Indonesia Conference Directory

Understanding the physical mechanisms of bone fracture represents a major challenge in biomechanics, since it allows the enhancement of injury criteria commonly used by Euro NCAP benchmarks for the safety of passenger cars or pedestrians. It can also deliver a follow up of athlete-s safety during their trainers avoiding risk zones of injury especially in contact sports. This knowledge is essentially based on the use of the numerical models, whose prediction is assessed through the development of high resolution medical imaging and simulation softwares. Among these models, the anthropometric test device (ATD) commonly used for crash-test or other more detailed local models simulating the interaction between bone tissue and clinical equipment such as prostheses. Their advantage lies mainly in the diversity of configurations and loading conditions and hence the optimization of time and the total benchmark cost. However, one can observe that the material constitutive laws used are often derived from the experimental characterizations carried out at the macroscopic scale ignoring the bone microarchitecture. A micromechanical based approach revealed to be more suitable where the robustness of computation and accuracy of results are of interest. The present investigation is devoted to the theoretical formulation and validation of an ductile damage model applied to the human humerus bone in the thermodynamics framework. The approach consists in formulating the macroscopic material tangent operator by considering the linear local behavior of each phase. Due to the matrix-inclusion morphology of the bone microstructure, a Mori–Tanaka scheme was considered at the localization stage. In order to consider the strain rate effects on the humerus behavior, the standard model of Johnson-Cook was adopted as a preliminary trial. The obtained micromechanical model was implemented using a User Material subroutine (UMAT) within the explicit dynamic code LS-DYNA. The validity of the resulting finite element model was validated by comparing numerical predictions with experimental measurements at different length-scales. The outcome of the proposed ductile damage model appears to correctly predict the general trends observed experimentally through the good estimation of the ultimate impact load that a human humerus may encounter at fracture. The fracture patterns predicted by the proposed micromechanical damage model are consistent with the physical humerus rupture even if this model is limited only to the fracture initiation. Further improvements will be performed to the present model to take into account the marrow effects and fracture patterns.