Prediction of rib cage fracture in computational modeling: effect of rib cortical thickness distribution and intercostal muscles mechanical properties

A finite element model of the 50 percentile male was developed based on geometrical surface generated from medical images. First, an interactive multi-block meshing approach was used to generate high quality quadrilateral and hexahedral meshes of the thorax anatomical structures. Second, a methodology based on the mesh blocks was developed to assign cortical thickness data taken from a micro-CT study to each of the nodes in the cortical shell elements of the ribs along the longitudinal direction and around the cross-sectional perimeter. The whole thorax model (rib cage, internal organs, muscles, skin) was exercised under a wide range of loadings that include inertial and non inertial loadings (blunt impacts, and table top). Although the response of the thorax model was reasonable compared to the experimental results at a “global” level—such as under hub or belt loading onto the entire body—, it has not been evaluated at a local level -such as the strain distribution in the rib cage. The structural response of the rib cage was therefore investigated to evaluate the effect of cortical thickness distribution and intercostal muscle mechanical properties on the thorax mechanical response. The need for node dependent cortical thickness to predict force and deflection at the time of fracture was demonstrated at the rib level by simulating anteroposterior dynamic bending of individual ribs. As for the intercostal muscles, there is no experimental data available to aid with the definition of their mechanical properties. Therefore the impacts to the lateral thorax recently performed by CEESAR for the THOMO project were used to carry out a sensitivity analysis to assess the effect of the cortical thickness distribution and intercostal muscles material properties on rib fracture prediction. The FE model of the thorax was run for three cortical thickness distributions (one distribution with thicknesses defined for each node, and two distributions of uniform thickness values) and three values for the intercostal muscles’ Young’s modulus. The variation of the strain field was compared for the various combinations of parameters and loading conditions to assess how the fracture prediction was altered. In particular, the rib strain profiles measured in the experiments as well as the locations of the rib fractures were compared to the FE results. This study represents a major effort in the development and validation of the thorax finite element model for the Global Human Body Modeling Consortium, and provides insight for the development of anatomically detailed computational models for injury prediction.