Multivariate head injury threshold measures for various sized children seated behind vehicle seats in rear impacts

Government recommendations to place children into the rear areas of motor vehicles to avoid airbag induced injuries have been complicated by the fact that most adult occupied front seats will collapse into the rear area during rear-impacts, and thus pose another potentially serious injury hazard to rear-seated children. Many variables affect whether or not a front seat occupant will collapse into the rear child, and whether that interaction could be injurious to the child. For instance, the severity of rear impact, coupled with front and rear occupant sizes (mass and stature), and the level of front seat strength, all interrelate to influence whether or not a rear seated child is likely to be impacted and possibly injured. The most common types of child injuries in these instances are head and chest injuries. In this study, a "high-low" experimental method was employed with a multi-level "factorial analysis" technique to study "multivariate" biomechanics of child head injury potential determined from rear-seated 3 and 6 year-old child surrogates in different types of vehicle bodies mounted to a sled system. The sled-buck systems were towed rearward into crushable barriers that matched the crash pulses of the vehicle types being tested. Various sizes of adult surrogates (i.e. 50 kg up to 110 kg), seated in both the "typical" low strength "single recliner" collapsing type front seat (i.e. 3.2 kN) and a much stronger "belt-integrated" seat design (i.e. up to 14.5 kN), were tested in the two different "sled body-buck" set-ups at various impact levels (i.e. 22.5 to 50 kph). One set-up used a popular minivan vehicle body with "built-in booster" seats for the 3 year-old. The other used a 4-door family sedan vehicle body with the 6 year-old in a standard rear bench seat. The parameters of the tests enabled the experimental data to be combined into polynomial "head injury" functions of the independent variables so the "likelihood" of rear child head-injury potential could be "mapped" over ranges of the key parameters. Accident cases were compared with predictions to verify the methodology.