An image of the 3D printed squeezing component used for molding to create a cavity that is the shape of what we want.
3D printed squeezing component for use in molding
An image of the non-sacrifical liquid plastic mold made to take on the shape of the 3D printed ovular squeezing component
Polyurethane mold of the Squeezing Component
An image of the black angled insert being cast inside the non-sacrificial plastic mold using Mold Max 30
The angled insert being cast inside the mold using Mold Max

After a check-up meeting with Lai, we realized that using a one-piece mold that is permanent would save us more time and resources compared to a one-piece mold that is sacrificial. For the permanent mold, Lai recommended that we try out the plastic materials that we have. We ended up choosing Smooth Cast 305, as it has a relatively fast cure time of 30 minutes, and is a hard plastic that we can use multiple times. We also realized that there were not any inhibitions or issues with using the Smooth Cast 305 alongside the casting material.

We cast our angled insert (shown in the pictures above), and also the cylindrical insert with the 4 indents, using Mold Max 30 in a liquid plastic mold. Although we initially planned to use OOMOO as the casting material, it was out of stock, and Professor Lai mentioned more was on order. Not wanting to delay our prototyping, we reviewed our materials list for alternatives with a Shore hardness of 30A. As its name suggests, Mold Max 30 met this requirement, so we decided to give it a try. Fortunately, it turned out to be even more suitable than OOMOO, offering improved squeezeability. As a result, we decided to pivot and adopt Mold Max 30 as the primary casting material for our device.

After waiting 24 hours—the cure time for Mold Max 30—we attempted to remove our final molded and cast product from the non-sacrificial plastic mold. However, the rigid nature of the plastic mold made it impossible to coax the silicone out. It was completely stuck. We were running out of options and were reluctant to consider this attempt a failure, so we approached the Nolop staff for possible solutions.

The first suggestion from the Nolop staff was to drill a nail into the black plastic insert and then try to pull both the insert and the silicone cast out together. However, this plan failed because the silicone and insert were so tightly wedged that, when force was applied, only the nail came out, leaving everything else still stuck in the mold.

We consulted additional Nolop staff, and someone with substantial molding and casting experience suggested bandsawing the one-part plastic mold into two equal halves. This would allow us to open it and remove the contents. The idea seemed promising, so we went ahead with it. Once the mold was split into two halves, it became much easier to claw the silicone out of the plastic mold.

An image of the plastic mold after it has been badnsawed in half. The pink silicone cast and green insert are still visible inside.
The plastic mold after it was sawed in half, with the silicone + insert still inside

After successfully removing the silicone from the mold, we realized that the sawed plastic mold could now serve as a functional two-part mold for future iterations. This realization offered a practical solution to the problem we had faced earlier about being unable to extract the silicone from a rigid, single-piece mold after curing. With the mold now split into two halves, we could simply separate the parts after each casting, making it much easier to release the cured silicone from the cavity without damaging either the mold or the cast. We also discovered that a large quantity of mold release spray was absolutely necessary to ensure that the silicone and the plastic would not bond to each other too tightly.

An image of the makeshift eyedropper device we made after the two silicone halves were sawed into half. A rubber band holds the two parts together.
The makeshift device we made by tying together the two halves of the bandsawed squeezing component

We also received encouraging results from the silicone squeezing component. Despite the silicone cast being cut in half, we were able to use rubber bands to hold the two halves together, creating a functional, makeshift prototype. Each team member took turns squeezing the structure to dispense eyedrops, evaluating how difficult it was to generate sufficient pressure for proper dispensing. During this testing, we realized that the pressure points were effectively helping to transfer more of the applied force onto the eyedrop bottle. Wanting to further optimize the design, we quickly 3D-printed two hard PLA shells to fit over the extruded pressure points. These rigid additions provided extra support, allowing the user to press more firmly against the bottle. As a result, squeezing the bottle became noticeably easier, making the dispensing motion feel almost effortless. This improvised prototype gave us a much-needed boost of confidence at a point when we were beginning to doubt whether we were heading in the right direction with our design.