Occupant ejection has been identified as a safety problem for decades, particularly in rollover crashes. While field accident studies have repeatedly demonstrated the effectiveness of seat belts in mitigating rollover ejection and injuries, the use of laminated glass in side window positions has been suggested as a means to mitigate occupant ejection. Limited data is available on the field performance of laminated glass in preventing ejection. This study utilized 1997-2015 NASS-CDS data to investigate the reliability of the glazing coding variables in the database and determine if any conclusions can be drawn regarding the effect of different side window glazing types on occupant ejection.
An initial query was run for 1997-2016 model year vehicles involved in side impacts to evaluate glazing coding within NASS-CDS. Sixteen individual cases were identified where the first-row side window glass was coded as both laminated and as in-place and holed, out-of-place and not holed, out-of-place and holed, or disintegrated from impact. NASS-CDS case summaries and photographs were reviewed and compared against original equipment manufacturer (OEM) specifications for the involved vehicles. The results demonstrated that the presence of laminated side window glass was incorrectly coded in 11 out of the 16 cases.
The coding definitions were revised to query NASS-CDS for vehicles known to be equipped with standard or optional laminated glass in the first-row side windows based upon sources including the National Auto Glass Specifications (NAGS) database and vehicle manufacturer information (e.g., sales brochures, press releases, and other specifications). The new definitions were used in conjunction with the 16 previously identified side impact cases and found to correctly differentiate between the cases with and without laminated first-row side window glazing.
Based upon this revised coding strategy, another NASS-CDS query was run to evaluate glazing type as it relates to rollover crashes involving 1997-2016 model year vehicles to determine the rate and frequency of front seat occupant ejection by glass type. This query resulted in a weighted 2,785,988 front seat occupants involved in a rollover, of which only 0.5% (13,640) occupants were in vehicles equipped with standard laminated front door window glass. Using the weighted occupants counts, there were 193 completely ejected and 277 partially ejected front seat occupants in vehicles equipped with standard laminated front door window glass; these numbers were based on unweighted samples of 10 or fewer. The sample sizes were too small to draw any statistically significant conclusions regarding ejection rates in rollover as a function of glazing type.
The results of this study indicate that NASS-CDS glazing coding alone is not appropriate to identify the presence of laminated glass in specific vehicle window portals. This study also demonstrates that the resulting sample of field accidents is too small to assess the effectiveness of laminated glass in reducing the incidence of occupant ejection. The revised coding strategy may, however, be expanded with newer data and may be used in other databases such as Fatality Analysis Reporting System (FARS). It should be noted that, given the increased implementation of side curtain airbags in the vehicle fleet, it may become increasingly difficult to assess any individual contribution of laminated glass to occupant containment versus ejection.