Traumatic Brain Injury (TBI) is a significant cause of death and disability in the United States, with approximately 2.5 million emergency department visits, 288,000 hospitalizations, and 56,800 deaths reported to be related to TBI. The majority of TBIs are caused by blunt impacts resulting from falls and motor vehicle crashes. The head injury criterion (HIC) is the only standard used to assess the safety related to vehicles, personal protective gear, and sports equipment. A new brain injury criterion (BrIC) assessing the rotation velocity effect was recently proposed by the US government. Unfortunately, its derivation from animal data scaled to a computer model of the human head raises questions about its reliability. Therefore, there is a need for a more dependable brain injury assessment function that can accurately distinguish between different types and severities of brain injuries.

Blast-induced TBI (bTBI) has emerged as a growing concern, particularly due to the fact that over 80% of mild TBIs experienced by military personnel are a result of primary blast waves, which cause no external damage. However, the biomechanics of bTBI are not well understood. To address this knowledge gap, a validated computer model of the head capable of simulating the blast effect was developed. This model can serve to advance our understanding of bTBI mechanisms and aid in the design of protective equipment to mitigate blast-related injuries. The primary objective of this dissertation research was to create an advanced anisotropic visco-hyperelastic finite element (FE) head model, which can be utilized to establish tissue level injury thresholds capable of predicting different types and severities of traumatic brain injury resulting from blunt impacts and blast loading. This dissertation research work focuses on three specific aims. Firstly, an advanced finite element (FE) human head model was developed, incorporating anisotropic visco-hyperelastic brain properties. This model enables the prediction of injury severity based on impact direction and neuronal architecture. The biomechanical responses of the head at the tissue level were validated against various measurements from Post-Mortem Human Subject tests. Secondly, this validated FE head model was utilized to reconstruct over 160 impact tests encompassing a range of outcomes from noninjury to mild, moderate, and severe brain injuries of various types. Various local and global brain response parameters, such as strain, strain rate, intracranial pressure, and pressure rate, were analyzed, and machine learning techniques were employed to identify the best injury predictors. The resulting injury risk function (IRF) effectively distinguishes between different severities of brain injuries. The predictive capability of the IRF developed in this study was confirmed through simulations of various standard crash tests involving direct head contact with short-duration impacts, as well as airbag deployments with long-duration impacts. Thirdly, open-field blast experiments conducted on PMHS were simulated. The incidental overpressure and intracranial pressure were validated against experimentally measured results to ensure the accuracy of the FE blast models. The analysis indicated that while the IRF is effective in predicting TBI caused by blunt impacts, its applicability for predicting bTBI is limited.