Injury Mechanisms and Severity in Narrow Offset Frontal Impacts

Abstract

Current standard frontal crash tests include full frontal or 40% offset. Frontal impacts with offsets less than 40% and corner impacts have received little attention. Because of the limited engagement of vehicle structures that would permit less energy dissipation, these crashes have the potential for severe trauma to the near-side occupant. Narrow offset and corner-impact crashes under a frontal impact classification were analyzed using data obtained from the United States Department of Transportation National Highway Traffic Safety Administration crash databases: Crash Injury Research and Engineering Network (CIREN) and the National Automotive Sampling System (NASS) 2000–2006. A subset of crashes that could be defined clearly as narrow offset crashes were then examined. The Collision Damage Classification (CDC) classification was used to obtain only crashes with a “FLEE” code for drivers and a “FREE” code for right front passengers. These codes were used to separate out crashes that had bumper and lower vehicle engagement and less than 41 cm of front vehicle damage. The NASS data indicated that corner impacts (“Y” and “Z” type) and narrow overlap (“L” and “R” type) constitute 26 % each of all frontal impacts. When occupants were severely injured in frontal crashes 20 % of the occupants were in “L” and “R” type crashes and 31 % were in “Y” and “Z” type crashes. The percentage of near side “FLEE” and “FREE” crash occupants with severe injuries by body region demonstrated that head, thorax, upper, and lower extremity injuries dominate the data set. For the 71 CIREN cases, lower extremity injuries dominated. The injury severity score did not correlate well with assumed severity parameters like extent of crush and delta-V, however, crashes with extent zones 2–5 had a greater percentage of occupants with chest and spine injuries than crashes with extent zones 6–9. Vehicle rotation after front impact in extent zone 2–5 crashes may be influential in airbag effectiveness.

INTRODUCTION

Frontal impacts represent the largest portion of crashes that result in serious and fatal injury to the occupants. Current standard frontal crash tests include full frontal or 40% offset. Frontal impacts with less than 40% offset and corner impacts have received little attention. Because of the limited engagement of vehicle structures that would permit less energy dissipation, these crashes have the potential for severe trauma to the near-side occupant. In some early work on crash reconstructions comparing head on crashes with front-corner impacts, Mackay observed differences between injury outcomes for drivers and passengers and presumed that the A-pillar and door structure played a greater role in generating injuries in front-corner impacts (Mackay, 1968). When Mercedes Benz engineers first designed vehicles to perform well against their own 40% offset frontal crash test, they found that, of the total AIS=3+ injuries in a limited real world data set, injuries in frontal crashes reduced from 54% in vehicles without the design to 44% in vehicles with the design feature (Pletschen, et al, 1990). A recent study out of Sweden indicated that crashes with less than 30% overlap and no engagement of the drive-train or longitudinal structures accounted for 26 of 61 (43%) fatally injured belted occupants in frontal crashes (Lindquist, et al, 2004). Because current standard methods to assess vehicle crashworthiness do not typically define narrow offset crashes by examining vehicle structural members, a combined approach using two national databases was used. The purpose of the present study was therefore, to examine narrow offset frontal crashes in more detail and to ascertain mechanisms of injury in these types of crashes.

METHODS

Narrow offset and corner-impact crashes under a frontal impact classification were analyzed using data obtained from the United States Department of Transportation National Highway Traffic Safety Administration crash databases: Crash Injury Research and Engineering Network (CIREN) and the National Automotive Sampling System (NASS) 2000–2006. The NASS database was examined for the number of occupants sustaining injuries in 11-1 o’clock frontal crashes by collision damage classification (CDC) code (Figure 1). The number of occupants with MAIS=3+ injuries by CDC code and risk of MAIS=3+ injury by CDC code were obtained. Risk of injury was defined as the number of injured occupants divided by the number of crashes in a given CDC classification. A subset of crashes that could be defined clearly as narrow offset crashes were then examined. The CDC classification was used to obtain only crashes with a “FLEE” code for drivers and a “FREE” code for right front passengers. These codes were used to separate out crashes that had bumper and lower vehicle engagement and less than 41 cm of front vehicle damage. Occupants with injuries by body region were examined for this subset of crashes.

Figure 1.

The NASS collision damage classification (CDC) coding for frontal impacts with different types of crush profiles.

The CIREN database was queried for frontal crashes with narrow offset by obtaining crashes with < 41cm crush profile to the left or right corners of the vehicle. These type of crashes were isolated using the CDC code in NASS for frontal collisions where the third and fourth codes were “FL” or “FR” (Figure 1). A subset of these CIREN crashes were analyzed using the extensive photo documentation in the database where it could be implied that vehicle structural members were not engaged. Occupant injuries were examined, in general, from the NASS database, and specific injury mechanisms and patterns were analyzed using the CIREN database.

