For our final project, we were tasked to, first, choose an object on thingiverse to create a negative mold of and then create multiple multiple casts of that object. We had to choose an object that was not symmetrical so that we had to make two unique molds. After flipping through a number of possible objects, we decided on choosing this stormtrooper helmet for our project:

thingi:16776

Once we had our .stl file, we began prepping the file for CNC milling. Preparing the file was definitely the most challenging and time consuming part of this project. We used Fusion 360 to cut the helmet into two pieces, the front face and the back face, and draw keys on our design, which helped us align both halves of the mold during the molding process.

using a plane to slice the helmet in half

front face with backboard and keys

Once we were happy with our design, we used the CAM feature on Fusion 360 to set up our tool paths. At first, we only used the adaptive clearance option but when we ran the simulation, we realized that we couldn’t see many of the details on the helmet, especially on the front face. So, in addition to adaptive clearance, we used both the parallel and pencil options, which proved to make all the difference. With our g-code ready, we started the CNC process.

Importing our g-code onto flashcut CNC proved to be much harder than we expected. Initially, the program didn’t want to even open our g-code so we had to convert our code into a .txt file before we were able to import the code onto flashcut CNC. However, the settings weren’t exactly the way we liked it to be (eg. the bit on the CNC started at a much higher position than we wanted it to be). So, we had to find another way of exporting our file from Fusion 360 so that it would run more smoothly on the program. Dr. Wettergreen’s solution was to use easel to export our g-code and it worked perfectly. Here are some pictures of the CNC milling process:

our tool paths on flashcut CNC

We didn’t run into any major problems when we were CNC milling our piece of wax. We did encounter an issue with the wax melting onto the sides of the mold and the drill bit occasionally but we were able to counter that by stopping the milling every 10-15 minutes and then vacuuming out any access wax stuck onto the sides and by chipping away any melted wax stuck onto the bit. All together, each side of the helmet took approximately 2 hours to CNC (including some stoppage time to clear the melted wax).

We used oomoo to make our negative molds and Smooth On Smooth-Cast 300 Q (liquid plastic) to cast our stormtrooper helmets and added some dye to color our models.

After we made our first stromtrooper helmet, we realized that excess smooth on had seeped into and set in the cracks between our oomoo molds. Luckily, we were able to remove any excess smooth on using a file and sandpaper. Here are our final products:

 

The biggest issue that we noticed on our model was that the front and the back face didn’t align perfectly with each other. We think that an issue occurred when we scaled down each face on Fusion 360. There may have been a discrepancy between the scale ratios of the two faces.

Overall, this assignment was a great way to end the course! The assignment had many components to it (setting up the design, exporting the g-code, CNC milling, and finishing) and while each step of the process had its challenges, we were able to troubleshoot any issues we were facing and create a repeatable process for when we have to do another assignment of this nature. We are pleased with our final product and we had a lot of fun doing it.