This item is available under a Creative Commons License for non-commercial use only
The increasing number of cars and collisions on the roads of Europe ensure that the need for accurate speed estimation is an ever-more pressing requirement. This requirement has provided the motivation for an investigation into the accuracy and reliability of existing crash models An in-depth literature review was undertaken to create a greater understanding of the structural collapse behaviour of vehicles during full frontal collisions. It was decided to concentrate the work on frontal collisions as this type of accident makes up the majority of road accidents. This review underlined some of the phenomena that need to be considered when assessing a crashed vehicle. To create a foundation for understanding and criticising the existing models, derivations of the fundamental theories related to crash analysis was considered. This “first principle” approach ensured that the models were verified against well established momentum and impact mathematics. Physical frontal impact testing was carried out on six vehicles with closing speeds ranging from 36 kph to 84 kph. The results obtained from these tests were amplified by test data from the National Highway and Traffic Safety Association (USA). All these tests were used to determine the accuracy of the six most common models over this speed range. It was found that each model had differing accuracy over certain speed ranges. This was particularly true of the linear-based models which showed increasing inaccuracies at higher speeds (>90 kph). It was also noted that there was a lack of high speed collision data available and unsurprisingly the models based on empirical regressions became inaccurate at these higher speeds. Another observed discrepancy was the omission of the peak force data at the point when the engine breaks away from its mountings. Because of this, any speed estimation from a collision with a crush level coinciding with the front of the engine block should be regarded with a degree of scepticism. This research suggested inaccuracies of up to +/- 20 kph in closing speed prediction as a result of this omission.
Rice, R. (2004). Crash analysis: an investigation of current speed estimation techniques. Masters dissertation. Technological University Dublin. doi:10.21427/D7X02B