RESULTS

There were just over 9.8 million (20,000 raw cases sampled in NASS) frontal crashes with almost 16 million occupants (35,000 raw cases) for the 2000 to 2006 time period. The NASS data indicated that corner impacts (“Y” and “Z” type) and narrow offset crashes (“L” and “R” type) constitute 26 % each of all frontal impacts (Figure 2). When occupants were severely injured in frontal crashes 20 % of the occupants were in “L” and “R” type crashes and 31 % were in “Y” and “Z” type crashes (Figure 3). The risk of severe occupant injury however, is more evenly distributed – between 2 and 3 % – by crash type (Figure 4). For the subset of frontal crashes with the “FLEE” and “FREE” codes there were a total of 12,832 and 13,310 occupants with AIS=3+ injuries. To aid with identification of injury mechanism, this subset of crashes was further analyzed to include only drivers for “FLEE” crashes and right front passengers for “FREE” crashes. The percentage of these occupants with severe injuries by body region demonstrates that head, thorax, upper, and lower extremity injuries dominate the data set (Figure 5).

Figure 2.

The percentage of frontal impacts in the NASS database with specific CDC codes.

Figure 3.

The percentage of frontal impacts in the NASS database that had occupants with severe injuries by specific CDC codes.

Figure 4.

Bar graph representation of the risk of serious injury in frontal crashes by type of crash from the NASS database.

Figure 5.

The percentage of drivers in “FLEE” and right front passengers in “FREE” type crashes in the NASS database with AIS=2+ and AIS=3+ injuries by body region.

A total of 71 CIREN cases were found that met the inclusion criteria. All but one had frontal airbag deployment and 72 % were belted. There were 50 drivers in left sided crashes and 21 right front seat passengers in right sided crashes. Demographics included 44 males and 27 females; 32 were age 16–34 years and 15 were 55 and older. The vehicle model years were distributed with 46 vehicles from 1993 to 1999 and 25 vehicles from 2000 to 2006. The data set was predominantly passenger cars (47) and the balance were SUV, LTV, or van (24).

For the 71 CIREN cases injuries to various body regions were documented. Pelvis inuries were separated out from other lower extremity trauma. The distribution of the number of occupants with AIS=2+ trauma (Figure 6) by body region was similar to the distribution of the number of occupants with AIS=3+ trauma (Figure 7). There were 11 occupants with AIS=2 loss of consciousness or concussion as their most severe head injury. There were five occupants with AIS=3 brain hemorrhage or hematoma, two with basilar skull fracture, one with AIS=5 diffuse axonal injury, and two with AIS=6 brain stem laceration or head crush. For lower extremity trauma, 33 occupants had AIS=3 femur fractures; 22 occupants had AIS=2+ tibia/fibula fractures; 24 occupants had AIS=2+ pelvis fractures. Lower extremity injury and chest injury were associated most with head injury (Figure 8). For thorax injury, 17 occupants had AIS=2+ rib fractures, nine had lung contusions, and five had AIS=5+ aorta or heart laceration. Lower extremity trauma was associated with 80 % of the occupants that had thorax injury (Figure 9).

Figure 6.

The number of occupants with AIS=2+ injuries in various body regions of the 71 occupants from the CIREN database involved in narrow offset frontal crashes.

Figure 7.

The number of occupants with AIS=3+ injuries in various body regions of the 71 occupants from the CIREN database involved in narrow offset frontal crashes.

Figure 8.

The percentage of occupants with AIS=2+ head injuries that also had injuries to other body regions.

Figure 9.

The percentage of occupants with AIS=2+ chest injuries that also had injuries to other body regions.

The Injury Severity Score (ISS) was used as an overall measure of trauma severity. The ISS however, did not correlate well with vehicle maximum crush (R² = 0.099), crash delta-V (R² = 0.060), or amount of damaged wheelbase (R² = 0.0004). The average ISS was 18 for pole/tree impacts, 18 for matched vehicle crashes, and 24 for mismatched vehicle crashes. Crashes were separated between extent zones 2–5 (crush up to the A-pillar) and crashes with extent zone 6–9 (crush through or beyond the A-pillar). Crashes with extent zones 2–5 had a greater percentage of occupants with chest and spine injuries than crashes with extent zones 6–9 (Figure 10).

Figure 10.

