Development and demonstration of a tendon adaptation model to aid in the evaluation of repetitive jobs
The aim of this dissertation was to develop, demonstrate and evaluate a tendon adaptation model (TAM), which predicts the relative tendon damage incurred by wrist muscles during repetitive wrist motion. This research was divided into three major goals.
The first goal was to adapt the muscle model of Loren and Lieber (1995) with an optimization routine in order to predict muscle forces during dynamic motions. The muscle model was combined with the cumulative strain model of Goldstein (1981) and the tendon healing and damage models of Wren, Beaupre and Carter (1998) and Liao and Belkoff (1993) respectively. TAM predicts the amount of tendon adaptation given job/worker specific force, repetition, duration, and anthropometry and wrist posture. Individual components and the entire model were subjected to benchmark tasks to demonstrate how TAM produces muscle forces, tendon strains and damages for a given task. Results showed that the TAM reproduced results found in animal studies.
The second goal was to demonstrate the model's capabilities. The first of three experiments was designed to collect maximal contractions at the wrist to determine the maximal strength (PST) of five wrist muscles. The second experiment was designed to evaluate the muscle model under dynamic situations. TAM predicted muscle stimulation patterns were compared to those recorded from subject electromyograms. Predicted stimulation patterns synchronized well with electromyograms but did not match as well in amplitude. The third experiment was designed to collect necessary data over the course of one hour while subjects performed a repetitive wrist flexion task to determine tendon strain parameters K1, K2 and α. The results suggest that subjects change their kinematic and kinetic profiles over the course of an hour to minimize the effects of fatigue and damage.
The third goal was to perform a sensitivity analysis on the model by varying the estimated parameters and subject specific measurements. The muscle model was found to be sensitive to the parameter PST. The cumulative strain model was hardly sensitive to the parameter K1 while changes in the parameter α were found to have little effect on ending damage levels for simulations lasting approximately 1 hour.
0575: Sports medicine
0354: Occupational safety