My wife was given a silicone ice cube mold with a particular brand embossed into it. It makes ice cubes with words on them. For some reason, this amused me, so I decided to make my own version just for fun.
This document is a running project log of designing and casting my own silicone molds. The journey is still in progress.
**Current Status**: To provide a quick overview, here are some of the key outcomes and challenges so far:
Successes:
- Achieved minimal flashing and no significant leaks in the mold.
- Successfully removed ice without cracking or sticking issues.
Failures:
- Sides of the ice cube bulge during freezing, affecting shape.
- Current embossing depth doesn't achieve the desired boldness.
Open Questions:
- How can the shape stability be improved further?
- What adjustments will give a clearer embossing on the ice?
## Project goal
Make a one-cube silicone ice mold that:
- produces a crisp, readable embossed mark
- stays square while freezing (no bulging sides)
- releases the ice without a fight
For the first tests, I kept the “word” simple: a monogram of my initial using the Luminari font. The current version is purely a proof of concept.
## Materials and tooling
- CAD: Fusion (modeled the cube and monogram)
- Mold housing kit: [https://www.amazon.com/dp/B0CRVCFMJN](https://www.amazon.com/dp/B0CRVCFMJN?utm_source=chatgpt.com)
- Silicone under test (Siraya Tech): [https://siraya.tech/collections/silicone](https://siraya.tech/collections/silicone?utm_source=chatgpt.com)
- testing both 10A and 25A
- Master pattern fabrication:
- revision 1 printed on [[Trellium]] in PLA (FDM)
## Design approach
I modeled a positive “master” (the cube with raised monogram), then cast silicone around it to create the negative cavity.
Emboss orientation: the monogram is on the bottom of the mold, opposite the opening. I placed the embossed face of the ice cube against the closed end of the cavity.
One cube per mold.
## Mold box setup
I was supposed to tape the seams of the housing kit, according to the manufacturer. However, I decided to experiment by skipping the tape and instead ensuring a precise fit. My hypothesis was that a snug fit using just the model base could prevent leaks and minimize flashing.
**Result**: minimal flashing and no significant leaks.
## Revision 1
### Design parameters
- Form is 50 mm
- Cube is 50 mm - 4 mm (46 mm)
- 2 mm walls
- 1 mm emboss
### Master fabrication
- FDM PLA print
- No post-processing after removing the part from the printer
- No sealing
### Silicone casting process
- Mixed 1:1 as directed by the manufacturer
- Batch size: 60 mL
- No vacuum chamber
- If I keep enjoying silicone work, I’ll likely buy one (also useful for epoxy)
- No mold release
- Cured for 4 hours
- I went with the low end of the recommended range because the walls were thin.
- Room temperature during cure: 72°F
I do see trapped bubbles in the silicone. They don’t impact the surface finish of the cube.
### Freezing test
- about 3 hours at -6°F
## Results of revision 1
- 25A is more well-suited than 10A
- 3 mm walls will be implemented to improve structural integrity. 2 mm walls are too thin, leading to bulging during freezing and distorting the cube.
- 1 mm embossing is too shallow
- The monogram transfers, but it isn’t as bold/crisp as I want
- FDM layer lines and surface texture transfer very clearly into the silicone
- A smooth surface was not a design goal; it was just a discovery from the first attempt.
Demolding notes:
- Removing the ice was easy.
- I could peel the silicone away from the formed ice.
- The ice did not crack around the emboss during release.
- I did not notice frosting or opacity differences between the embossed face and other faces.
## Plans for revision 2
- 3 mm walls
- 3 mm embossing
- Resin-printed master instead of FDM
- Standard wash/cure processing (no special finishing planned right now)
## Potential future tests
- Add a draft angle to encourage the ice to expand upward toward the opening. This adjustment is expected to reduce side bulge, resulting in a more uniform cube shape.
- Add a brim to stiffen the walls, which should improve mold stability and decrease wall deformation.
- Insert silicone sleeves into a rigid tray to enhance structural support and keep the sides square. Enable stacking.
## Next steps
- Update Fusion model for 3 mm walls and 3 mm emboss depth
- Resin print the revised master.
- Cast again (likely focusing on 25A)
- Freeze test and compare:
- squareness
- emboss readability
- surface finish
- ease of release
![[Discussions]]