Comparison of lung response and thoracic injury metrics

Abstract

Pulmonary contusion resulting from side impact collisions is a common and significant occupant injury. The goal of this study was to demonstrate the dependence of predicted injury on the loading using a detailed numerical human body model, previously validated for standard impact tests. The development of injury metrics for pulmonary contusion are challenging since it is a physiological response that cannot be evaluated directly using cadaveric tests or numerical models. Two commonly used global criteria for predicting thoracic injury are the Viscous Criterion (VC) and deformation criterion. In contrast, numerical models can be assessed based on the local tissue response, which can be correlated to injury. Two different loading conditions, based on cadaveric tests from the literature, were considered in this study: A free-flight side pendulum impact to the thorax at 6.67m/s, and a full body impact using a NHTSA side-sled configuration at 6.67m/s. For both simulation cases, the predicted VC were 0.99 and 0.98m/s and the chest deformations were 118 and 110mm, respectively for the thoracic region. Injury prediction for the lungs was investigated using four different tissue-level injury metrics. This allowed for an estimate of injury severity for comparison to the VC and the deformation criterion. The four injury metrics predicted a consistent range of contusion lung volume (30 to 37% in the pendulum case versus 53 to 90% in the sled case). The difference in injury suggests that global measures such as VC are not only direction dependent but also loading condition dependent (the area over which the impact occurs). These dependencies are inherent challenges with injury criteria based on large scale measures of response. The application of detailed numerical models provides a consistent basis to predict local organ response and injury when evaluating different loading conditions. © 2010 International Federation for Medical and Biological Engineering.