Finite element (FE) models of the brain are regularly used to investigate brain injury mechanisms. Validation of these models against cadaver impacts is usually restricted to intracranial pressure data. However, the low shear modulus of neural tissue means that injurious strains result from shear deformations. It is shown that brain injury models that are validated for pressure alone can give a wide range of shear responses to the same impact. Holbourn's arguments on the harmlessness of pressure are extended by introducing separate wave equations for pressure and shear, derived from the Helmholtz vector decomposition. Two idealised models of traumatic brain injury are used to show that there is no one-to-one relation between pressure and shear in head impacts lasting a few milliseconds. The first is an analytical model of wave propagation in the brain under the action of local skull bending. The second is a strain-validated FE representation of the coronal plane of the human head under rotational acceleration. As there is no one-to-one relation between dilatation and distortion in typical head impacts, it is not acceptable to validate FE models for pressure and then use them to predict injury.
Abstract