Computer simulation of the sprint start

The aim of this project was to investigate the mechanics of the sprint start through the use of computer simulation. Experimental data was collected on one male athlete in accordance with a procedure agreed by Loughborough University Ethical Advisory Committee. The data provided subject specific data for the creation of a four and fourteen segment, angle and torque driven models of the sprint start. The models simulated the start from the moment of onset of force production until takeoff from the starting block. The four segment model comprised a head and trunk, thigh, shank and foot whilst the fourteen segment model also included a lower spine and pelvis, upper arms, forearms and hands, as well as the other leg including two segment feet. Subject specific torque data was combined with EMG data to provide input to the torque models Results from the four segment angle driven model demonstrated that the participant will benefit from using smaller joint angles than usual in the set position as this resulted in increased velocity on takeoff with minimal increase in movement time. The model also showed large joint torques during such starts and so suggested that this is likely to limit start performance. The four segment torque driven model also revealed that optimal joint angles exist for the hip and knee but such a result was not clear for the ankle. For this model the optimum angle at the hip was 73 (the smallest tested) and 108 at the knee which was the athlete’s usual angle. Increasing the athlete’s strength parameters resulted in a small increase in horizontal velocity on takeoff for some simulations and all simulations had enhanced acceleration. Increasing initial muscle activations didn’t increase horizontal takeoff velocity but did also increase horizontal acceleration. The fourteen segment angle driven model was used to optimise spring parameters for input into a torque driven model. The fourteen segment torque driven model simulated movements and forces realistically but an adequate match was not found to the sprint start performance of the participant due to long simulation times and lack of computing power.