I designed, iterated, and cast molds for two rubber seals as part of my Summer 2023 internship at iRobot. I used surface modeling in CREO to design the molds and printed them on an Object 3D printer with a VeroClear. The parts were cast with Castaldo QuickSil firm 2-part jewelers rubber. This was my first time creating molds and I learned a lot — I can't get into specifics about the parts, but I'll share my process and learnings!
Skills
CAD, Surface Modeling, Mold Design, Casting
Mold Design
Seal #1
Undercut
This design was much better, I was able to take each outer part of the mold off the part, and then de-mold the part from the middle-insert.
This first seal was tricky because of the undercut in the part. The casting material has a rated hardness of Shore 40 A, so I initially decided to make this a two-part mold with the hopes of being able to pull the part out from the undercut.
I identified a good place for the parting surface and constructed the mold around the part. Here are some design features I included to improve the molding process:
- Drafted corners for alignment
- One corner larger than the others for orientation
- Overflow holes for excess material
- Gas escape groove
- Through holes to bolt the parts together and embedded nuts for easy clamping
- Screwdriver pry slots
The casting material is a playdough-like puddy, so I packed the molds with it and was able to squeeze it into tight corners and nooks on both sides of the mold. Lots of flash was generated this way, but it ensured there were no air pockets.
​
The two-part design didn't work for this part! Unfortunately, the rubber near the undercut was too thin for the material to be pulled out and under, and the only way to de-mold it was to tear the part. I design a three-part version with an insert for the undercut.
The last thing to do is clean up the flash and extra material. I used scissors and an X-acto knife for this.
Seal #2
The most applicable thing I learned was to think about how the material flows and acts when clamped in the mold and imagine the cured material sticking a lot during the demolding process when designing a mold. I originally put the overflow vents in deep corners of the part in hopes that those corners would be filled out before overflow occurred, but this made it extra hard to de-mold as the material in the vents stuck to the walls in those deep corner locations. Because this material can be placed wherever I want when loading the mold, it would make more sense to place these vents in open spaces on the part.
​
I also undervalued the amount of skill it took to actually use the molds. If I didn't use enough material, it would be a short shot and the part would be unusable. If I used too much material, the squeeze-out would be larger and the vents wouldn't be able to handle it. This could mean the part becomes thicker if the clamping pressure isn't high enough. De-molding was also tricky because the material stuck to the mold even with mold release. If I wasn't careful, I'd tear the parts.
​
Overall I really enjoyed this design opportunity and I particularly liked how the mindset required to design a mold is different than that needed to design a part.
This second seal has no undercuts or complex geometry, so I made a simple two-part mold and used the techniques I learned from the first seal to quickly design this one.
Because of the ridges in the part, applying mold release was important in ensuring the seal didn't tear.
In order to get even more clamping force, I put one end in a vice, tightened the screws, then repeated the process with the other end.
This second seal has no undercuts or complex geometry, so I made a simple two-part mold and used the techniques I learned from the first seal to quickly design this one.
