April 11, 2005
Expected NHTSA Roof Crush Standard Will Not Improve Survivability for Most Occupants in Vehicle Rollovers
Statement of Joan Claybrook, President, Public Citizen*
Today we are going to present evidence that demonstrates beyond any doubt that an effective
Department of Transportation roof crush safety standard could readily be written to save thousands of lives and mitigate thousands of injuries of rollover crash victims – a combined total of close to 10,000 each year. This is in stark contrast to the statements by the National Highway Traffic Safety Administration (NHTSA) over the past year that the proposed rollover safety standard it is preparing to issue in June or July will save only 50 of the 10,400 lives now lost in rollover crashes. Again and again we have been asked how the estimated number of lives saved by a safety standard could be so different. Today we will answer that question.
I am joined here by Don Friedman, an engineer and consultant who will present his analysis of data demonstrating why NHTSA’s expected new roof crush standard will be ineffective. Mr. Friedman developed the sidewinder missile for the military, was an engineer at the GM research labs for eight years, and designed the NHTSA Research Safety Vehicle in 1977 that crashes without occupant injury in 30 mph dynamic dolly rollover tests. In the past 15 years, Mr. Friedman has investigated some 400 rollover crashes in depth, gathering far more data than are available to NHTSA in NHTSA’s data system. In addition, he has published more than 20 peer-reviewed papers on head and neck injuries in rollovers, and in the past four years he has submitted 19 comments to the roof crush docket.
Ensuring that occupants can survive when vehicles roll over is probably the single most effective step that the automotive industry and NHTSA can take to reduce the unacceptable carnage on our highways. Rollover crashes are responsible for about a fourth of all traffic fatalities and about one-third of all occupant fatalities each year. There are about 27,000 casualties (deaths and very serious injuries) each year in rollovers, including about 12,000 head injuries and 3,000 spinal cord injuries, including 500 cases of quadriplegia. Overall, about 27 percent of all harm in rollovers involves the head and neck, an area of the body in which injury often has severe and deadly consequences.
About 6,000 to 7,000 of the rollover deaths occur because vehicle roofs do not withstand the impact of the rollover, but instead intrude into the occupant compartment, reducing the survival space for the occupant, undercutting the vehicle structure that is supposed to protect occupants and permitting ejection. Many of these deaths are entirely preventable with vehicle designs similar to the Volvo XC 90, which has roll-cage roof strength, pre-tensioned belts, side window laminated glazing, and side head air bags that inflate in rollover crashes, drastically reducing not only head injury but ejection.
Given the huge number of lives and injuries involved, it is imperative that every effort be made to mitigate this problem with an effective government safety standard as rapidly as possible. The existing roof crush safety standard, Federal Motor Vehicle Safety Standard (FMVSS) No. 216, has not been upgraded since 1971. It was originally intended to be an interim standard until adoption of a mandatory dynamic dolly rollover test that is included as an optional test in FMVSS 208, which covers restraint systems. But this never happened.
All vehicle manufacturers have dolly rollover test systems and regularly test their vehicles with them. But the manufacturers now say that the dolly rollover test is unacceptable because when the same vehicle is tested again and again, the roof crush patterns are not identical or repeatable. In fact, they don’t like any dynamic rollover tests. There are a variety of them now, including the newest one, the Jordon Rollover System test device that Don Friedman has been using. As we showed last week, occupant injury patterns – the most important measure – are in fact repeatable with a dynamic test.
Unfortunately, rather than require a new test for FMVSS 216 that comprehensively measures the full impact of rollover crashes, NHTSA has signaled that it intends to barely upgrade a long-outdated test that will do little to improve safety for the occupants who are most at risk. Mr. Friedman’s research demonstrates that the occupant who is far more likely to be killed or seriously injured is the one who is sitting on the trailing side in a rollover crash, in large measure because of damage to the A-pillar (the pillar that hold the windshield in place). That’s because the first impact usually deforms the roof structure, particularly at the A-pillar, knocking out or crushing the windshield (which provides 30 to 40 percent of the strength of the roof) and side windows, making the second impact far more likely to cause the roof to crush into the occupant’s head and neck and opening portals for full or partial ejection. Belts often do not hold occupants in place because there currently is no safety standard for belt performance in rollovers, and they frequently spool out during the course of the rollover. At the end of our presentation we will show a very short film revealing how the roof deforms and then is pushed back into place during a rollover, causing injury along the way. It makes the case for a dynamic test.
But under NHTSA’s expected new requirement, automakers will continue to test vehicles with only a “static” test that involves a metal plate pressing down on just one side of the vehicle, primarily putting force on the B-pillar beside the occupant’s shoulder. We anticipate that under the new standard, the vehicle will be required to withstand pressure equal to only two-and-one-half times the vehicle weight compared to one-and-one-half times under the current standard. This does little to improve occupant protection in the seat below the second impact point because it does not increase the roof strength or the A-pillar enough. Also, the roof strength average of vehicles on the road today is already 2.3 times the vehicle’s weight.
What is needed is a dynamic test that actually rolls the vehicle to simulate the injuries that occur in a real-world crash. This is the only way to test what happens in a rollover crash to the roof, its supporting structures, the windows and the belt system, and of course, to the occupant.
Today, we are also releasing a compendium of industry documents dating to the 1960s showing that General Motors’ own tests demonstrated a clear link between roof crush and injury. When GM tested the second side of vehicles’ roofs in 1970, the second side of the roofs on six of seven vehicles failed. And a 1982 GM study agreed that ejection, window breakage and roof strength were related. Yet NHTSA rulemaking documents ignore ejection risks when calculating the benefits of a roof strength standard.
The documents are also important because the auto industry for years has maintained that roof strength is unrelated to injuries in rollover. Industry officials claim that people sustain head and neck injuries when they “dive” (are thrown) into the roofs of their vehicles, not when the roofs crush into the people’s heads. The documents bolster findings in a report released last month by Public Citizen showing that automakers have misled government regulators and the public for years by claiming that roof strength and injuries in rollover crashes are unrelated.
The government has a golden opportunity to save lives that are lost in rollover crashes, but all indications are that officials are going to blow it because NHTSA has succumbed to unsupportable data from automakers that are simply not true. The only way to determine if a vehicle’s roof will hold up in a rollover crash is to test it under real-world crash conditions. Strong roofs save lives, and weak roofs kill. NHTSA should not have waited 30 years to implement a weak test that ignores what happens when vehicles roll over.
*Joan Claybrook was administrator of NHTSA from 1977-1981.
To read the press release, click here.
For more information, click here.