The Quest for Dynamic Consistency - A Comparison of OpenSim Tools for Residual Reduction in Simulations of Human Running

The use of synchronous kinematic and kinetic data in simulations of human running will typically lead to dynamic inconsistencies (i.e. residual forces and moments) being present. Minimising the residual forces and moments in such simulations is important to ensure plausible model outputs (e.g. joint moments, muscle forces) are obtained. A variety of approaches suitable for residual reduction are available in OpenSim, however a detailed comparison of these is yet to be conducted. This study compared a variety of OpenSim tools applicable for residual reduction in simulations of human running. A series of approaches (i.e. singular and iterative Residual Reduction Algorithm, MocoTrack, AddBiomechanics) designed to reduce residual forces and moments were examined using an existing dataset of 10 male participants running on a treadmill at 5.0 m·s (n = 3 gait cycles per participant). The computational time, resultant residual forces and moments, and output joint kinematics and kinetics from each approach were compared. A computational cost to residual reduction trade-off was identified, where lower residual forces and moments were achieved using approaches that required longer computational times. All of the tested approaches regularly reduced residual forces below recommended thresholds, however only the MocoTrack approach could consistently achieve acceptable levels for residual moments. The AddBiomechanics and MocoTrack approaches produced variable lower and upper body kinematics, respectively, versus the remaining approaches; with minimal other qualitative differences were identified between joint kinematics from each approach. Joint kinetics were qualitatively similar between approaches, however MocoTrack generated much noisier joint kinetic signals. The MocoTrack approach was the most consistent and best performing approach for reducing residuals to near-zero levels, at the cost of longer computational times and potentially noisier joint kinetic signals. This study provides OpenSim users with evidence to inform decision-making at the residual reduction step of their modelling and simulation workflow when analysing human running.