Biomechanical Modeling of Head Impacts: A Critical Analysis of Finite Element Modeling Approaches (Biomechanische modellering van hoofdimpacten: een kritische analyse van eindige elementen modelleringsmethoden)

Over the years, the use of finite element (FE) head models for head impact research increased significantly as computing power became cheaper and more readily available. Besides experiments with animals, cadavers or volunteers, impact simulations with FE head models are used to identify the mechanisms of head injuries and the responses and tolerances of the head to various impact conditions. FE head models are also frequently used for the development and improvement of safety equipment such as airbags and bicycle helmets.Today, a large variety of FE head models exist, each with their own approximate representation of the true physiological and anatomical details of the human head. This approximation includes simplifications in geometry, boundary and loading conditions as well as in constitutive properties and element formulations. Despite the large differences in modeling approaches, these FE head models all aim to model the head and its impact behavior as accurately as possible. The question then arises whether these FE head models are able to predict similar head responses and injuries for a variety of impact conditions?The general research aim of this dissertation is therefore a critical analysis of the modeling approaches used in FE head impact models. To achieve this goal and contribute to the improvement of head injury assessment, four research objectives were formulated which lay the foundation for the four major parts of this dissertation.First, an in-depth comparison was made of the design and the responses and injury predictions obtained from three different FE head models when simulating simple falls. The observed differences in material properties defined for the intracranial tissues as well as the large variation in mesh size and element types give rise to discrepancies in the prediction of head responses and injuries with the three models. Despite the observed discrepancies, no statement was made on the biofidelity of each model s behavior or which model performs best. This is due to the lack of a golden standard to which the FE head models can be compared. However, the results highlight the need for accurate material properties of the intracranial tissues and for further validation of the FE head models to increase their biofidelity.Second, a thorough evaluation of the effect of various FE head model design specifications on the intracranial head response was performed based on a developed simplified FE head model. A number of variations of this simplified model were created using different mesh densities, element formulations and boundary conditions as well as a wide range of material properties for brain and cerebrospinal fluid (CSF). As such, the influence of each model parameter separately on the head model response was investigated. The importance of hourglass treatment as well as the high sensitivity of the brain and CSF response to their material parameters were addressed. Furthermore, a patch test was performed to verify the implementation, accuracy and convergence of hexahedral elements. The test revealed that not all hex-elements available in commercial FE software packages pass this test.Third, a pilot study was performed where two FE head impact reconstruction approaches, i.e. the boundary prescribed motion (BPM) and the initial velocity (IV) approach, were evaluated and compared based on the reconstruction of a real life head impact. For each approach, the internal head response was calculated and the predicted injuries were compared to the actual sustained injuries. It was found that both approaches were able to predict the sustained injuries. Whether one reconstruction approach should be preferred over the other depends on the available information on the impact circumstances and the requested FE results. Where the ability to predict skull fractures is one of the advantages of the IV approach, the main advantage of the BPM approach is the presence of the full body in the head impact simulation which improves the accuracy of the estimated head impact motion.Last, the material characteristics and failure behavior of the bridging vein - superior sagittal sinus (BV-SSS) complex were investigated in order to provide a strain threshold level for ASDH as well as a value for the BV Young s modulus, specified in FE head models. Furthermore, a strain rate dependency as well as the effect of the gender of the specimens on the biomechanical behavior of the BV-SSS complex was investigated. No significant effect of strain rate was found, whereas gender had a significant effect on both the BV-SSS geometry and failure characteristics.In general, it was found that the use of the FE method for the biomechanical modeling of head impacts brings many challenges with it. This dissertation highlighted the necessity to have a good knowledge of the effects of the different FE modeling approaches and of the limitations and pitfalls which can be present when modeling head impacts with an FE head model. This knowledge will allow the user to better interpret the obtained results and make improvements to the models.nrpages: 222status: publishe