A simple computational method to estimate stance velocity in running.

Abstract

Running dynamical analyses typically approximate a runner's stance velocity as the average stride cycle velocity (the average running speed). That approximation necessarily overestimates stance velocity and biases subsequent results. Stance velocity is crucial in kinetic spring-mass analyses of running, where approximation of a runner's impact angle and calculation of leg stiffness require that input. Here, a new method is presented to approximate a runner's stance velocity via measurement of contact and flight times with the runner's average speed, leg length or height, and mass. This method accurately estimated the stance velocity of simulated spring-mass systems across typical running speeds of 3.5-5.5 m s-1 (r>0.99) and more accurately estimated the impact angle and leg stiffness. The method also accurately estimated the peak horizontal ground reaction force across speeds (r=0.82), but the bias magnitude increased with speed. Robustness of the new method was further tested for runners at 2.5, 3.5 and 4.5 m s-1, and the new method provided steeper impact angles than those from traditional estimates and correspondingly higher leg stiffnesses, analogous to the observations in the simulation models. Horizontal ground reaction force estimates were weakly correlated in braking and propulsion. They were improved by a corrective algorithm, but the intra- and inter-individual variation persisted. The directionality and magnitude of angle and stiffness estimates in the human runners were similar to simulations, suggesting the new method more accurately modeled runners' spring-mass characteristics by using an accurate approximation of stance velocity. The new method can improve traditional kinetic analyses of running where stance velocity recordings are not captured with kinematic recordings and extend opportunities for accurate field-based analyses with limited measurement sources.