15. Mechanical design & 17. Machine Design¶

For this week’s assignment, we decided to build a 2-axis marble maze puzzle.

Input : The objective is to control Marble Maze with hand movement and/or a joystick. Thus the input would be accelerometer mounted on hand for the first control method. and joystick in the second.

Output : The two servo motors will be the output whose rotation would be limited to 20 degrees in each direction.

Inspiration¶

After spending weeks on deciding what to make ranging from an automated flamethrower to a vending machine, we finally decided on making a marble maze. This was our inspiration for the project.

The [amazing] Maze made by Katie Levine and our regional instructor Ohad Meyuhas.

The maze¶

Since this was a group week, we have uploaded the entire prosess on the Group Page.

Here is my contribution in this project.

Assembly¶

• For 2-Axis movement, we needed the gears driving the board in the same vertical axis.
  The below image was our first sketch in making that.

• This was the first small prototype for the same.

• But since the individual axis were open, the system was not stable.
• So for the second attempt, we connected the system at top so both would give stability to the other.
• But in this case, we’d have to add flexibility for the top pivot to move in the 3rd dimension as well. Hence, we replaced the pivot with a 3d printed ball socket joints.

• This is the video of the first attempt of it working with the motors.

• We then changed the design of the base to make it more stable. This was the new conceptual 3d diagram

• These are the lasercut parts for this updated base.

• This is how it turned out to be.

The Final Prototype¶

• After multiple models and iterations, it was time to start making the final thing. This is the cad design for it.

• The gear axis are perfectly aligned here and they are also aligned with the intersection of the axis.

• And this is how it looked after the first time it was assembled.

• The material used here was a recycled one and hence the kerfing came out to be uneven and thus the fitting wasn’t upto the mark. We made a refined version at a later stage. This was used to test out the ball socket mechanism.

• We added a lager base to this to add more stability and we decided to make a maze of a larger size.

• This base was dsigned to not only support the structure from the base but also fron the side using triangular notches.

• The model fit perfectly on this base and gave enough support for it to take the weight of the maze.

• This is how the maze looked after it was up and functioning.

The Final Maze¶

Slide¶

Video¶

The remainder of the process is documented on the Group Page.

Downloadable files¶

You can download all the files here