff is water soluble so getting it out of the brush wasn’t too hard. The working time on this stuff is only five minutes so I didn’t manage to get any images of pouring or mixing. After removing it from the mold I broke off the excess flash and got started on prepping to paint.

Painting

I started off with making a stencil to cover the bottom surface while spraying. I created a sketch in Fusion and used Project & Include to get the outlines of the shape before exporting that as a .DXF and cutting it on the vinylcutter on some leftover scrap.

I then applied that as a stencil onto the piece and started spraying. Each color is sprayed at an angle 120° offset from the other. I used the faces of the triangle as a reference and sprayed at a low angle.

I forgot to take a photo of the part sprayed with the stencil still applied. But here’s the stencil and the part after finished painting:

Ultimately I could’ve post-processed this better but it does not show any production marks on the final product. If I had the time to give this a once-over I would re-think the way I machined it, Using the ball-nose more efficiently to create smoother surfaces and sanding more.

Design File

I did all of this in Fusion. So here’s the file:

Filename	Type	Link
NormalMapsMolding	.f3d	Link