Using 3D printing to manufacture end-use high quality robotic tires.
ADDRESSING THE NEED
This project started as a pursuit of a manufacturing a better tire for our robotic swerve drive. While I was on the FRC team 33 Killer Bees, we had built a custom swerve drive for the 2021 at home challenges. Because this platform was powered by Falcon motors on 3in wheels, we really started struggling with traction, as the tires couldn’t keep up with the power of the motors, despite having a very light frame to help agility. At the time, the two commercial solutions consisted of riveting blue nitrile tread onto wheels, which created an inconsistent contact patch because of the rivet heads, or adapting colson tires to work with the swerve design.
DESIGN CREATION
I started off toying with the idea of using my recently built dual extrusion 3D printer to print both the wheel and tire as a whole, as shown in Fig. 2 below. This is actually a replacement tire for a common rc car, something much easier to test. These tires were a great point to start, but we quickly realized their drawbacks.
Firstly, printing them as a single piece not only took too long but was too specialized. Ideally we would be able to swap tires independent of the rim, so the dual extrusion wasn’t the right application here. Secondly, the flexible TPU that we used was way too hard, we needed a softer compound for more grip. Finally, while the airless “fin” design was cool, it didn’t leave much for the practicality of tread wear, so that was out the window. Therefore, from here we moved on to the second iteration, shown in Fig. 3 below.
This was our first tested design. It consisted of a two-piece clamped wheel which was printed in blue nylon here, as well as a 90A studded TPU design for the tread, both originating from my mentor at the time Nick Coussens. This worked pretty well, and was definitely a step in the right direction, but the Nylon was not rigid enough and the TPU still needed to be softer.
Then, I took the two piece clamped idea for the wheel, but changed up the tread design as shown in Fig 4. on the right. This tread design would allow for a larger contact patch for the tire, allowing for more tread. This tread idea was quite promising, and we tested this extensively, but we still thought that a softer TPU would give us a better performance.
While there was a bit more testing and a few different tweaks from here, this idea moved us to the final design shown in Fig. 5 below. For this design, we wanted to get the softest TPE compound that we could find. After doing some research, I found a company out of Spain called Recreus who made a 60A shore hardness TPE that was more grippy than any substance that I have ever printed. And surprisingly, it wasn’t that hard to print at all, with a little bit of tuning on my machines, it was easily manageable with no stringing at all. The tires that resulted from this are still in use and continue to supply superior traction for robotic tires.
