Development and Testing of an Unpowered Ankle Exoskeleton for Walking Assist

Abstract

Assistive technologies traditionally rely on either strong actuation or passive structures to provide users with increased strength, support or the ability to perform lost functions. At one end of the spectrum are powered exoskeletons, which significantly increase a user’s strength, but require strong actuators, complex control systems, and heavy power sources. At the other end are orthoses, which are generally unpowered and lightweight devices that rely on their structure’s mechanical behaviour to enhance user’s support and stability. Ideally, assistive technologies should achieve both systems’ characteristics by enhancing human motion abilities while remaining lightweight and efficient. This can be achieved by using distinctive actuators to harness gait energy, towards enhancing human mobility and performance.

Pneumatic Artificial Muscles (PAMs), compliant and flexible, yet powerful and lightweight, present a unique set of characteristics compared to other mechanical actuators in human mobility applications. However, given the need of a compressor and power source, PAMs present a significant challenge, limiting their application. In contrast, PAMs can be implemented as unpowered actuators that act as non-linear elastic elements.

This thesis aims to develop a wearable lightweight unpowered ankle exoskeleton, which relies on the PAM to harness gait energy and compliment the human ankle biomechanical abilities at the push off movement, thusly assisting the user in propelling the body forward during walking. Presently, limited PAM models have been developed to analyse PAM passive behaviour and to assist in designing and selecting the appropriate PAM for unpowered application. Thus, this thesis aims to develop a passive model for the PAM.

To mechanically validate the proposed exoskeleton design, a prototype is fabricated, and tested within an Instron tensile machine setup. The unpowered exoskeleton has shown its ability to provide significant contribution to the ankle timed precisely to release at the push off phase of the gait cycle. Furthermore, the proposed PAM stiffness model is validated experimentally, and accounts for muscle pressure, geometry, material and stretching velocity. This enables the evaluation of the impact of various parameters on the muscle behaviour and designs the PAM accordingly for the unpowered ankle exoskeleton