This paper examines the repeatability of the Insurance Institute for Highway Safety’s small overlap frontal crash test, based on repeated tests of six midsize vehicle models. Vehicle accelerations, structural measures, restraint system performance, dummy kinematics, and dummy injury measures were compared. Vehicle longitudinal acceleration pulses were similar in repeated tests of the same vehicle. The testto- test differences of the least repeatable vehicle structural intrusion measurement targets ranged from 4 to 8 cm, with the vehicle models having higher levels of structural intrusion showing the most variation. Restraint system deployments were not always repeatable because many vehicle restraint systems were not yet tuned for this crash mode. In vehicles where restraint systems performed consistently, similar dummy kinematics was observed. Head, neck, chest, leg, and foot injury measures were similar in repeated tests for these vehicles. In the vehicle where the restraint system did not perform consistently, different dummy kinematics was observed. This resulted in large variations in femur, knee, and tibia injury measures.

None of these vehicles would have received different component or overall ratings whether the ratings were based on the results from the original or repeat test. The largest variations observed in this study were unimportant to the overall assessment of the vehicle, as measures from either test would promote the same design changes. Higher levels of variability likely reflect the fact that many of the vehicle structures and restraint systems were not specifically designed for this load case. Repeatability can be expected to improve as vehicles are redesigned to take the small overlap crash into account.