Evaluation of the approximations involved in analyzing high rate shear experiments of brain tissue using finite element analysis

Abstract

The results of brain tissue finite element (FE) models under high rate shear deformation are affected by several factors. This thesis evaluated the effects of hourglass control, Poisson's ratio and element type in such simulations. Moreover, a comparison of FE and analytical models were performed related to boundary conditions. The simulations and optimizations were executed in ANSYS, LS-DYNA and LS-OPT. A Rivlin hyperelastic material model with linear viscoelasticity was used to describe the mechanical response of brain tissue. Examples of inverse FE material characterization of representative brain shear experiments at strain rates of 800, 500, 120 and 90 S-1 were studied and the results were validated by the ability to predict wave traveling times and deformed configurations. The difference between experimental and idealized shear strain increased with aspect ratio. One-point-integrated brick element combined with stiffness hourglass control gave the best result. A smaller Poisson's ratio that is still physically meaningful, e.g. 0.495, is preferable.