A National Automotive Sampling System–Crashworthiness Data System (NASS–CDS) based Similarity Scoring Methodology (SSM) is presented for the quantitative comparison of real-world crashes to crash tests. Using NASS–CDS 2000–2008, five categorical and five continuous crash, vehicle and occupant parameters were utilised for frontal and side impacts. Mahalanobis metric results revealed that 1% of frontal and 23% of side NASS–CDS cases scored received similarity scores of <0.14 demonstrating greater similarity to standard crash tests. These included 20,334 frontal and 442,511 side impact case occupants (weighted). On average, the best scores occurred for NASS–CDS cases compared to FMVSS 214 side crash tests. The majority of real-world crashes had lower delta-Vs and maximum crush than associated crash tests. The results will aid in the study of occupant safety and vehicle crashworthiness by helping researchers identify population groups to study real-world injuries versus the injury risk predicted by anthropomorphic test devices.
Similarity scoring methodology for comparing real-world cases to crash test standards
International Journal of Crashworthiness ; 19 , 1 ; 57-70
2014
14 Seiten, 7 Bilder, 5 Tabellen, 36 Quellen
Aufsatz (Zeitschrift)
Englisch
Similarity scoring methodology for comparing real-world cases to crash test standards
Taylor & Francis Verlag | 2014
|SAE Technical Papers | 2024
|