In an effort to better understand thoracic trauma in frontal impacts, sled tests with cadaveric specimens and test dummies in various restraint environments were conducted.

Analysis of the measured mechanical and physiological responses and anthropometric data of the human subjects suggested that for the same mechanical response (either acceleration, deflection, or a combination of both), belt restraint systems have a higher associated injury rate than air bag restraint systems. The Dicholomous Process was developed to provide better injury evaluation from measured mechanical parameters without prior knowledge of what restraint system was used. Based on multiple chest deflection measurements, this process first determines whether the chest deflection patterns were localized (belt like) or distributed (bag like). It then uses separate injury criteria for each of the two categories.

Analysis of the test data and the first version of the Dichotomous Process has already been presented in the 38th Stapp Car Crash Conference in 1994. This paper presents further validation of the Dichotomous Process and its application to the Hybrid III dummy (with additional chest deflection gages) in vehicle crash tests. Separate injury criteria based on chest acceleration, chest deflection, and age of the cadaver for the belt like and bag like categories are presented. A relationship to scale internal rib deflections measured on the Hybrid III dummy to represent external deflections of a tensed human chest is presented. A method of eliminating the age parameter from the injury function is also demonstrated.