Document Type

Thesis

Date of Degree

Spring 2012

Degree Name

MS (Master of Science)

Degree In

Mechanical Engineering

First Advisor

Jia Lu

Abstract

New methods are employed to develop an anatomically accurate, analysis-ready isogeometric model of skeletal muscles. Current modeling techniques for the analysis of skeletal muscles include the utilization of finite element meshing, which inherently poses a few well-known problems that provide motivation for isogeometric analysis. In addition to those issues, standard FEA meshing cannot preserve smooth geometries, therefore the accuracy of the foregoing model and analysis is reduced. Moreover, there is no easy means to characterize fiber direction in the FEA framework due to discontinuities at element boundaries. Additionally, material property distributions such as the transition of the muscle-tendon complex along the longitudinal axis through FEA are prescribed on an element by element basis, leading to abrupt, unrealistic property changes at element boundaries. The current research builds on the idea of an isogeometric tensor-product rod using harmonic coordinates and NURBS [1]. Through a direct comparison between a meshed, FEA model and the tensor-product rod model, it can be shown that the tensor-product rod model preserves smoothness, enhancing the geometric representation passed through to analysis while reducing the total DOF of the model. Muscle fibers can be easily implemented as parametric lines with muscle-specific orientations along the muscles' longitudinal axis that match distinct fiber orientations existent within common skeletal muscles. This technique not only allows for the representation of perfectly parallel-fibered structures, but also those that do not directly follow the longitudinal axis such as a helical twist. Utilizing this geometric method also provides the framework for implementing material properties using an interpolative-style scheme. Varying properties at specific longitudinal control point cross-sections near muscle termination areas can be designated to more accurately represent the muscle-tendon complex. These new techniques allow for the creation of an analysis-ready, realistic skeletal muscle model of the male human arm. The model contains 28 muscles complete with muscle-specific geometric, fiber, and heterogeneous property characterizations all compiled into a complete "digital muscle library."

Keywords

Fiber, Isogeometric, Muscle, NURBS

Pages

viii, 101 pages

Bibliography

Includes bibliographical references (pages 99-101).

Copyright

Copyright 2012 Matthew John Boss

Share

COinS