Is there a minimum complexity required for the biomechanical modelling of running

Mathematical models have the potential to provide insight into human running. Existing models can be categorised as either simple or complex, and there appears to be a lack of natural progression in model development. By sequentially adding complexity, there is the potential to determine how different mechanical components contribute to the biomechanics of running. In this study, a series of four models, of increasing complexity were developed in OpenSim: a simple spring-mass model, a two-segment model with a torsional spring at the knee and two three-segment models, one with a sprung knee and ankle and another with a sprung knee and actuated ankle. For each model, a forward simulation was developed and model predictions compared with experimental data from 10 forefoot runners. The results showed the spring-mass model overestimated the vertical displacement of the centre of mass (percentage difference: 43.6(22.4)-67.7(21.7)%) and underestimated the vertical ground reaction force (percentage difference: 13.7(8.9)-34.4(10.9)%) compared to the experimental data. Adding a spring at the knee increased the match with the vertical centre of mass displacement (percentage difference: 4.4(25.2)-18.4(40.2)%), however, geometry restrictions meant it was only possible to model approximately 60% of stance. The passive three-segment model showed a good match with centre of mass movements across most of stance (percentage difference in the vertical centre of mass displacement: 4.3(24.5)-21.3(19.2)%), however, actuation at the ankle was required to obtain a closer match with experimental kinetics and joint trajectories (e.g. vertical ground reaction force RMSD decreased by approximately 0.4BW). This is the first study to investigate models of increasing complexity of distance running. The results show that agreement between experimental data and model simulations improves as complexity increases and this provides useful insight into the mechanics of human running.