Manufacturing and Assembly - Group 10

Due to time constraints, we started prototyping and assembling the mechanism as soon as the subsystem design was finished. The first subsystem we assembled was the finger mechanism, then came the slider mechanism, and finally, the cam and follower mechanism was assembled.

From the first iteration of the finger mechanism, we found out that the compliance of the bearing that went into the finger mechanism that allowed rotation for positioning created a considerable amount of play in the finger mechanism, especially due to the weight of the steel rail. The compliance on the bearing caused up to 3mm of vertical movement at the end of the finger mechanism. Since our target value for the vertical movement of the finger was 3mm, the compliance from the bearing was too big for the mechanism to work. To fix this issue, we redesigned the bearing mount so that there are two fixture points of the shaft that serves as the axis to rotation using a U-shaped part. This eliminated the vertical play at the end of the finger. 

Another issue with the finger mechanism was that the tip of the finger kept rotating about the rail it slides on because the interface between the slider mechanism to finger mechanism was a ball joint for added degrees of freedom. To address this issue, we added another shaft. Having 2 shafts minimized the rotation of the end of the finger, and thus achieved a more accurate position.

Moving on to the slider mechanism, we had to print multiple iterations to achieve the tolerance we wanted for the bearing press fit. This was crucial because the tolerance on the bearing interface influenced the overall rigidity of the slider mechanism. In the end, we were able to create the right dimension for the press-fit that we are able to assemble and have minimal play in the overall structure. The second issue with the slider mechanism was that the bearing compliance caused the link that positions the finger mechanism to bend downward. We added an acrylic beam underneath the link to support it from excessive deformation.

Lastly, we assembled the cam and follower mechanism. The initial design of the follower was a cantilever beam-type design, with the follower tip at the end of the beam. Due to the inertia of the mechanism from its mass and the friction due to the surface roughness of 3D printed parts, when we tried to move the mechanism by rotating the cam, the follower beam kept bending. This issue was solved by adding bottom support for the left cam and top support for the right cam. The support location is different for the left and right cams because the cams rotate is the same clockwise direction, and the follower touchpoint was opposite for the cams. To further improve the smoothness of the mechanism, we added fillets on the cam path edges.

Also, the mounts for the cam were noticeably deforming as we applied torque on the cam to operate the mechanism. We redesigned the rear mounts of the cams as one piece for added structural rigidity and kept individual mounts for the front cam mounts for ease of assembly and adjustability when syncing the 3 cams together.

In the final iteration, which is the 14th iteration, we were able to achieve a smooth mechanism that typed ‘H’, ‘E’, ‘L’, ‘L’, ‘O’, ‘Enter’, and back to ‘H’. We attempted motorizing the mechanism, but even the minimum speed the motor was able to operate was too fast. We were concerned that the high angular velocity would result in the failure of the 3D-printed gears that are integrated into the cam design due to the high impact force applied at the start of the operation. Therefore, we chose to eliminate the motorization.