Jump to Content

CR 176: Passenger Roof Crush Strength Requirements (1998)


Rollover crashes, especially in the country, are usually very destructive events. About 15% of passenger cars in fatal crashes in Australia have overturned. Between about 13% and 16% of all passenger-car occupants killed in Australia died primarily as a result of injuries received in a rollover. Vehicle damage often includes deformation of the roof and its supporting structures. Head and neck injury are common, and associated with roof deformation. Strengthening of the roof is often suggested as an appropriate countermeasure for such injuries.

The Department of Transport, through the ATSB Road Safety, requested a review of the costs, benefits and feasibility of introducing a new Australian Design Rule based on the relevant US rule, which is Federal Motor Vehicle Safety Standard 216. This would apply in Australia only to passenger cars, and not include convertible models.

Review of the local statistics indicates that if a roof crush standard (or any other measure) were perfectly effective in preventing death or injury in rollover, the maximum benefit for belted occupants would be in the order of 30 deaths and 140 serious injuries prevented a year.

The cost to Australian manufacturers of introducing a standard based on FMVSS 216 is estimated by the Federal Chamber of Automotive Industries to be, for those small minority of current models (some 2%) which are believed not to comply with FMVSS 216, $125,000 (average) per body style for development and testing programs.

Of those current models that are believed currently to comply with FMVSS 216, $375,000 (average) per body style for certification requirements is estimated. In respect to future models, estimated costs for design, development, testing and certification to FMVSS 216 are estimated to be in the order of $85,000 per body style.

The relationship of roof crush and strength to injury is the fundamental issue in determining whether roof strength standards (including FMVSS 216) might be of value in Australia. It has long been taken for granted that roof crush is directly and causally related to occupant injury. It is envisaged that the roof is forced "down" upon the head and neck of the occupant as the car overturns, and that this mechanism is the direct cause of the injury.

However, although there is almost certainly an association between roof strength and head/neck injury in rollover, whether this association is causal remains a matter of debate. Recent, comprehensive statistical studies have confirmed a positive relationship between roof damage and occupant injury. What has not been shown, however, is any relationship between differences in roof strength as measured in the test used for FMVSS 216 on injury outcomes.

Rollover testing has also resulted in debatable conclusions on the relationship between roof deformation, roof strength and occupant injury. The dummies used for such testing are not well suited to rollovers, because of the lack of biofidelity of the dummy neck. Further, rollover test conclusions have been based on biomechanical criteria - in particular, axial neck loads - that in themselves are open to doubt on their real-world validity.

For the purpose of this project, slow motion analysis of the videos of real-world rollovers in rally competition was performed. This revealed that substantial changes in the angular velocity occur as parts of the vehicle contact the ground. This results in high tangential forces on the occupants. The head and arms of occupants, despite restraints, commonly extend well outside open or broken side windows.

The combination of vertical acceleration/deceleration, horizontal decelerations and rotational acceleration/deceleration generally results in complex occupant kinematics during a roll-over. Occupants are thrown from side to side and up and down in a chaotic manner. Partial ejection through open or broken side windows is a strong possibility, even for restrained occupants. Roof damage mostly results from a combination of vertical and horizontal loads on the roof and its supporting structures.

Some testing of Australian vehicles to FMVSS 216 has been performed in Australia, at Monash University. With the exception of a 1990 sedan all vehicles passed this test. The results from these tests confirm that the loading which in the end defines the crush is a bending one on the A pillar, rather than an axial load. This is in accordance with field observations. The windscreen, and its bonding to the body structure, therefore has great influence on the resistance to crush, because the screen is supporting the pillar.

In the view of the present consultants, the FMVSS test method used to assess strength is unrelated to the kind of strength that is required in rollovers - particularly, resistance to bending of the A pillars (and to some extent bending of the B pillars) after the windscreen has broken. The kind of strength that is required will be more able to withstand inverted impact in the presence of forward motion, as well as impact with the ground after end-over-end and launching rollovers, where height from the ground has been gained and the vertical velocity on impact is substantial.

Thus, the main conclusion of this review is that the FMVSS 216 is an inadequate standard, and that there would be little or no incremental benefit in introducing an Australian Design Rule based on it.

It is recommended as follows:

  • that through international forums the ATSB Road Safety should closely monitor, and where appropriate encourage, moves to update and improve the existing FMVSS 216;
  • that in the short term the ATSB Road Safety review the feasibility of introducing an Australian Design Rule based on the newly amended FMVSS 201 for head impact protection;
  • that investigations be mounted into the incidence of ejection of restrained occupants, which could in turn be related to inadvertent unlatching of the seat-belt buckle and instability or weakness of the seats and their mountings;
  • that related vehicle design improvements identified in the work of Rechnitzer and Lane, at Monash University, should be the subject of further study.

Rollover is an important cause of injury in road accidents in Australia. It is considered that improvement in roof strength (perhaps in certain key impact directions), along with other countermeasures, would decrease the incidence of rollover-related injuries, not only from contact with the ground during inversion but also from contact with roadside obstacles.

However, it is not considered that the introduction of an ADR based on FMVSS 216 would have more than a minimal effect among such countermeasures, and thus its introduction cannot be justified in its present form.

Download Complete Document: Roof_Crush [PDFPDF: 478 KB]

Type: Research and Analysis Report
Sub Type: Consultant Report
Author(s): M Henderson, M Paine
ISSN: 0810-770X
Topics: Occ protection, Vehicle design
Publication Date: 01/10/98


Last Updated: 6 May, 2013