This thesis presents experimental results from tests performed to measure the response of a human knee joint in dynamic lateral loading. A critique of previous knee joint studies was done and used as a motivation for developing a new test methodology. In the present study, isolated knee joint specimens from Post Mortem Human Subjects (PMHS) were tested in 4-point medial-lateral bending (15 tests) and 3-point medial-lateral bending with varying proportions of shear force (8 tests). The loading rate replicated the pedestrian impact at 40km/hr which corresponds to knee bend angulation rate of 1°/ms.
The bending moments and shear forces measured at the load cell were inertially compensated to estimate the moment and shear response at the knee joint center. The mean bending moment, scaled to 50 th percentile adult male, for the first injury in bending tests was 122 ± 15 Nm (95% confidence interval) which is much less than previously reported in the literature. It is shown that procedures like geometric scaling alone do not truly account for subject variability and have little effect on reducing the response variance. The sample size of 15 chosen in the study for bending tests is justified by comparing its effect on the failure moment confidence interval. Post test necropsy showed repeated incidence of MCL injury in all specimens tested which is consistent with observations from pedestrian traffic injuries.
In the presence of small proportions of shear, no appreciable change in bending moment response was observed. However for larger proportions of shear, inertial effects dominate and rigid body assumptions are not sufficient to compensate for knee joint response. Similar to the bending tests, post-test necropsy showed that MCL avulsion is the only injury in most combined loading tests as well
Thus it was concluded that MCL elongation upon tension in valgus bending primarily dictates the knee bending response for laterally struck pedestrians. The bending response curves and first failure moments as reported in this study can be used for estimating injury thresholds and validating computational and mechanical lower limb impactor models.