Research presented in this thesis stems from rising concern about blast-induced traumatic brain injury (TBI). It has been hypothesized that brain tissue is damaged by the blast wave generated during an explosion, but the mechanism of tissue injury is unknown. The pressure wave produced by a typical explosion includes a peak pressure, or overpressure, and positive and negative pressure phases;​ each component of this blast wave may make a unique contribution to injury. A simple device called a shock tube is capable of generating the characteristics associated with the blast wave. This thesis presents a computational model of a shock tube being used in our laboratory to investigate the above-mentioned characteristics of the blast wave. The shock tube is approximately 135 cm long and has a 2.54 cm inner diameter.

This research has two primary objectives. The first is to characterize blast wave properties as a function of shock tube independent parameters. In our shock tube, the independent parameters are driver section length and initial pressure. Because the purpose of this research is to study injury due to pressure wave loading alone, the target is placed outside the tube to avoid interaction with venting gases. Quantifying the appropriate region for testing is the second main objective of this research.

Shock behavior within the shock tube was characterized with 1D simulations, while expansion of the wave after it exits the tube was modeled using primarily 2D but also 3D simulations. The numerical code used for this research (called Uintah) was previously developed at the University of Utah. Results show that peak overpressure and positive phase duration increase with driver pressure, and negative phase duration decreases with driver pressure. Response time of the expansion waves is controlled by the driver section length. Expansion waves travel in the reverse direction to the shock wave and reflect back from the shock tube wall. These reflected expansion waves eventually overtake the shock wave, and decrease its peak pressure, increasing the positive phase duration. Results from 2D simulations show that the region lying above 45⁰ angle from the shock tube axis is the most appropriate region for testing of primary blast effects. Preliminary 3D simulations generally agree but suggest that this boundary may be overly conservative so that some of the region below this line is likely also appropriate for testing. As anticipated, the 2D approach has quantitative limitations in modeling 3D behavior. However, comparison with the 3D solution indicates that the 2D approach effectively simulates trends in shock tube behavior. In addition to these findings, an investigation of boundary conditions and potential sources of error in the numerical code are also discussed.