5.4 - Implementation
Fabrication
Linkages
All of the links were laser cut on 1/4 inch acrylic to create a smooth polished design and prevent potential areas of friction where the links could catch and impede the motion of the robot. Additionally because of the expected clamping movement of the mechanism, we wanted to use a material that would be lightweight and not crack or splinter when the mechanism released the rope to move upwards because it created a rough movement.
Complex Parts
We 3D printed the parts that were more complex and not suited for laser cutting or required multiple iterations because we had easier access to 3D printers than laser cutters. This included the battery enclosure, motor enclosure, spacers, components of the gear train, and components of the clamping mechanism. Since these parts required many iterations as we tested the robot, we felt that 3D printing would allow a higher degree of customization with our designs and would let us test different ideas faster as we experimented with different sizes of components, numbers of teeth on the gears, and spacing.
Assembly
Mechanism
After fabricating all of our components, we assembled and tested the robot in tandem. We began by assembling the linkages in the current orientation and placement and ensured manually cranking the mechanism would still achieve the desired movement. At this stage we also experimented with spacers widths and adjusted them as we saw places the movement would catch. We used bearings inserted in precut holes in the laser cut links and pressed steel dowels to hold them together. We also used screws to keep static components together and adjusted the tightness of the screws to control the friction between different links and between the robot and the rope.
Gear Train
Next, we assembled the gear train and battery enclosure which were attached with screws and custom 3D printed enclosures. We inserted the switch into a pre-cut insert in the acrylic link and put the battery in the enclosure. The wires were soldered to the appropriate connectors so that we could control power to the receiver with our switch. We experimented with a few different designs for the gear train in order to generate enough torque from the motor to drive the weight of the mechanism. In the final version we used a 25:1 cycloid gear train with 34 teeth on the motor pulley and 30 teeth on the cam pulley. We chose this configuration for ease of placement and weight distribution while maintaining enough force to power the robot.
Dwelling
After the prototype phase, we added a second cam-follower mechanism to clamp and release the rope. As the motor runs, the mechanism converts the rotation driven by the motor to a linear motion along a slider that clamps and releases the rope. The allows the mechanism to move up the rope while giving us more control over the friction holding the rope than the original 2-pin design in the prototype used. By adjusting how tightly the component at the end of the slider was pressed to the rope, we were able to experiment with different levels of clamping in the rope until we found one that allowed for the smoothest motion. The cam-follower is driven at the same linkage as the crank-rocker, providing simplicity in design by minimizing the number of actuators we need and ensuring the timing of its motion was consistent.
Clamping
The most complex part of this project was determining the ideal balance between maintaining enough force to provide tension and allowing enough room for the mechanism to move. We iterated through many designs as we tested the robot and added two components with teeth to clamp on to the rope. At the top, we added teeth to the clamping component driven by the cam-follower described above which allows for more friction along the rope. This helped solve the issue of the robot moving up inconsistently and allowed constant motion. The second component was added to the bottom of the linkage using a modified rope cleat used on boats powered by a rubber band. It allows the link to move up the rope, but provides resistance when the rope tries to go the other direction, helping prevent the rope from slipping.
Electronics
The electronics were design to be simple to account for the need to carry the weight of all the electronics on the robot itself. However, the weight of the robot was greater than anticipated, so we made modifications to the original plan using components we had access to from previous projects to allow for a greater amount of torque from the motor. This involved adjusting the gear train and using a brushless motor controlled by a Radiolink controller. This provided the additional benefit of being able to control the speed which had a significant impact on the smoothness of the movement. We used a switch connected to the battery that controlled power to the receiver wirelessly connected to the controller. As we adjust the controller, the motor receives varying voltage inputs based on the speed we are inputting. The input from the controller is transmitted to the receiver which sends PWM to the Electric Speed Controller (ESC) which then converts it to an input to the motor.