Calibrating multibody ulno-humeral joint cartilage using a validated finite element model

Correct prediction of contact characteristic in multibody models is a challenging issue, which requires both employing an appropriate contact model and choosing its corresponding contact parameters. This study employed two subject specific multibody models of the ulno-humeral joint where the humerus cartilage was discretized and interacted with the ulna cartilage through a deformable contact. Parameters for the deformable contact were optimized by matching the maximum predicted contact pressure in the multibody models with contact pressures predicted in identical finite element models under axial 110 N compressive load. The performance of the optimized contact parameters and employed approach was evaluated under two other loading conditions of 80 and 140 N. Moreover, the effect of discretized element size on the contact prediction was also addressed. After optimization of the contact parameters, the difference between the multibody and FE models in terms of peak and average contact pressure, and contact area was significantly reduced (more than a factor of 10). Smaller discretization size of the humerus cartilage in the multibody models resulted in better predictions especially on highly curved regions. However, the optimized parameters of our study were different between the two elbows. This indicates that optimized contact parameters can be different not only among different joints, but also among different specimens and optimization conditions. Therefore, in order to improve contact predictions, separate contact optimization should be performed for each subject.