Finite Element Modeling of Skull Fractures : Material model improvements of the skull bone in theKTH FE head model

Abstract: The main aim of this project was to develop a model for predicting skull fractures of a 50th percentile male, using a finite element head model developed at the Neuronics department of KTH, Royal Instituteof Technology. The skull bone is modeled as a three layered bone, where the outer and inner tables are modeled as shell elements, while the diploë is modeled by two layers of solid elements. The material model of the tables was changed from the material model MAT_PLASTIC_KINEMATIC to a material model including a damage parameter to soften the damaged material and to enable ploting of the damage of the skull bone. Due to the coarse mesh of the FE head model the material model was not allowed to include any erosion, deleting element as they reach their ultimate strain. With these requests, two materials from the LS-DYNA material library seemed appropriate: material 81,MAT_PLASTICITY_WITH_DAMAGE and material 105, MAT_DAMAGE_2. To evaluate these materials and adjust the input parameters a dog bone FE model was developed and tension tests were simulated with this model, equivalent to tension tests performed on equally shaped skull bone specimens. The material simulating a behavior most similar to the behavior from the tension tests turned out to be material 81. This material model was then implemented in the full FE head model for further input parameter adjustment and validation. Four different cadaver experiments were simulated, with different impacting objects: sphere, box, cylinder and flat cylinder surface, and impacted areas of the head: vertex, temporo-parietal and frontal. The forces obtained in the simulations were compared to the forces of the cadaver experiments. The fracture prediction was based on the damage parameter, which could be plotted to visualize the areas where the ultimate strain was exceeded and thereby the area most likely to be fractured. This parameter was then compared to the documented fractures from the cadaver experiments. The result showed that using material 81 with the input parameter EPPFR=0.05 gave the overall most accurate forces and fracture predictions. The breaking stress, σB, did not affect the fractures significantly but a reduced σB resulted in reduction of the peak forces. The thickness of the diploë did not have any significant impact on the fracture occurrence, but a thinner diploë had a reducing impact on the peak forces as well.