【Prosthetic Hands】11 Second Generation - Relocation of actuators and Double 8-shaped Wiring

There were two problems to be solved -- the insufficient grasping strength and the too-long forearm. The higher grasping strength (or torque output) generally means larger actuators (especially for low-budget actuators), whereas shortening the forearm means that the volumes occupied by actuators needed to reduced. To simultaneously satisfy these two features, it is inevitable that the number of the actuators be reduced.

I reduced the number of actuators to 6, eliminating the active DoFs on the PIP. Nonetheless, the flexing movement on PIP joints is still vital for the hand to grasp objects tightly, so I tried to link the movement of PIP joints to MCP joints. 

Most of the 3D-printed hands that can be found on the Internet link the phalanx through simply attached the wire to the distal phalanx, with the other side attached to the actuators. This type of link is simple but  

Hand by Kenneth V. (https://www.youtube.com/watch?v=KSP4o_WCqVs)

However, the relative movement of the phalanx is in fact only determined by the resistance in each joint and is thus uncertain. The movement of the phalanx varies when the hand is placed and oriented differently, which is inconvenient for researchers to model the hand in computer environment. More significantly, the uncertainty of finger movement damages the grasping integrity. An inappropriate finger shape may cause the hand to grasp with the dorsal side of the phalanx, which could potentially result in dangers.

Here I refer to the 8-shaped wiring developed by my Lab. seniors and designed a new "double 8-shaped wiring". The idea was simple: using two sets of 8-shaped wire to link every phalanx so that when the MCP joint moves, PIP and DIP joints would move as well.

This mechanism possesses a lot of advantages over the simple one. When applying this mechanism, every movement of the finger parts are defined by the pulley radii. By changing the radius ratios between different finger parts, I can determine the relative movement of the phalanx. Moreover, this mechanism eliminates the affect of resistance between the channel surfaces and the wire. Last but not least, the 8-shaped wire makes it possible to implement non-slip materials in the middle of the phalanx, which can help grasp objects.

After I made sure that the reduction of actuators was achievable and the missing joint movement could be compensated well with the new mechanism, I redesigned the hand as follow.

In the next article, I will reveal the final design and display some videos and photos.