Experimental investigation of the inversion/eversion motion for the design of a two degrees of freedom ankle prosthesis

Bipedal walk is a very complex gait. Lower limb amputees know this very well since it is challenging to recreate healthy ankle motions with a robotic ankle prosthesis. It always leads to some comfort and stability reduction, which often limit the daily life activities participation for amputees. In order to propose some guidelines that can improve the design of two degrees of freedom ankle prosthesis, this master thesis focuses on the analysis of the inversion/eversion movement of a healthy ankle. In a first step, the normal life outside conditions were characterized to be able to recreate similar situations inside the lab. Therefore, an electromechanical device has been developed to measure ground angles that are reached during a normal walk in the street as well as the ankle inversion/eversion angle during some daily life movements. The device is designed to fulfill the specifications needed to allow accurate measures without disturbing the user. These preliminary tests enable us to acquire experience in the experimental domain. They allow us to identify which kinds of gait use the ankle inversion/eversion degree of freedom. These results were used to select the moves analyzed further in the lab. The second part of this study focuses on the in-lab measurements of the inversion/eversion of 4 gaits selected on base of the outside conditions observations: turning step, lateral step, 15° inclined plane step and 30° inclined plane step. The torque of inversion/eversion in the ankle is computed for these different gaits in order to analyze the behavior of a healthy ankle in the frontal plane. The results reveal that the inclined plane steps use more the inversion/eversion movement than the normal walk. The behavior of the ankle in the frontal plane appears to be correlated with the inclination of the ground. In the turning and lateral steps, the inversion/eversion movement was comparable to that in the normal walk. The determination of the characteristic impedance shows a linear profile for all gaits and was very comparable between the two inclined plane tests. The computed power involved in the frontal plane shows a correlation with the sagittal plane. Finally, some guidelines are proposed for the future design of an assistive prosthetic device targeting the eversion/inversion motion, in order to have the best two degrees of freedom robotic ankle prosthesis as possible. The characteristic impedance being linear, a semi-active or even a passive system may be possible.