Where the dummy is positioned in a crash test isn't necessarily where a real person would sit in a vehicle, and the difference matters. In fact, the dummies in federally mandated crash tests may be positioned so far forward that real drivers sitting this close to their steering wheels might be at risk. Fortunately, most drivers choose to sit farther back.
The whole point of crash testing is to assess how well real people in real-world crashes of similar severity would be protected. But there's a problem with the tests mandated by the National Highway Traffic Safety Administration (NHTSA) to comply with motor vehicle safety standards. The test procedures can result in dummy positions that don't reflect where many people choose to sit in their vehicles. So the results might not reflect actual risks in real-world crashes.
In 2002 the Institute and University of Michigan Transportation Research Institute (UMTRI) petitioned NHTSA to take steps toward more realistic seating positions in compliance tests to meet federal standards as well as in tests the agency conducts for consumer information. The petition spelled out the problems with the positions NHTSA currently specifies for testing and recommended new procedures based on extensive study of driver seating preferences.
NHTSA denied the petition. Now the Institute and UMTRI are asking the agency to reconsider the denial, saying NHTSA's analysis of the 2002 petition was "flawed by error and misunderstanding."
Tests don't reflect actual risks in crashes
To comply with Federal Motor Vehicle Safety Standard 208 (occupant crash protection), automakers run frontal tests using dummies representing average-size men and small women. The male dummy has to be placed in the driver seat adjusted to the middle position of the front-to-back seat track range. The female dummy's seat is adjusted full forward.
"Testing with these dummies is supposed to give an indication of the protection afforded to both small and average-size occupants and people in between," says Institute chief operating officer Adrian Lund. "But this is true only if the dummies are seated where real people of their sizes would sit."
In fact, the dummy positions often are farther forward. In some cases the dummies are in positions considered hazardous because they're in such close proximity to driver airbags. A number of studies, including one conducted by NHTSA and Ford staff, indicate that men and women typically choose to sit farther back than prescribed by the standard. These differences average about 42 millimeters, or less than 2 inches, which is small, but it's enough so that real motorists who are involved in real crashes might not fare as well as the dummies do when they're positioned in crash tests according to NHTSA procedures.
Under NHTSA procedures for positioning dummies in crash tests, an average-size male dummy in a Land Rover Discovery would sit about 9 inches from the steering wheel. This is closer than most drivers choose to sit in their vehicles, and it's closer to the steering wheel than NHTSA recommends for safety.
Under UMTRI procedures, the same dummy positioned in the same vehicle would sit about 13 inches from the steering wheel. These procedures are based on where 120 test subjects chose to sit in 36 different vehicles. For the second time, the Institute is asking NHTSA to adopt UMTRI's seating position protocol.
Small differences can affect test results
During tests to comply with the occupant crash protection standard, a vehicle slows down while the dummy continues moving until it's stopped by the restraints or contacts the vehicle interior. The farther a driver dummy is positioned from the steering wheel, the more space there is to continue its forward motion while the vehicle slows down. This means there will be a bigger speed differential and, in turn, higher impact forces and/or greater strain on the restraints.
Positioning dummies more forward can reduce these forces and the injury measures recorded on the dummies.
"This issue of dummy positioning is key to assuring adequate protection across a range of occupant sizes," Lund points out. "How can you know if someone as big as the average-size male dummy would be protected if you don't put the seat as far back in the test as a man this size would sit?"
Put the dummies where real drivers sit
After observing 60 to 120 test subjects choose seating positions in each of 36 vehicles of various types, UMTRI researchers developed a mathematical formula relating driver stature to chosen distances from vehicle controls (steering wheel and pedals). Using this formula, UMTRI and the Institute created procedures to put dummies approximately where people of similar heights would sit. These procedures are based on the dummy's relative distance from the controls instead of the length of a vehicle's seat track.
In denying the Institute's 2002 petition, which asked NHTSA to use the same procedures, the agency cited "a lack of compelling beneficial evidence supporting the UMTRI procedure." NHTSA also cited its own "views about the adequacy of the current seating procedures" but offered no defense of such procedures — an important impetus for the new petition, which says it's "perplexing that NHTSA can dispute the relevance of the large body of data used to develop our proposed dummy seating procedure, while failing to acknowledge that the agency's current procedure is based on no information about human driver positions."
In contrast UMTRI procedures result in positions that "reflect those that would be chosen by real people and are less subject to manipulation by automakers chasing good test results," Lund says.
The Institute and UMTRI are asking NHTSA to reconsider the new procedures for both compliance test specifications and New Car Assessment Program (NCAP) tests for consumer information. The Institute already has begun using UMTRI procedures in its own consumer information tests.
Automakers can influence test results
An important consideration in establishing seating position procedures is that seat tracks can be shortened or lengthened in ways that change dummy positioning and test results but have no effect on where people actually sit in their vehicles.
In particular, Institute researchers believe automakers are manipulating the positions of dummies in crash tests to improve the results. This belief is based on unexplained differences the researchers have found in the front-to-back ranges of seating positions in vehicles built for sale in the United States versus Europe. Such differences are apparent in vehicles that are sold in both markets.
A clip applied to the seat track in this Mercedes C320 has the effect of shortening the track. The most likely reason is to change the dummy positions for optimal results in crash tests mandated by the National Highway Traffic Safety Administration.
The European version of the 1999 Volvo S80, for example, has bolts in the front of the seat track to restrict forward movement. The U.S. version doesn't. This means dummy positions in U.S. compliance tests will be farther forward than if a European version of the S80 were being tested. A similar modification is apparent in the 2001 Mercedes C class.
Dummy seating in both U.S. and European compliance tests is based on seat track length. But the tests themselves differ, which may explain the different track lengths in cars sold in the two markets. In European NCAP side impact tests, for example, dummies' heads are instrumented and capable of recording impacts. If a head hits the interior hard enough, a vehicle can fail the test. Moving the dummy forward or back can help keep the head from contacting hard interior structures such as the B-pillar or assure that the dummy's head will hit the side airbag in the best spot. But this isn't the case in U.S. side impact compliance testing because the dummies' heads aren't instrumented.
"The different track lengths certainly cannot be explained on the basis that same-size U.S. and European drivers would be likely to choose different positions in the same vehicle," Lund notes. "The most likely reason would seem to be to manipulate the dummy positions in whatever crash tests are required in the United States versus Europe."
In denying the 2002 petition, NHTSA ignored the potential for manipulation. "This is troubling," Lund says, "because manipulations could be affecting both regulatory and consumer information test results. Equally troubling is that NHTSA didn't offer any credible evidence that sticking to its current policy is the better way to go. So now we want the agency to take another look and explain how it can be known that someone as big as an average-size man will be protected in a crash if the compliance test never puts the seat where an average-size man would sit. This is why we're petitioning NHTSA again."