20.2 - Design Process

Iteration Documentation

This is the initial compliant finger design. It was parametrically designed, so that we can easily alter our design as well as feed design parameters into the kinematics model to exhibit our force and motion profiles for each design.

The first iteration was a basic proof-of-concept used to test the physical construction and motion of the mechanism. In this version, a TPU finger was fixed to a PLA base, and a strip of denim was sewn to the finger and routed through the design to act as the tendon. The finger was fastened into an a PLA component attached to a simple spool, which we manually rotated to mimic motor-driven actuation.

Here is a visual of the manual rotation. This first prototype allowed us to validate the general tendon-driven curling motion, but it also showed that the system was fairly stiff and that the denim tendon was, as you would imagine, not ideal for smooth motion. Also, this design was not made to be automatically actuated, or scaled.

With this next design iteration, we redesigned the pulley to fix directly into the motor, which makes actuation for this finger possible. The interior was made more robust to allow for more torque than the previous iteration.

Unlike the first iteration, this strap was looped around the back and then sewn to create path for the strap to wrap around the distal of the finger. This fixture version is also stronger and more robust, ad the nylon webbing is sewn into itself, and the holding part is embedded in the body of the finger.

In the final second iteration, we moved to a motor-actuated design and made several important changes to improve functionality and packaging. We used a lower-stiffness TPU material and reduced the flexure thicknesses so the finger would bend more easily. We also redesigned the finger geometry to make it smaller and more proportional, while keeping the same overall length. In addition, we switched from denim to nylon straps for tendon routing, since the nylon interfaced better with the pulley fixture and moved more smoothly through the finger slots. This version was mounted to a PLA fixture and designed specifically for motor actuation, making it a more realistic prototype for the final gripper direction. We also made the design slimmer and more modular so that additional fingers or straps could be added later.

Although the second iteration improved compactness and actuator integration, it also revealed a new problem. Because the flexures were thinner, the TPU was less stiff, and the tendon slots were narrower, friction became too high and the finger became overly compliant. As a result, the finger did not reliably return to its resting position after curling. This iteration was still valuable because it highlighted the tradeoff between compliance and recovery. Based on this result, the next iteration should increase stiffness by either using a stiffer TPU, increasing flexure thickness, widening the tendon slots, or combining those changes. Overall, these iterations helped us move from a simple manually driven concept to a more integrated motor-actuated prototype while identifying the main design changes needed for the next build.

Bill of Materials

Sources

[2] L. L. Howell, Compliant Mechanisms. New York, NY, USA: Wiley, 2001.

[3] L. L. Howell and A. Midha, “A method for the design of compliant mechanisms,” Journal of Mechanical Design, vol. 116, no. 1, pp. 280–290, 1994.

[5] S. Timoshenko and J. M. Gere, Mechanics of Materials. New York, NY, USA: McGraw-Hill, 1972.