The following is a quick overview of some of the contributions that I made to my Two-Month rover team rover: affectionately named Cookies N' Cream.
To simplify the drive train and enhance maneuverability, I opted to design the rover without a conventional steering system. Instead, I opted to guide the group's general design in the direction of a differential steering system. One of the other groups opted for a complex omnidirectional steering system. They had a lot more experience in robotics than my team and were, on average, further along in college. But, due to our conservative design choices, our rover worked while there's struggled to drive at all.
To complete many of the challenges in the competition, the rover had to have an arm capable of pressing buttons and hooking onto a payload (the hook was made of bent coathanger and was not drawn in cad). To simplify the movement of the arm, the only axis of motion I chose to consider was an aproxametly 210° range of motion arc. The other two axes of motion for the arm were covered by the tank steering. Tank steering, or differential steering, allows the rover to turn about its center. Tank steering also enabled us to move the arm forward and backward by simply moving the rover forward and backward. This was definitely not ideal in terms of arm precision, but it was more robust and succinct than any other solution I could think of.
The keyed pin (white) connects the bracket to the arm and transmits the torque from the servo to the arm. The pin has a hole down its center to allow for a screw to fasten the pin in place.
There were a number of challenges in the process of making the arm assembly work together. The arm's pivot had to put up with a good bit of force, but due to the fact that the rover's body is thin, laser-cut MDF, screws weren't going to cut it. I needed to make a design that anchored itself on a large area of the material to make it stronger, similar to drywall anchors. The design of the pivot has the main body of the bracket (grey) made all as one piece, It sticks up through the top layer of MDF, firmly anchoring it in place.
The blue piece slots into the grey bracket holding it firmly in place. The piece's primary function is to fixture the servo and help keep it from moving as it actuates the arm. The same two screws that hold the servo in place hold the blue and grey pieces together.
This bracket was the most over-engineered part of the entire rover. This bracket connects the motors, gearboxes, and wheels to the rest of the rover. It is insanely over-engineered. I could have just made an L-shaped piece with four holes in it and screwed the motor on like all of the other teams, but I decided that making sure that the wheels didn't fall off was more important than my natural engineer laziness.
The bracket completely encloses and protects the motor and connects to the base of the rover at three points, making it strong in all directions. To model this part, I made a complete model of the motor and gearbox together and then offset a surface from the model. I then thickened the surface and added screw-hole tabs to the outside.
This wheel is made of plastic. That one simple fact made me think way too much about what should have been the simplest part of the entire rover. Rubber wheels are grippy; they have a high coefficient of friction. Plastic wheels are not. If I wanted our team's rover to have any chance of gaining traction, I would have to find a way to raise the coefficient of friction. Luckily some engineer ages ago came up with a way to do just that, treads-- turning a friction problem into a sheer force one.
But, there is a problem with treads, rubber squishes and absorbs vibration, plastic doesn't. If I put straight treads on the wheel (making it look like a gear), the distance between the axle and the edge of the wheel would not be constant, which would lead to rapid bumps up and down as the rover drove. My first thought to mitigate this problem was to make the treads helical instead of straight, but that would turn the wheels into screws, pushing them side to side as they drove. I mirrored the helix to fix this final problem, making the side forces oppose each other. And that's how I reinvented the wheel.