Hermes I

Custom 3D Printed Swerve Module Designed for FTC Drivetrain Use

Project Overview

The Hermes project originated as a high school engineering assignment and developed into a passion project for my robotics team. I designed a mechanically functional coaxial swerve module using both commercial off-the-shelf (COTS) and 3D printed parts. A key objective was to ensure the design was “easily manufacturable,” emphasizing 3D printing as the primary method.

The swerve module required low backlash due to the precision demands during the autonomous phase of the FIRST Tech Challenge, where robots must execute exact movements. To achieve this, I minimized gear usage and opted for a system using non-planar pulleys and a toothed belt. This approach aimed to eliminate gear-related motion issues, although it introduced challenges regarding tension and pulley radii.

Developments

▹ Designed mechanical prototype entirely in Fusion360 CAD software.
▹ Conducted finite element analyses (FEAs) on the standoff plates and structures.
▹ Performed center of mass (CoM) analysis to identify optimal mounting locations.
▹ Specified parts to meet our robotics team's standards.
▹ Performed calculations for optimal speeds/accelerations for available hardware.

About Hermes I

Beginning both as a project for one of my high school engineering projects and as a passion project for my robotics team there, I designed a mechanically functional coaxial swerve module primarily manufactured with COTS (commercial off-the-shelf) parts and 3D printed parts. For my objectives listed in the project proposal for the class, I even listed one of the overall objectives as "easily manufacturable" and specifying 3D printing as a main manufacturing method.

This project required low backlash due to the precision(s) required in the autonomous period in the FIRST Tech Challenge that requires extremely precise movement of an autonomous, pre-programmed robot to complete a set of actions. For this reason, I tried to minimize the use of any and all gears in the system, including the bevel gears typically used for wheel rotation. Instead, I opted for non-planar pulleys with a toothed belt rotation: I was hoping that by removing the gears in the system altogether, the motors (and the encoders integrated directly into the system) would not be affected by even the slightest of motions, up to the belt tension itself. This system likely would have worked with enough testing, however, rotating belts like this is typically considered bad practice and would need accounting for different kinds of radii in both the pulleys and rotation axes (foreshadowing).
‍

For this project, I designed the prototype entirely in the Fusion360 CAD software where I performed various FEAs on the standoff plates and structures using rough estimations for 3D printed plastic and CoM analysis on the modules to determine several best places to mount these modules. These assemblies were also digitally "jointed" in CAD to be able to both animate and conceptualization the motion of these parts in the physical world, which gave a good medium for determining the module performance.

Several calculations were performed for optimal max speeds and accelerations based on typical FTC drivetrain performances and motor specifications. Each of the parts were carefully spec'd to fit the standards and typical parts used by our robotics team.