A methodology for physically-based contact and meniscus properties in rigid-body computational knee modeling

Determining natural inner knee mechanics is a longstanding goal for researchers with applications to prevention and treatment of knee trauma and osteoarthritis. Physical testing has only provided limited information of knee mechanics due to technical challenges and cost. Modeling has been used for decades to obtain some of this otherwise inaccessible information, and recently nite element analysis (FEA) has become a popular means to this end. However, FEA requires time intensive mesh-creation and has large computational requirements. Ideally, model creation should be easy and simulations should be fast to allow for sensitivity analysis. Although allowing easier model creation and o ering over an order of magnitude more computational e ciency than FEA, current rigid body modeling of the knee is limited by imprecise methodologies for de ning material properties. Cartilage and meniscus are particular points of weakness. The following thesis develops an improved methodology for cartilage contact which is user-friendly and allows for precise de nition of contact via user-supplied material properties while accounting for changes in sti ness due to discretization. Additionally, meniscus modeling is improved by developing and implementing equations which directly de ne stress-strain relationships to match values reported in literature or those selected by the user. Results from two implemented knee models are compared to experimental results in literature and sensitivity to material properties and driving kinematics is investigated.