External mechanical work of locomotion from inverse dynamics
insight from different body plans
The external mechanical work (WEXT) estimation primarily depends on the trajectory of the body centre of mass (COMb). Inverse dynamics (ID) provides reliable tools to reconstruct the COMb position from different kinematic models. We measured and compared the WEXT from a full body model and a simplified one, in human locomotion and in octopedal locomotion of terrestrial spiders, in order to quantify the difference and evaluate the reliability of the latter model.
Analyzing the COMb displacements by means of one or two landmarks fixed to the main body segment can be a simple approximation, useful for different purposes. Conversely, the full models take into account the movement of all the body segments, with different complexity levels. In our protocols, the simplified model was a subset of the full model. Therefore we could collect both in the very same trial, using a motion capture system. Spider kinematic data were collected during free displacements in a calibrated space. Humans performed walking, running and skipping at different controlled speeds on a treadmill.
The simplified model always resulted in a variable, speed dependent, overestimation of the WEXT. 3D kinetic energy of the COMb was affected more than the potential energy. Therefore in bouncing gaits like skipping and, on minor extent, running, the differences were proportionally smaller than in walking. In skipping the error was almost constant (30%) throughout the speed range. The error was also affected by the relative weight of the body segments. For the purpose of estimating the mechanical energy of the COMb, a full body model is highly recommended, at least in vertebrates.