As front crashworthiness has improved in relatively low-exposure, high-severity
configurations, performance in high-exposure lower severity crashes has become
an area of increasing interest. Identifying representative test conditions
requires accounting for known inaccuracies in crush-based estimates of real
world crash severity, and for the possibility that whole-body injury exposure
distribution masks body region differences. This study analyzes front impacts
from the Crash Investigation Sampling System (CISS) with event data recorder
(EDR) information to characterize crashes based on longitudinal ΔV, pulse
duration, pulse deceleration, PDOF, and number of events. These factors also
were used to model injury risk along with driver sex, age, and BMI. Results
demonstrated EDR-based ΔV estimates were systematically higher than crush-based
estimates and the difference increased with injury severity. EDRs also reported
higher ΔV for single-event crashes producing injury than multiple-event crashes.
These factors can produce large differences in how crashes are characterized:
while median WinSMASH ΔV for all crashes producing MAIS3+ injury was around 30
km/h, median EDR ΔV for single-event MAIS3+ injury crashes was 60 km/h. Body
region differences also were considerable; MAIS2+ upper extremity or
head/face/neck injury occurred in crashes with median ΔV less than 30 km/h,
while thoracic or lower extremity injury crash medians were around 50 km/h.
Results are relevant to several areas of ongoing work to improve assessment
tools. In the near term, a test producing a ΔV of 50–55 km/h and a mean
deceleration of approximately 18 g may provide the best opportunity to evaluate
crashworthiness performance under conditions representative of the greatest
share of injuries.