Meeting Abstract
Legs are traditionally considered to be compliant during running and rigid during walking. This derives from the paradigm of a compliant spring-loaded inverted pendulum (SLIP) mechanism during ‘bouncing’ steps of running and a rigid inverted pendulum mechanism during ‘vaulting’ steps of walking. We examined this dichotomy by combining experimental step length manipulations in walking humans with a serial actuator-spring model of measured leg dynamics to determine changes in the modeled radial-leg spring constant with step length. Our model finds the radial-leg spring constant that minimizes the total actuator work. Across walking step length conditions from shortest to longest, the radial-leg spring model shows that stiffness decreases as a reciprocal function of leg excursion angle. This curve fit of stiffness versus excursion angle reveals a horizontal asymptote at 19.5 kN/m, which matches the radial-leg spring constant for running, indicating a confluence with the compliance of running legs at unnaturally long walking step lengths. At the other extreme, as step lengths become unnaturally short the vertical asymptote of this fitted curve is 38 degrees – indicating that human legs become arbitrarily stiff as they approach the excursion angles used by rigid-legged passive dynamic walkers at the upper limit of their walking speed. However, at natural step lengths, humans achieve compliant walking with a relatively modest 25-50 percent increase in radial spring constant, which does not support a qualitative change from compliant-legged running to rigid-legged walking at the step lengths and speeds typically used by human walkers.
