A Loosely-Coupled Passive Dynamics and Finite Element based Model for Minimising Biomechanically Driven Unhealthy Joint Loads during Walking in Transtibial Amputees

The primary objective of a prosthetic foot is to improve the quality of life for amputees by enabling them to walk in a similar way to healthy individuals. Amputees su˙er from health risks including joint pain, back pain and joint inflammation. The aim of this thesis is to develop a new computational approach to reduce the likelihood of biomechanically driven joint pain in transtibial amputees resulting from sustained exposure to Unhealthy Loads (ULs) during walking. This is achieved by developing a computational methodology to achieve a customisable sti˙ness design solution for prosthetic feet so that the occurrence of unhealthy joint loads during walking is minimised.It is assumed that the healthy population is able to spend energy most optimally during walking at all walking speeds. During walking, the force exerted by the body on the ground is measured by the ground reaction force (GRF). The GRF value is normalised with the body weight defining a dimensionless parameter . The values are similar for both legs in healthy populations but are di˙erent for the sound and a˙ected leg for amputees. A new hypothesis has been proposed in this thesis that walking is comfortable for an amputee when the di˙erence between values is minimal between the amputee and an equivalent healthy population. The values for healthy adults, as well as amputees, follow a finite number of patterns. The pattern of the values (or the GRF curve) depends on the walking speed of an individual, categorised as slow, fast or free walking. However, it is observed in the literature that free walking speed (FWS) varies over a wide range for healthy individuals (e.g. 1.1 m/s to 1.5 m/s). As a result, it was diÿcult to establish a relationship between walking speed and GRF pattern. A novel parametrised description of GRF curves for a healthy population and amputees is proposed so that a new dimensionless velocity ratio parameter and the corresponding value of the FWS can be predicted by observing the GRF pattern of a healthy adult or an amputee. A new classification approach based on the parametrised description of GRF curves, along with the dimensionless velocity ratio parameter, has been recommended for categorising very slow, slow, free, fast and very fast walking. The GRF result predictions are validated on healthy adults in an experiment conducted in a gait lab. A group of candidates who walk a lot in their daily life were specially selected for this experiment. This classification approach is used to develop a new measure of ULs based on the parametrised GRF description for healthy population and amputees. An innovative computational methodology is proposed to design an optimal sti˙ness response of a prosthetic foot that minimises the occurrence of ULs. This is achieved by transferring the roll-over shape (ROS) information of the prosthetic foot and the corresponding information for a given velocity ratio across a passive walking dynamic (PWD) and a finite element model via a newly defined form of loose coupling. A theoretical case study is presented in which an amputee walks in a gait lab with a representative C-shaped prosthetic foot. The thesis explains how the proposed novel computational methodology is able to redesign the prosthetic foot in a way that is better suited to minimising ULs. The redesign process of the prosthetic foot has led to the development of an innovative 3D printable double keel and double heel design. With the advancement of carbon reinforced polymers and additive manufacturing technology, the sti˙ness customisation methodology proposed in this thesis has the potential to create a new generation of energy-eÿcient prosthetic feet.