MATE 2018
Carbon Contingency
The third year of MATE we switched from the acrylic frame to a modular carbon fiber/3D printed chassis. This was done primarily because carbon fiber was readlily avalible and more astheticly pleasing, the weight we saved was offset by the addition of two more T-100 Thrusters which provided our sixth degree of freedom.
The ROV was rechristened the “Carbon Aurelia” since it was still the same hardware package as the Aurelia with the only substantial change being the frame.
Because of the print orientation of these early pieces, they were being used in their weakest load direction and would quickly snap. The two pins to the right show the first rendition (blue) next to the final version (turquoise) which I printed flat on the print bed and added a fillet to every surface. The final part versions were incredibly strong and did not break.
The rendering below is from my SOLIDWORKS assembly model and took hours of trial and error to get to this point.
All the 3D-printed blue and green parts were designed in SOLIDWORKS and printed with donated ABS filament. I had a great time designing these parts and spent hours improving and revising them. I learned a lot about how to design and orient the part during printing to minimize the effects of the inherent structural weakness of a laminated plastic part. The thruster mounts were first printed flat on the print bed as shown in the below two pictures.
Working with carbon fiber presented an unexpected challenge. It was difficult to drill due to the tendency of the fibers to frey or the tube snap. We overcame this by using sharp drill bits and thoroughly de-burring the holes after drilling at a snail’s pace. We used a tile saw to cut the tubes which worked very well but still put a cancerous powder in the air so an N-95 mask was used whenever cutting or drilling.
The images to the left and bottom are from initial testing of the thruster configuration. The four internal thrusters were flipped to the outside in the final rendition.
The control system was completely redesigned, we onboarded an enthusiastic software guy who convinced us to use a Javascript microcontroller called the Espruino. The Espruino was a mixed bag, on one hand, it added control latency to our system and was quickly overloaded with commands. The advantage was its compatibility with the custom control app on our ThinkPad control station. This allowed us to use commercial game controllers and even a 3D mouse. I was always in favor of using the 3D mouse because our ROV had 6 degrees of freedom and all six axes could be used at once with one hand.
