A novel hydraulic energy-storage-and-return prosthetic ankle : design, modelling and simulation

In an intact ankle, tendons crossing the joint store energy during the stance phase of walking prior to push-off and release it during push-off, providing forward propulsion. Most prosthetic feet currently on the market – both conventional and energy storage and return (ESR) feet – fail to replicate this energy-recycling behaviour. Specifically, they cannot plantarflex beyond their neutral ankle angle (i.e. a 90° angle between the foot and shank) while generating the plantarflexion moment required for normal push-off. This results in a metabolic cost of walking for lower-limb amputees higher than for anatomically intact subjects, combined with a reduced walking speed. Various research prototypes have been developed that mimic the energy storage and return seen in anatomically intact subjects. Many are unpowered clutch-and-spring devices that cannot provide biomimetic control of prosthetic ankle torque. Adding a battery and electric motor(s) may provide both the necessary push-off power and biomimetic ankle torque, but add to the size, weight and cost of the prosthesis. Miniature hydraulics is commonly used in commercial prostheses, not for energy storage purposes, but rather for damping and terrain adaptation. There are a few examples of research prototypes that use a hydraulic accumulator to store and return energy, but these turn out to be highly inefficient because they use proportional valves to control joint torque. Nevertheless, hydraulic actuation is ideally suited for miniaturisation and energy transfer between joints via pipes. Therefore, the primary aim of this PhD was to design a novel prosthetic ankle based on simple miniature hydraulics, including an accumulator for energy storage and return, to imitate the behaviour of an intact ankle. The design comprises a prosthetic ankle joint driving two cams, which in turn drive two miniature hydraulic rams. The “stance cam-ram system” captures the eccentric (negative) work done from foot flat until maximum dorsiflexion, by pumping oil into the accumulator, while the “push-off system” does concentric (positive) work to power pushoff through fluid flowing from the accumulator to the ram. By using cams with specific profiles, the new hydraulic ankle mimics intact ankle torque. Energy transfer between the knee and the ankle joints via pipes is also envisioned. A comprehensive mathematical model of the system was defined, including all significant sources of energy loss, and used to create a MATLAB simulation model to simulate the operation of the new device over the whole gait cycle. A MATLAB design program was also implemented, which uses the simulation model to specify key components of the new design to minimise energy losses while keeping the device size acceptably small. The model’s performance was assessed to provide justification for physical prototyping in future work. Simulation results show that the new device almost perfectly replicates the torque of an intact ankle during the working phases of the two cam-ram systems. Specifically, 78% of the total eccentric work done by the prosthetic ankle over the gait cycle is returned as concentric work, 14% is stored and carried forward for future gait cycles, and 8.21% is lost. A design sensitivity study revealed that it may be possible to reduce the energy lost to 5.83% of the total eccentric work. Finally, it has been shown that the main components of the system – cams, rams, and accumulator - could be physically realistic, matching the size and mass of the missing anatomy