Properly functioning ligaments are an essential characteristic of healthy joints. Despite numerous investigations, questions remain regarding the exact mechanical role of ligaments as well as the cause and effect of specific injuries and surgical procedures. This is in part due to inherent limitations of experimental and clinical studies. Computational modeling techniques such as the finite element method have tremendous potential for the study of ligaments, and the use of such models reduces some limitations of experiments. Unfortunately, most finite element models to date have either used a simple one-dimensional ligament representation or a three-dimensional representation that uses a simplified application of ligament initial strain in combination with average material properties and geometry. These simplifications limit the ability of previous models to make accurate predictions of complex three-dimensional ligament mechanics.

This dissertation investigated the mechanics of the human medial collateral ligament (MCL) and developed a number of tools that will facilitate future study of other ligaments and biological tissues. A detailed experimental and computational study quantified the strain distribution in the human MCL during passive flexion as well as with subsequent valgus loading. Subject-specific finite element models of the femur-MCL-tibia complex were developed for a series of knees based on experimentally measured geometry, material properties, and boundary conditions. The models featured a transversely isotropic, hyperelastic representation of the MCL and were validated usingexperimental measurements of MCL strain. In addition, a new methodology was developed to characterize the response of MCL samples to large deformation shear loading. Simple shear tests were performed along the fiber direction, allowing the normally dominant tensile resistance of collagen to be eliminated from the tissue response. The inhomogeneous deformation during finite shear loading prompted development of a nonlinear optimization procedure that used specimen-specific finite element models to determine material coefficients for a transversely isotropic constitutive model. This framework for parameter estimation will be used to develop new ligament constitutive models that will improve future finite element models. The currently developed modeling methodologies can ultimately assist with patient-specific surgical planning and education and to study the effects of tissue level deformations on local cellular mechanotransduction.