Of the 71 CIREN occupants in narrow offset frontal crashes, the percentage of occupants with AIS=2+ injuries (top) and AIS=3+ injuries (bottom) grouped by the extent zone of the vehicle damage

DISCUSSION

This study examined narrow offset and corner impact frontal crashes from two NHTSA databases. Narrow offset crashes have not been examined specifically for injury patterns to various body regions and possible injury mechanism differences with respect to large overlap or full frontal crashes. The NASS data indicated that the full frontal type of crash (“D” type) constitutes about 41 % of all frontal crashes, whereas narrow offset (“L” and “R” type) crashes plus corner impacts (“Y” and “Z” type) account for 52 % of all frontal impacts. The standard 40 % offset frontal crash is similar to a “Y” type crash, but there is no standard crash test for the narrow offset scenario. When the vehicle longitudinal structure is engaged in the crush, it is generally thought that the frontal crash is easier to classify (Hill, et al, 1993). When these structural members are not involved it is more difficult. Indeed, the current standard classification schemes do not consider the analysis of longitudinal members. The CDC code uses a dimensional quantity (less than 41 cm) to distinguish the type of crash. Recently Lindquist, et al, used a new methodology to characterize frontal crashes that classified according to which structural members of the vehicle were loaded in the crash. They found that a relatively high percentage of fatalities in frontal crashes occurred when the load path was outside of the longitudinal members of the vehicle (Lindquist, et al, 2004). This type of method could not be used for NASS data.

The CIREN data however, contained more photographic documentation of the case vehicle and therefore, the crashes included in the analysis were those in which it could be implied that vehicle longitudinal members were not engaged. The CIREN analysis did not result in many cases with fatalities but had many occupants with severe trauma to multiple body regions. Lower extremity and pelvis injuries dominated the CIREN data set. Thirty-eight out of 71 occupants (54%) had AIS=3+ lower extremity trauma. Similar occupant injury patterns were documented by Volvo researchers that defined “severe partial overlap collisions” as those when less than 50 % of the front end was engaged and exterior longitudinal deformations exceeded 80 cm. These researchers indicated that intrusion of the occupant compartment in these crashes was responsible for a more than three-time higher risk of severe injury compared to other frontal crashes (Planath, et al, 1993). These studies indicate that a more universal definition of narrow offset crashes is needed.

Comparing the CIREN results to the NASS results revealed a slightly greater distribution of AIS=2+ injuries among body regions for the NASS data set. It appeared that the extent of head, thorax, and upper extremity trauma were similar in the NASS data set whereas lower extremity trauma dominated the CIREN data set. These observations were qualitative because the two data sets do not result in sampling “equal” types of crashes. The NASS data set is population-based and uses a sampling schema to “weight” each sample and records all injuries from all occupants from a sampled crash. The CIREN data set samples only severely injured occupants in select recent model year vehicle crashes. There are a greater percentage of fatalities recorded in the NASS database compared to the CIREN database which could account for recording more severe head and chest injuries in the NASS data set. In a recent study out of Australia that used a data set that samples similarly to the CIREN database, the authors also found that lower extremity data dominated their data set. They found that AIS=2+ lower extremity trauma was between 2.6 to 2.9 times more likely in narrow or wider offset crashes compared to fully engaged frontal crashes (Logan, et al, 2007).

A further difficulty in analyzing narrow offset crashes is that the usual indicators of crash severity in other frontal crashes like crush measures and vehicle delta-v don’t apply clearly to these crashes. The current effort could not find any clear indicator of vehicle damage or speed change that correlated well with the occupant injury severity score (ISS). A recent study that examined calculated delta-v using standard crash analysis algorithms with delta-v recorded by onboard event data recorders (EDR) indicated that the amount of frontal overlap influenced the accuracy of the calculated delta-v. In general, a calculated delta-v in crashes with less than 50 % overlap significantly underestimated the delta-v recorded on the EDR (Niehoff and Gabler, 2006).

Perhaps the mixture of crashes with different objects also plays a role. The narrow overlap crashes in the current CIREN data set that engaged a tree/pole demonstrated the same average ISS as those for matched vehicle collisions (ISS=18); the mismatched vehicle crashes had a significantly higher average occupant ISS (ISS=24). Further study of how these multiple factors interact may help define the severity of these crashes by combining factors.

An unusual finding in the current CIREN data set was a higher percentage of occupants in extent zone 2–5 impacts had severe spine and chest trauma than occupants in extent zone 6–9 impacts. All other body regions indicated the expected pattern of more trauma in the higher extent zone crashes. One possible explanation for this difference is that the occupants in the lower extent zone crashes may be subject to different kinematics than those in higher extent zone crashes. Because of the narrow and off-center engagement of the vehicle with the other object, vehicle rotation in the latter stages of the crash may induce occupant movement laterally and place the occupant in a position where airbag effectiveness is sub-optimal. The amount, rate and timing of vehicle rotation may influence occupant injury patterns and interaction with the frontal airbag. For the higher extent zone crashes the vehicle may be engaged in the frontal movement for a longer period of time and thus fully benefit from airbag deployment. This may be why only chest and spine trauma are affected.

CONCLUSION

Because the type of trauma and the injury patterns are different in narrow offset frontal crashes compared to wider offset crashes, countermeasures that are designed with standard frontal crash tests that employ large overlap may not address the specific injuries associated with narrow overlap crashes. Rotation of the vehicle and subsequent occupant movement lateral to the airbag could be a factor in less severe impacts.