🤖Week 9: MAchine Design

Project Heros

Finished body of robot:

My Contribution

Outside of the collaborative planning of the project, my contribution to the project was the mechanical design of the body of the robot. Below is that relevant section:

Mechanical Design

We started by desiging the hubs that would hold the 2020 chassis together and allow us to mount the wheels and the motors to them. We figured that if we left enough 2020 exposed on the chassis, we would have no issue figuring out how to mount the esky.

The mount was designed as it was modelled in onshape, starting with the wheel. We had very slow motors and decided to max out our printer's build volume and make the largest wheels we could to remove the need for any gearing - meaning we could directly drive the wheels with the motor. The outter design of the wheels where they contact the ground were copied from that of a dirt bike (but without the directional V shape as we wanted it to turn both ways). The inner spokes were chosen so that when the model is sliced, it will have some nice solid walls for support and not have to rely on infill. The axle was also modelled into the wheel.

Then the hub to connect the 2020 extrusions was modelled. As with all parts it was designed to be printed without support. This hub mounts to the 2020 extrusions, and holds the axle of the wheel with the bearings.

On the end of the axle, a cap (the yellow thing) bolts to it with 3 long bolts. The wheel firmly hugs the hub and the bearing on the outside, and this cap hugs the bearing on the inside, meaning that the wheel and axle are locked in place. We also modelled the 6mm Polulu mounting bit for the motors

And finally we modelled the motor and a mount for it. We decided to make the motor mount a seperate piece that could independantly slide along the 2020 as it improved printability and would make it easier to assemble. And with that, our CAD phase was done for now as we would need 4 identical sets of these parts to make the robot.

3D Printing

With long print times ahead for the wheels, we decided to utilise the Cura Ultimaker 7 we had access to at work and printed them overnight. We chose PETG as it has improved mechanical properties over PLA, and is still quite easy to print. We printed everything on the thickest layer height of 0.3 mm and chose a rectilinear infill of 8% as these parts were mostly reliant on wall thickness for strength (which was left at the defaults).

The only different part to slice was the wheel which the top 30 mm was modified to have a 60% infill. This is because the cap mounts to the wheel's axle with 25 mm long bolts, and this is how power will be transfered from the motor to the wheels, so a little extra reinforcement was chosen.

Another issue with the wheels is that the axle is printed parellel to the Z-axis, meaning that it runs along the weak orientation as a result of the print lines. This is an issue that will be fixed by inserting the 12mm aluminium rods through. We have a cool process of how this is done coming up.

And after about a week of printing, all of our parts were done and ready to be assembled.

Construction

Before we could build it though, we hit a terrible snag. The 3D printed parts were left in a car during a terribly hot summer day and all the printed parts had warped a little. The 2020 extrusion no longer fit in the hubs, and the wheel's axles had bent over a little in the heat.

However, after a few cycles of heating and cooling in the oven and with a heat gun, the parts mostly returned to their original states as they seemed to have a bit of "memory" about their original shape.

After that process, we could start building. The first thing was to sink the 12mm aluminium tube into the centre of the tyres. There was 1 wheel that wasn't in the car in the heat, and it was easy to sink the tube into by cooling it in the freezer for 20 minutes, then malleting it into position. When it reheats and expands, it braced against the inside slot in the wheels axle.

The other wheels though had warped too much in the heat and this trick wouldn't work, so i heated the up with the biggest soldering iron I had, and sunk them in like threaded inserts.

Then the bearings were malleted into the hub.

And the Cap and shaft hub mounted on top.

And then 2020 extrusion was mounted, as well sa the motor and motor mount.

This process was repeated until the chassis was complete.

We were quite happy with how this came out, and it was a great excersice in not only CAD design, but CAD design for printed parts and desiging assemblies for ease of construction.

This where the niceties of the project ended as the deadline aproached and we had to mount the esky and electronics. We intially planned to epoxy some mounting points to the esky which would mount to the 2020 extrusion, and a similar idea for the electronics. However, we achieved the same result (close enough) with a large amount of duct tape and cable ties.

And with the final addition of some googly eyes, the mechanical phase was finished.