Home » Status Report
Status Report, Vol. 44, No. 2 | March 7, 2009 Subscribe

Crash data of good-rated vehicles could be used to develop new tests

This wreck was fatal, even though the vehicle earned good ratings in frontal crash tests.

Real-world crashes don't always mirror tests. That's why Institute researchers recently looked at how people are dying in crashes of vehicles that earn good frontal ratings. The injury patterns and crash factors they uncovered could be used to develop new programs to help make vehicles even more crashworthy.

Based on patterns of damage to the vehicles in which people died or were seriously injured, the researchers identified 5 frontal crash modes: full-width crashes that resemble the government's 35 mph consumer tests, moderate overlap offset crashes like those the Institute conducts, offset crashes with a smaller overlap than the Institute's test, center impacts into narrow objects like poles, and underride crashes with medium and heavy trucks, SUVs, and pickups. To understand what problems a new test program should address, the researchers also sorted the frontal crashes according to whether the injuries appeared related more to intrusion, restraint factors, or occupant factors.

"We mark the success of our 14-year-old frontal test program by the vehicles that earn good overall ratings and the lives saved by improved designs," says Institute president Adrian Lund. "Still thousands of people die each year in frontal crashes. The continued presence of more than 14,000 such fatalities — 13,374 to front-seat occupants in 2007 — compels us to focus on these deaths and new crash test programs that could encourage vehicle designs to help prevent them."

Researchers analyzed 2000-06 case files from the National Automotive Sampling System/Crashworthiness Data System on frontal crashes in which 116 drivers and right front passengers were seriously injured or killed despite using safety belts. The vehicles had to have good frontal crashworthiness ratings in the Institute's offset test, and all but 1 earned 4 or 5 star ratings out of the maximum 5 in the federal government's 35 mph fullwidth New Car Assessment Program (NCAP) test.

"We found that a lot of frontal crashes are happening in configurations that aren't represented in the crash tests being conducted right now, either by us or the National Highway Traffic Safety Administration," says David Zuby, Institute senior vice president for vehicle research and an author of the study. "Automakers design vehicles to perform well in these tests, so we wanted to explore if other kinds of frontal tests could address the life-threatening injuries that occur in the field."

Some of the crashes in the study were so extreme that the deaths and injuries probably can't be addressed in crash tests. Some involved very high speeds or unusual conditions. In one fatal case a car went off an open drawbridge.

One of the biggest crash groups in the study involved people killed or injured in vehicles whose safety cages collapsed despite good test ratings. Intrusion was an obvious factor.

In other cases, the vehicles' safety cages did a good job of managing crash forces, but people died because factors such as their age, health, weight, or stature increased their vulnerability. Or the restraints weren't ideally suited for the occupants and the crashes, either allowing an occupant to hit something inside a vehicle or holding an occupant tightly enough to cause injury.

Center impacts

Nineteen percent of crashes in the study were center impacts into trees, poles, or posts. In all but 2 of these cases the vehicle structure prevented substantial intrusion. Almost half of the vehicles hit something else before running off the road and striking a narrow object. Researchers identified 10 of the center crashes as having restraint factors that contributed to occupant injuries and 7 as having occupant factors or both restraint and occupant factors.

Neither the Institute nor the government includes frontal pole tests in their consumer information programs. For research purposes, the Institute has conducted some frontal tests into a 10-inch diameter pole using vehicles rated good, but the injury risk concerns these tests raised didn't seem to match the chest and abdominal injuries real people sustain in similar crash modes.

The dummies used in frontal crash tests and the chest injury criteria may need to be revised to better reflect injury risks, Zuby says. This work also has prompted the Institute to consider how the position of occupants and the performance of their restraints are influenced by off-road excursions like tree impacts.

Full-width crashes

Only 6 percent of the occupants were in full-width impacts like those simulated in NCAP tests. None of the occupants died, and the vehicles had little or no intrusion.

"Based on this sample of cases, relatively few restrained occupants seriously injured in frontal crashes are in impacts resembling the NCAP test setup," Zuby says. "This likely wouldn't have been the case 20 or 30 years ago, back in the early days of the testing program before safety became a sales strategy. Because most vehicles now earn top marks for frontal crashworthiness in NCAP tests, it's clear this program has done a good job of influencing designs that protect people in head-on full-width crashes, and the real-world crashes we studied bear this out."

Moderate overlaps

The configuration of the Institute's offset test is a moderate, 40 percent overlap of the driver's side of the vehicle. This type of crash accounted for 24 percent of the cases studied.

About half of the belted people seriously injured in moderate overlap crashes were in vehicles that otherwise held up pretty well, so other factors besides vehicle structure came into play.

In the other half of the moderate overlap crashes studied, the vehicles' safety cages allowed substantial intrusion into the occupant compartments, which likely contributed to injuries. These intrusion crashes also occurred at much higher speeds than the Institute's test, so the conditions were more severe.

"Manufacturers design cars, minivans, SUVs, and pickups to do well in our offset test," Zuby says. "But these real-world crashes with intrusion show that for some vehicles, at least, occupant protection drops off as frontal impacts become more severe."

This suggests that the Institute's offset test "may not be driving the kinds of improvements in vehicle structures and restraint systems that are needed to protect people in higher severity crashes. We might need to consider conducting additional tests," Zuby points out. He adds, "One concern, though, is that the kinds of restraint system and structural changes that more severe offsets would necessitate could be counterproductive in other kinds of crashes."

Small overlaps

Another 24 percent of the study cases were small overlaps. That is, they didn't involve very much of the vehicles' front ends. These were most common among crashes in which intrusion contributed to injuries. Forty-three percent of crashes with a lot of intrusion were small overlaps involving around 20 percent or less of the vehicles' front ends. The Institute's current offset test isn't intended to evaluate crashworthiness in such cases.


New crash tests could help spur vehicle designs that do a better job of protecting people in small overlap crashes like these — and in some full-width and moderate overlap crashes, too.

Few vehicles have energy-absorbing structures that extend to the edges of the front end, and this may be why so many small overlap crashes involved intrusion into the occupant space. Instead, more metal is concentrated toward the centers of vehicles, which helps limit intrusion in center impacts with things like trees, posts, and utility poles.

"Based on results of this study, we think small overlap tests have good potential to improve the crashworthiness of new vehicles," Zuby says. "This would help spur vehicle designs with beefed up structural protection across the full width of the front, including the outer corners. It also would result in better protection for some occupants in moderate overlap and full-width crashes. We plan to conduct some small overlap crash tests to see if an evaluation program on a wider scale is justified."

Crashes that involved underride

These are when passenger vehicles go partially or all the way under a larger vehicle such as a truck rig. After small overlap crashes, underrides accounted for the most impacts in which intrusion was an injury factor.

Eight of the underride crashes involved heavy-duty truck rigs or medium trucks, and the other 6 involved impacts with pickups or SUVs. Half of these 14 underride cases were fatal.

Underrides that involve SUVs and pickup trucks have been addressed to some extent by the voluntary work auto manufacturers are doing to alleviate the incompatibility in crashes between these large passenger vehicles and cars. Still the prevalence of truck underride crashes suggests a need for improved guards on large trucks.

This isn't a new idea. In fact, a 1997 Institute study estimated that about half of all fatal crashes between large trucks and cars involved front, side, or rear underride (see "FARS undercounts fatal large truck-car underride crashes," Feb. 15, 1997).

©1996-2018, Insurance Institute for Highway Safety, Highway Loss Data Institute | www.iihs.org