Bajelan, Soheil ORCID: 0000-0001-5526-401X (2020) Biomechanical design and evaluation of a Self-Powered Ankle Exoskeleton. PhD thesis, Victoria University.

Abstract

Walking is critical to many everyday activities, but it can be impaired by conditions including ageing, neurological disorders and muscular pathologies. One of the most serious consequences of gait impairments is the inability to maintain the adaptive lower limb mechanics typical of the characteristic, unimpaired swing foot trajectory. A principal requirement for walking safely is maintaining clearance between the forefoot and ground at critical events in the walking cycle. Foot-ground clearance, represented by the vertical component of the sagittal swing phase trajectory is, therefore, associated with tripping probability. High risk foot trajectory control can be mitigated by providing an external assistive force to lift the foot and a variety of ankle assistive devices have been designed to assist those with walking impairments. Both passive (unpowered) and active (actuator powered) Ankle Foot Orthoses (AFOs) have been evaluated. Passive devices are low-cost, light and mechanically uncomplicated but tend to restrict ankle motion and reduce plantarflexion at push-off. Active devices have the advantage of regulating the actuator’s timing and intensity. It enables them to overcome ankle restriction during stance but are disadvantaged in being more costly, bulky and mechanically complex. An alternative approach to low cost, lightweight exoskeletons is progress toward a “minimal device”. Passive exoskeletons are preferred due to their simplicity, durability, customizability, affordability, compactness, lightness and ease of use; but two challenges limit their practicality. The first is how to produce assistive forces without an actuator and external power source and the second is controlling the timing and magnitude of assistive force without sensors and control units. A further important requirement in designing such devices is detailed information concerning the swing kinematics and kinetics of ankle-controlled walking. A substantial literature has documented stance phase biomechanics, but the swing phase literature is limited and has rarely been employed in the design and evaluation of AFOs. To address the above device design and biomechanical challenges three aims were; (i) investigate the effects on kinematic and kinetic gait variables of foot trajectory modulation, using either the ankle joint only or walking without ankle joint motion, (ii) design, construct and test an ankle exoskeleton which would be able to modulate swing foot trajectory using heel strike energy harvesting and a mechanical controller and (iii) evaluate the constructed exoskeleton’s effects on lower limb swing phase kinematics. Accordingly, three studies were designed to realise the construction and evaluation of the ankle-assisting device described in this project. The first was an experiment to predict kinematic and kinetic effects on gait mechanics of intentional ankle-controlled walking. Using a real-time biofeedback, swing foot trajectory was displayed on a monitor and a range of predefined target foot-ground clearances were projected. Participants were then asked to walk on a force-sensing treadmill while matching the pre-defined clearances in two experimental conditions; (i) using either the ankle joint only or (ii) achieving the target foot-ground clearance with no ankle joint modulation. Intentionally ankle-controlled walking reduced the hazardous Minimum Foot Clearance (MFC) event by increasing foot-ground clearance and, in some cases, eliminating MFC. Ankle-controlled walking also decreased swing phase time to MFC and foot maximum horizontal velocity, with no effects on gait symmetry. Kinetic analyses using AnyBody, showed no significant increase in ankle moment required to lift the foot using a highly dorsiflexed ankle, but greater tibialis anterior muscle force was required. Moreover, increasing the foot-ground clearance by using the ankle joint only showed less mechanical energy than knee or hip action. Design and construction of the Self-Powered Ankle Exoskeleton (SPAE) was then performed using the kinematic and kinetic variables derived from the first study by adapting a systematic engineering design procedure (second study). The design was then evaluated with a preliminary single-subject test, showing that when walking the SPAE could successfully harvest adequate energy during heel strike and actively dorsiflex the ankle during swing, with minimal gait disturbance. A second experiment (third study) evaluated the final SPAE design. SPAE-assistance increased the vertical component of the swing foot trajectory, incrementing MFC and tending to wash-out the hazardous MFC event, as reflected in the Mx1-MFC height ratio. SPAE-controlled walking did not restrict ankle, knee and hip joint motion and showed no effect on the unassisted limb kinematics. The project demonstrated, therefore, that a biomechanically designed SPAE could provide functional active ankle assistance during swing, without an external energy source or electronic control system.

Search Google Scholar

Repository staff login