Tricopter chassis v0.2 complete
The new chassis is now complete!
Changes from the SolidWorks model:
Instead of plastic (delrin, HDPE, ABS, whatever), I used 1/2″ thick MDF I found in a scrap heap near the robotics lab. This was good because I was able to obtain a sizeable supply of good raw material for free, but MDF has a much weaker tensile strength compared to most plastics and is more slippery in contact with polished aluminum. I solved the latter problem by putting double-sided tape between the MDF and the aluminum to make the tube stay, but I have been unable to do anything about the former problem except to reinforce the bolt holes with superglue:
In the same scrap pile in which I found the MDF, I found some 3/16″ plywood. I had modeled the chassis to use 1/8″ plywood—which would have been just thin enough for me to use my #2-56 machine screws—so I had to massively countersink the holes in the top plate to make the screws reach through to the other side. I reinforced these holes with superglue, as well.
I made the motor mount sleeves out of delrin (as designed) so they would not break like the MDF did.
There are now two pairs of slots in the bottom plate of the chassis through which I have put 1/2″ wide velcro straps (for the battery). Although not shown in the picture, I also added some foam at the base of the velcro straps so that the battery can be mounted with some pressure and friction. These should be more convenient to use compared to zip ties.
My motivation for building a new chassis comes largely from the fact that my old chassis broke after crash landing on rigid plastic feet. To help prevent a reoccurrence of the event, I decided to install shock-absorbing landing gear. After considering several possible designs—springs (complicated), foam (godawful ugly), flexible plates (simple and effective, a labmate’s suggestion)—I chose to use the Mars Rover Team‘s leftover 1/32″ thickness 3/4″ width steel strips.
Steel is hard! When drilling holes, I had to be careful not to break my drill bit, but I had to lower the drill bit fast enough for it to bite into the steel. Otherwise, the bit walked all over the place (much more so than it would on a softer material like aluminum). After drilling two holes at one end of each of the strips, I bent it in a vise, bent the longer end into a curve, and bolted them to the bottom plate.
In the current configuration, there is about 1.5″ of clearance from the bottom of the battery to the ground, which can be changed by bending the steel strips some more. The three strips themselves weigh 68 g in total, which is reasonable for the functionality they provide. They will not prevent things from breaking if the tricopter crashes to a hard tiled floor from eight feet up, but they should at least minimize the small but frequent stresses of landing.
All in all, the tricopter weighs 2150 g, which is actually about 100 g lighter compared to the old chassis. Admittedly, things could be a lot lighter still, but I am going to avoid premature optimization.