Concept generation
See this page for the constraints established for concepts: https://wiki.uwaterloo.ca/x/9gQuD
Assumptions:
- Current chassis configuration remains the same: same vertical height and offset from outer wheels
- Assume the wheels do not give in/sink
- Assume level ground
- to reach a height of 4.5 feet off the ground, 2 links with a length of 600 mm each connected end to end give a comfortable amount of working space at that height. (about 450mm)
- This sketch really only works if the last two axes (axis 5 and axis 6 are stacked on top of each other in the pumpkin.
Concepts:
Mathieu
Ethan
Austin
Might have missed the mark on this one a bit, instead of multiple sketches i kinda just focused on one fully fleshed out arm design. I'll try to generate more before the meeting tho.
BIGBELTTM Overview
Axis 1
Motor in red, coaxial gearbox (cycloid or planetary) in green, belt driven parallel shafts in blue, encoder in pink. Two stage reduction, with second stage belt drive reducer allowing for easy encoder mounting, greater reduction ratios, and less backlash due to tensioned belt. Shafts will probably have to be strong, but housing should be relatively light and easy to make. Output hub will connect to axis 2 using two holes and rotate using bearings on shoulders. Easily 270 degree rotation.
Axis 2
Same kind of deal here, with a motor → coaxial gearbox stack up onto a second stage belt drive. Theoretically can achieve high ratios. IK we said that we wanted linear actuators on this axis, but after modelling some stuff out i think this design may also be viable as it allows for a large degree of rotation on this axis.
Axis 3
I moved the linear actuator to axis 3, as we did establish that we wanted to use this guy. I used the same model of linear actuator currently used on the arm, altho better options may be available. Looking back, rear mounting this actuator is prob better, but might add a lot of weight cuz then the rear will need space for a reducer and motor. Encoder is placed on the pivot point for readouts. Range of motion is approximately 45 degrees with this initial design, but I don't know if we need much more with ROM for this joint for our tasks.
Axis 4
Axis 4 features no real belt drive, and just has a one stage coaxial reduction as I think that the torque requirements on this joint will probably be lower. Gearbox and motor is mounted on the back of the joint housing, and there is a large interior shaft connected to output hub to allow for 360 degree rotation of this joint. There is a cutout in the housing to allow for a belt to go over the shaft, and connect to a top mounted shaft. There isn't gonna be a reduction ratio here, it should be 1:1. The purpose of this is to allow for an easy mounting spot for our encoders , as with this design there isn't really a good spot to mount encoders, unless they are stacked up on the inside of the shaft (but I don't think most encoders that we have access mount like that) Minor note, some stuff changed in pics. Last second changes lol.
The big blue ring is the output shaft part that actually rotates. Theres gonna be shoulders and bearings or whatever to account for thrust loads in this shaft in a more refined version of the design.
Axis 5
Very simple belt driven design that rotates with axis 4. Not too sure if an additional gearbox would be required here, as that obviously lends to higher reduction ratios. If needed, it could be integrated somehow. Belt drives aren't necessarily required here, but they are a good option as they should have low backlash when tensioned correctly. Using belts also lets you "pull back" the motor for axis 5 away from the joint. Gonna get approximately 210 degrees of rotation here.
Axis 6
360 deg. rotation, powered by belt drives once again lol Once again, I think having parallel shafts for axis 6 and it's motor just allows for super easy encoder stack up. Encoders can fit in a slot underneath/behind the axis of rotation. Output hub for axis 6 will be retained with bearings and shoulders for thrust loads.
Workspace Overview
Design cannot fully "fold" over itself due to limited range of motion on linear actuator, but this can be addressed thru better design lol. Below is max "fold"
Easy floor access, but there definitely is a limitiation to the horizontal range. In pink, this space is pretty much inaccessible. However, this also means we can mount the arm a lot further back on the robot and not on the edge of chassis
Max to min horizontal floor reaching range
Max horizontal range
Pros
- Very simple (IMO). Hardest part would be getting our gearboxes designed correctly
- Should be easy to take apart for individual joint testing as long as we are careful with the design. Should be able to take of the belt then unscrew the mounting flanges at most joints, although taking apart the linear actuator may be a little more annoying.
- Might be able to run all the wires thru structural tubing if we set up slots at joint connections
Cons
- With the inital design, it relies a lot on somewhat complex joint connections. These will probably be aluminum, so the arm might get heavy. Further design refinement may be able to address this tho.
- Somewhat small workspace due to linear actuator restrictions, but for our comp its prob fine
Claw Sketches
Jasmine
Alik
Medina
