Improved Side Impact Protection: A Review of Injury Patterns, Injury Tolerance and Dummy Measurement Capabilities
Monash University Accident Research Centre - Report #147 - 2001
Full report in .pdf format [1.6MB]
T. Gibson, B. Fildes, H. Deery, L. Sparke, E. Benetatos, M. Fitzharris, J. McLean & P. Vulcan
Side impact crashes account for a substantial proportion of injuries and Harm to Australian passenger car occupants. Twenty five percent of serious casualties and 28% of fatalities to vehicle occupants in Victoria occurred from side impacts. This crash configuration accounted for one-third of vehicle occupant Harm (approximately A$1 billion annually) during the early nineties.
This project aimed to develop a method for optimising the safety systems of new passenger vehicles to minimise occupant Harm in side impact collisions.
A representative sample of crashes and injuries to occupants in Australian passenger cars involved in real-world side impact crashes was analysed in terms of injury incidence and societal harm.
The analysis of side impact crash data has been used to set the priorities for the development of a means of assessing injury in side impacts. The available human impact tolerance data for side impacts was reviewed, as a basis for the development of the necessary Injury Assessment Functions IAFs. Preliminary values for lateral injury criteria have been proposed, mainly based on cadaver test results.
The steps required to develop and validate the required IAFs are described.
Side impact crashes account for a substantial proportion of injuries and Harm to Australian passenger car occupants. Twenty five percent of serious casualties and 28% of fatalities to vehicle occupants in Victoria occurred from side impacts, which accounted for one-third of vehicle occupant Harm (approximately A$1 billion annually) during the early nineties.
Current attempts at providing increased side impact protection have focussed on new car safety regulations. In Australia all new car models must meet either existing American or European standards (FMVSS 214 and ECE Regulation 95 respectively). Mandating these standards has been shown to be financially beneficial, but neither standard guarantees optimum benefits for Australian car occupants.
The objective of this project was to develop a method to allow vehicle designers to optimise the safety systems of new passenger vehicles and minimise occupant Harm in side impact collisions. The study was divided into four tasks:
The collection and analysis of a representative sample of crashes and injuries to occupants in Australian passenger cars involved in real-world side impact collisions;
The refinement of the Harm matrices at MUARC to obtain an indication of the societal cost of injuries of a given severity to particular body regions.
The development of a means of assessing injury in side impacts. This required the formulation of a series of dummy based Injury Assessment Functions for the major body regions to enable the prediction of the risk and severity of injury from the response of a surrogate /dummy occupant in side impact crashes.
The development of a computational model which incorporated the crash profiles, vehicle structural model, surrogate/vehicle interaction model, IAFs and Harm matrices. This model provided a basis for including the minimisation of Harm in design optimisation.
The work completed for Tasks 1 and 3 in the first year of the project has been summarised in this report.
Real World Data Analysis
Three main databases were used to determine the characteristics of side impacts. These included the Crashed Vehicle File of crashes involving hospitalization or death, the Federal Office of Road Safety's Fatal file and the Transport Accident Commission of Victoria's minor injury database.
The Crashed Vehicle File indicated that the modal side impact velocity was 25-30 km/h, and that the majority (53.8%) of side impacts are car-to-car impacts and a further 22.2% are car-to-pole impacts. In addition 70% of lateral impacts were near-side impacts and the remainder were far-side impacts.
The analysis found that head injuries constituted the majority of Harm in side impacts. The next most important sources of Harm were the thorax and then the lower extremities. The total head Harm however, was considerably larger in far-side as opposed to near-side lateral impacts despite the fact that near-side impacts constitute the majority of all side impacts. In addition it was found that the mean Harm from car-to-pole impacts was considerably greater than that for car-to-car impacts.
The door was the predominant source of injuries in side impacts although when a front passenger was present, then impacts with other occupants become the primary source of injuries. In car-to-pole impacts injuries from impacts with external objects were also significant.
The results of the real-world data analyses indicated that priority needs to be given to the development of injury functions that indicate the risk of head and then thoracic injuries in side impacts. Anomalies in the data meant that spinal and brain injuries were under-reported and may consequently deserve a higher priority than the statistics on Harm suggest. The data also indicated that the reliable assessment of Harm from dummy responses requires that the dummies reliably replicate the head impacts of occupants in real-world impacts.
The Injury Assessment Functions
The available human impact tolerance data for side impacts was reviewed, as a basis for the development of the necessary Injury Assessment Functions. The preliminary injury criteria developed were mainly cadaver test based and utilised measures suited to the response and measurement capabilities of the specific dummies to be used in the project (BioSID and SID IIs). In order of injury priority based on Harm, the following preliminary injury criteria were selected. All, except those for the neck, were based upon cadaver rather than dummy responses.
For the head, the injury criteria were related to HIC based on the work by Mertz, Prasad & Nusholtz (1996) for frontal head impacts resulting in skull fracture, with allowance for the lower threshold of skull fracture in lateral head impacts shown by JARI.
For the thorax, the injury criteria used were based on compression (as a percentage of full chest width) and the viscous criteria, V•C for cadavers, as proposed by Viano, Asbury, King & Begeman (1989).
For the lower extremities simple dynamic lateral three point bending strength of human long bones were used as a 50 percent risk of fracture for the lower extremities, based on work by Nyquist (1985, 1986) combined with a rotational load limit for the flexed knee, from Gibson Newman, Zellnor, & Wiley (1992). Other injuries to the lower extremities in side impacts were not well enough defined to suggest associated injury criteria.
Shoulder and Upper Extremities
For the shoulder and upper extremities no specific injury criteria were chosen as the current dummies are not able to make appropriate separate response measurements for these body regions. The shoulder region was treated as part of the thorax and no response measurements are yet available for the dummy upper extremities.
For the pelvis, the injury criteria used were based on the pelvic impact force to cadavers, as proposed by Bouquet et. al. (1998).
For the abdomen, the injury criteria used were based on compression (as a percentage of full chest width) and the viscous criteria, V•C for cadavers, as proposed by Viano et. al. (1989).
For the neck, there was little data available for an injury criteria for side impact. The suggested values were based on the single point injury assessment reference values for tension, compression, lateral shear and lateral bending in dummies, as proposed by Mertz (1984).
A brief review was conducted of the biofidelity of the responses of the two dummies used in the project. The dummies were the BioSID and the SID IIs, which represent a fiftieth percentile North American male and a fifth percentile North American female. The available side impact dummies were compared by the method proposed by the International Standards Organisation dummy working group, ISO (1997).
The review of the available human impact tolerance data for side impacts combined with the analysis of the injury priorities revealed those areas where developments of IAFs are necessary. For all body regions the IAFs currently available were inadequate: either, due to lack of knowledge of the injury mechanism for the individual body regions; or, due to mismatches between the injury severity probabilities available (ie. the available IAFs) and the injury severity distributions from real accidents. The priority areas for these developments were found to be the head which forms about 50% of the Harm due to side impact and the thorax, which forms about 22%. In both cases the continuous risk functions need to be extended to cover more of the injury severity distribution found in real accidents.
This project was funded by: Australian Research Council (SPIRT scheme); Holden; Department of Transport and Regional Services, Canberra; Australian Automobile Association, Canberra; and Autoliv Australia