The changing of the ankle stiffness is based on a system of a cam transmission and a leaf spring whose support can be altered [10]. In this current design, the adjustment mechanism is based on a lead screw which is used to move a support slider for the leaf spring. The adjustment is made possible by a radial knob at the back of the foot, which when rotated, will result in the translational motion of the support slider.

Figure 1 shows a schematic of the adjustment mechanism. In this figure, the screw is constrained at both ends in such a way that when rotated, a translational movement of the support attached to it is achieved. The spring is deflected by force Pemanating from movement of the cam transmission that forms the ankle joint. The movement of the support slider causes a change in the stiffness of the spring. This change in the stiffness of the spring will result in a corresponding change in ankle stiffness as described by Shepherd and Rouse [10]. The distance x, moved by the slider support upon rotating the screw through an angle θ, is given by

x = l θ / 2 π (1)

where l is the lead. The mechanism is designed to have the maximum travel distance of the support slider as 80 mm. The number of rotations of the screw for

this travel was arbitrarily taken as 10, and therefore the lead of the screw was determined as 8 mm. Thus, for a complete revolution of the screw, the slider moves through a distance of 8 mm, as described by the relation in Equation (1).

The forefoot section is incorporated in the design of the foot in order to mimic the functionality of the toe section in the natural foot. In normal gait, the foot rocker occurs when the heel is raised from the ground, while the toes remain until the end of the late stance where toe-off occurs [11]. Late stance begins from the moment the heel leaves the ground until the toes also clear the ground. In order to achieve the functionality of the toe section, the design is based on a cantilever attached at the front end of the prosthesis. As such, the forefoot section is modeled as a cantilever beam that deforms (deflection of the beam) and stores energy in the form of strain energy, which is released at toe-off, in order to propel the leg forward.

Figure 2 shows an illustration of the designed prosthesis as it progresses from midstance to toe-off. As the heel begins to rise, the foot rocker occurs, where the foot rolls over the metatarsophalangeal joint. Since the forefoot is a compliant fixture, it will deform as the heel rise continues, thus storing energy in the forefoot. This energy will be released at toe-off, towards the end of late stance, where the foot completely leaves the ground and begins the swing phase of the gait cycle. The forefoot is to be made from Nylon material, as this material exhibits a high strength to weight ratio, as well as high strain energy density [12]. The material is readily available and easy to machine. Energy stored in the forefoot in the form of strain energy, is given by,

U = F 2 l 3 / 6 E I (2)

where F is the ground reaction force at the terminal stance, l is the length of the forefoot beam, E is the modulus of the materialused and I is the moment of inertia.

The full structure of the ankle-foot prosthesis is made up of the forefoot section, the adjustment mechanism and a cam and leaf spring transmission for the ankle

joint. The prosthesis has one rotational degree of freedom about the ankle joint, while the forefoot is compliant. The dimensions of the proposed prosthesis are such that they closely resemble the dimensions of a natural ankle-foot complex of a 1.73 m tall human subject [13]. The dimensions of the natural limb and those of the designed prosthesis are presented in Table 1.

The solid model of the designed prosthesis is shown in Figure 3. Most of the parts of the mechanical structure are made of Aluminium 7075, with the exception of the forefoot, which is made of Nylon for greater flexibility and energy storage.

Structural analysis was conducted in order to assess the strength of the structure and to ascertain that the structure is strong enough to withstand the loads to which the prosthesis will be exposed to in use. The analysis was conducted on critical parts of the final model of the prosthesis.

The maximum force, which is achieved at heel strike is used as this is the maximum force that the foot will be exposed to during use. This heel strike impact force used is equivalent to 110% of the body weight as per literature [14] [15] [16]. The impact force, F_max, was therefore obtained using the conversion,

F max = 1.1   mg (3)

where m is the body weight in kg, and g is acceleration due to gravity, taken as 10 m/s². For this study, a body weight of 75 kg is taken as the design criterion, and as such the value of the vertical Ground Reaction Force (GRF) that the prosthesis would be exposed to on impact with the ground is 825N. This is the force used in the analysis. Von Mises yield stress criterion was used to analyze the design and the results obtained are presented in Figures 4-7.

Figure 4 shows the stresses of the ankle component based on Von Mises yield strength criterion, as well as the resulting displacement during static loading. The maximum stresses within the part are 4.69 MN/m², with the yield strength of the material used being 95 MN/m², thus giving a safety factor of 20. The maximum deformation of the part is seen to be 1.593 µm, which for practical purposes, is negligible.

For the foot base, as shown in Figure 5, the maximum stress was found to be 0.79 MN/m², resulting in a safety factor of 119. There is no significant deformation within the part, as the maximum deformation is 16.66 µm.

In the results of the analysis of the forefoot section shown in Figure 6, the maximum stresses are 88.7 MN/m², as compared to the yield strength of 103.6 MN/m² of the material used, resulting in a safety factor of 1.2. The large deformation observed is accepted as the forefoot is designed such that it emulates the flexibility of the toe section in a natural foot. The energy stored during this deformation of the section, as discussed in earlier sections, will then be released in order to assist in propelling the leg forward.

Another critical part on which analysis was carried out is the ankle bracket. The force used in analysis was half of the GRFs (412.5N) as there are two brackets that share the total load of 825N. The results are shown in Figure 7 above. The maximum stresses of 16.58 MN/m² within the bracket are significantly lower than the yield strength of the material, which is 95 MN/m², resulting in a safety factor of 5.7. This shows that the brackets are able to withstand the loads that they are exposed to during use. The maximum deformation is 4.617 µm, which is an insignificant.

The designed prosthesis was also evaluated for other properties, using the work by Zaier and Al-Yahmedi [13] as a standard for the natural human foot properties. The designed prosthesis closely resembles the dimensions of the natural ankle-foot complex. This is desirable as the prosthesis would mirror the missing limb. Another property which was evaluated is the weight of the prosthesis. The achieved weight, with all appropriate materials applied in the 3D model, was found to be just above 1.1 kg. For the natural foot, the weight is taken to be 1.37% of the body weight [13], and thus applying the 75 kg as before, the comparable weight of an intact ankle-foot complex would be about 1 kg. Thus, the prosthesis also closely mimics the weight of the natural limb.