Silicone Molding

Rapid prototyping using 3D printed molds

In order to create complex silicone parts, molding was the most desirable method of creating our suction channels.

To rapidly iterate through various designs of our suction channel, we used FDM 3D printing to quickly fabricate plastic pour-over molds.

Over 4 different iterations of mold creation and physical molding, we learned about the intricacies of silicone molding as well as important characteristics of silicone that make it more unique to mold compared to other materials like plastics and thermoplastic elastomers.

Risk Reduction Prototype: Week 1

At the beginning of this project, we decided to create a simple risk reduction prototype to accomplish the following:

Validate the use of PLA for the mold material
Understand the basic process of silicone molding (mixing, degassing, pouring/injecting, etc.)
Physically see/feel how Ecoflex (20A Shore Hardness) feels

The mold was fairly simple because all the features were contained in one part of the mold, and the parting line was made above the part by interfacing with a ¼” acrylic sheet.

We ultimately learned:

Smooth-On Ecoflex was a bit too flexible - could visit either geometry redesign or material reconsideration
Pouring may be a viable alternative to injecting due to the viscosity of silicone - more testing required
Overdoing the mold release may have negative effects on part outcome

First Prototype: Week 1 - 3

After we successfully learned the basics of silicone molding, we moved on to the next part of our project, which was designing the first mask prototype.

For this prototype, the geometry needed to match the curvature of a NIOSH headform, which made the mold significantly more complex.

We decided to pursue a two-part mold, with a parting line that follows the outer edge of the suction channel. An injection hole, injection channel, aligning pegs, pry locations, and aligning pegs were added.

One of the biggest learnings was that physically injecting silicone was not feasible due to the viscosity of EcoFlex. The injection channel and air vents were also too small, making it difficult to generate enough pressure to force out the air. To make this mold still useful, the silicone was poured into the mold directly and clamped shut once all the channels were filled.

For the first prototype, the results were mixed – the majority of the channel’s geometry turned out fine. However, there was a large air bubble that was caused by the mold falling over during the curing process, resulting in a leak of silicone through an air vent near the nose bridge area.

After evaluating the performance of the first prototype and deciding to dedicate some time to determining the ideal cross-sectional geometry, smaller molds were made to test other ideas in an efficient and cost-effective manner.

Intermediate Testing: Week 3 - 6

From the previous two prototypes, we learned that the suction channel geometry was incredibly important to creating a good seal. Thus, we went into an extensive testing phase of different suction channel geometries to determine the ideal shape, height, and material.

The goal of this was to determine the optimal cross-sectional geometry for seal quality, comfort, and manufacturability. We tested 5 different cross-sectional geometries with varying features.

The molds for these tests were easier to make and consisted of simple two-part molds.

Second Prototype: Week 6 - 8

For the second prototype, the mold had significantly more complex geometry since the base of the silicone channel both followed the curvature of the NIOSH headform and was oriented normally to all points of that curvature. This resulted in a mold that needed a particular parting line that ran along the outer edge of the silicone flap (shown on the left).

During the molding process, we discovered that the complex geometry of the suction channel resulted in parts of the mold not being able to fit vertically (i.e. geometry did not fit along a single axis, so the two parts of the mold had to be snap-fit together). The mold did end up working, but the final design mold was redesigned to address this.

Final Design: Week 8 - 10

The mold for the final design addresses the issue of the complex geometry by being made of four different parts, as shown in the figure below. Here, the positive part of the mold is split into three separate pieces while the negative part is held intact.

The reason for this is mainly because the parting line of the more complex suction channel resulted in parts of the mold that had caves/recesses around the cheek area that could not be reached by a simpler two part mold. In other words, the design required the mold to have an additional degree of freedom. The assembly process of molding is shown in the figures below.

Page updated

Report abuse