Experimental evaluation of underride analysis techniques and empirical validation of a new analytical technique

Abstract

Accident reconstructionists are often faced with damage patterns and locations on vehicles that are not well defined by available barrier impact data. One such example is a frontal underride collision. Underride impacts occur when there is a height mismatch between the primary structural components of the impacting vehicles, and the vehicle with the lower height is forced beneath the structure of the other vehicle. The lack of structural engagement typically allows for significantly different damage patterns due to the inherently lower stiffness of the underriding vehicle's contacting surfaces coupled with complex interactions between varying surfaces. In this study, a series of two-vehicle impact tests between a small pickup (bullet vehicle) and a large dump truck (target vehicle) were performed and studied. These tests involved a severe underride configuration in which the dump truck bed's vertical alignment was above the base of the windshield of the pickup. Coupled with these impacting surfaces was a single vertical support, a remnant of a commonly referred to ICC (Interstate Commerce Commission) bumper, which caused a narrow object-type impact, but did not extend down to the pickup's bumper. Multiple prior authors’ analytical and empirical relationships to predict impact speed based on crush damage were evaluated using the results of these tests as well as other published underride tests. No single model was sufficient at predicting the mixed mode of impact present in these impact scenarios. However, a system of equations was developed to predict the impact parameters utilizing a combination of previously reported methods and a new empirical relationship presented in this study. This new method shows high correlation and supports the authors’ hypothesis that separate crush models can be applied to multiple discrete areas of a vehicle and then combined to form a more complete predictive systematic model.