DOI: https://doi.org/10.15368/theses.2019.25
Available at: https://digitalcommons.calpoly.edu/theses/2015
Date of Award
6-2019
Degree Name
MS in Mechanical Engineering
Department/Program
Mechanical Engineering
Advisor
Brian Self
Abstract
The study of cycling biomechanics typically requires measurement of pedal loads through force transducer instrumentation. However, analysis of seated cycling often necessitates the additional measurement of the loads exerted on the rider by the seat and handlebars. A stationary exercise bicycle was instrumented with two commercial six-axis force transducers at both the seat and handlebar locations via a custom designed mounting system. The system was tested by applying known forces and moments to the fixtures and proved capable of accurately measuring the loads. Additional data collected from cycling tests were compared to values from the literature to add supporting evidence to the validity of the system.
The instrumented stationary bicycle was further used to study the accuracy of modeling seated cycling in OpenSim. Five participants cycled at a moderate resistance level for three trials. Force and moment data were collected by seat, handlebar, and pedal load cells, while kinematic data were collected by an optical motion capture system. Participant data were analyzed with the OpenSim residual reduction algorithm (RRA) tool excluding seat and handlebar loads. The RRA pelvic residual was then compared to an experimentally determined handlebar and seat equivalent (HBSE) calculated from respective load cell data. Graphical comparisons of the RRA and HBSE results showed strong correlations in Anterior-Posterior (A-P) and Superior-Inferior (S-I) force directions and to a lesser degree, Medial-Lateral (M-L) force and S-I moment directions. M-L and A-P moment plots showed the least correlation between RRA and HBSE. Statistical comparisons showed RRA errors likely within 5.2% body weight (BW) for forces and 2.4% BW*height for moments. Considering the average participant height and weight of 167.7 cm and 63.6 kg, respectively, recommended error ranges for RRA are roughly ±4.0% BW for forces and ±7.2% BW*height for moments. This indicates that the OpenSim RRA tool can be used for cycling analysis.