Investigating the hydrodynamic performance of two prototypes of prosthetic legs for swimming using computational fluid dynamics

Swimming is one of the physical activities mainly practiced in the world, with recent improvements to make it easily accessible to people with physical disabilities. However, proper devices for swimmers with limb amputation/deficiency have not been developed yet, mostly due to the ban on their use during competitions. The latter is often remedied using commercial paddles during training, which might limit the swimmer’s performance and cause some injuries due to the unbalanced muscles effort. In last years, a marked interest has emerged on the application of computational fluid dynamics (CFD) to study the flow patterns developing around the swimmer’s body or extremities, with the final aim to provide a numerical estimation of drag and/or lift force acting on the athlete and opposing his/her movements in water. Hence, allowing a better understanding of flow behaviour around swimmers and supporting their performance improving. In this scenario, here we used CFD to analyse, for the first time, the hydrodynamic performance of two customized prototypes of prosthetic legs for swimming. In details, the two prototypes were designed starting from the socket, the custom-made interface between the residual limb and the prosthesis, and modifying the pylon morphology to minimize the drag force during swimming. More in detail, the pylon geometry of model 1 was concave and non-linear, while that of model 2 was smooth and linear. One extremity of the pylon was attached to the socket, and one other to a commercially available prosthesis foot. A rectangular box full of water was reconstructed around each model to simulate the underwater condition. Models were then meshed and the standard k-ω turbulence model was used for drag force estimation. Steady-state numerical simulations of the flow around each prototype of lower limb prosthesis were performed for five different inclinations (i.e., -20°, -10°, 0°, 10°, and 20°) with respect to main flow direction and for three different swimming velocities (i.e., interregional 1.4 m/s; national 2.2 m/s; international 3.1 m/s). Additionally, unsteady-state numerical simulations were performed for the models aligned with the flow by simulating two swimming styles: front crawl and backstroke. The hydrodynamic performance of each model was quantitatively investigated in terms of total drag force and drag coefficient. It clearly emerged that: (1) model 1 promoted the development of two slow asymmetrical vortices at the ankle region, while only one slow vortex resulted for model 2 at the same region; (2) the lowest values of drag force resulted for both models when aligned as the main flow direction; (3) a prosthesis inclination of -20° with respect to the main flow direction is associated with the minimum drag coefficient; (4) as expected, the drag force increases as the swimming level progresses; (5) no relevant difference emerged in the hydrodynamic performance of the two prototypes under investigation (up to 3% of difference in drag coefficient). In conclusion, result from this study suggest that the differences in the pylon geometry of the two models do not produce significant difference in their hydrodynamic performance. Additionally, independently of the swimming level, the highest performances are guaranteed by the alignment of the prosthesis with the main flow direction.