Blast wave injury prediction models for complex scenarios

Blast waves from explosions can cause lethal injuries to humans. Development of injury criteria has been ongoing for many years, but with the main focus on free field conditions. However, with terrorist actions as a new threat, explosions in urban areas have become of much more interest. Urban areas provide a complex environment for blast wave expansion, thus increasing the difficulty of injury and lethality prediction. TNO and FFI have examined the topic of blast injury in a complex environment to find the most appropriate injury criterion and develop a quick analysis procedure. A review of available models found the Axelsson model to be the most promising. It can predict the injury of the air-filled organs in both complex and free-field blast situations. Unfortunately, it involves a cumbersome procedure, requiring four pressure signals on a so-called Blast Test Device (BTD) as input. However, several single point (SP) methods based on Axelsson to avoid the BTD have successfully been developed. A potential problem with the Axelsson model (both BTD and derived SP models) is that the injury prediction is only calibrated against data from very small charges (maximum 1.36 kg). To further examine the validity of the Axelsson model, results can be compared with predictions by the Bowen/Bass curves, as these formulas are calibrated to a different and much larger data set, both with explosives and in shock tubes. Since the Bowen/Bass models are only applicable for scenarios where the subject is either in an open field or next to a reflecting wall (whereas the Axelsson model is supposed to be valid for any complex scenario), a large range of such scenarios for different charge sizes were constructed and numerically simulated to provide input to the different models. Comparison between the injury predictions generally showed good agreement, except that the Axelsson and Bowen/Bass models diverged considerably for very short but high amplitude blast waves. The reason for this discrepancy was investigated and found to be due to uncertainties in the empirical formulas for blast wave parameters produced by a given explosive charge