Using real-world crash data to develop new crash tests

Crash tests at the IIHS Vehicle Research Center normally follow the same procedures day in and day out to produce comparable results for vehicle ratings. Lately, however, IIHS engineers and technicians have been mixing it up as they try out new test configurations — not for current ratings but for the evaluations of the future.

The new study of real-world crashes in which rear-seat occupants were killed or injured points to problems with rear-seat protection. Now research tests are being conducted to figure out what kind of evaluation will help distinguish vehicles that offer better protection in the rear seat from those that are lagging.

A second group of tests aims to solve another crash-testing puzzle: how to modify the successful IIHS side test to encourage even better protection in right-angle crashes.

For both research programs, the goal is to come up with a draft protocol for a new evaluation by the end of the year and incorporate a new test into IIHS ratings for 2022.

"All of our tests start with real-world crash data," says Joe Nolan, IIHS senior vice president for vehicle research. "We look for trends in the crashes that leave people with serious or fatal injuries. If those crashes have things in common that current ratings don't capture, we'll try to fill that gap in the ratings."

That was true of the very first IIHS evaluation, the moderate overlap front test. The Institute began the program in 1995 to complement the frontal evaluation that the National Highway Traffic Safety Administration (NHTSA) had been doing as part of the New Car Assessment Program started in the 1970s.

In the NHTSA frontal test, vehicles are crashed at 35 mph into a rigid barrier that covers the full width of the vehicle. This test spurred big improvements in occupant protection, but didn't reflect the full range of real-world frontal crashes. In the 1990s, IIHS researchers found that more than half of frontal crashes in which a vehicle was towed involved two-thirds or less of the front end.

In the IIHS moderate overlap test, 40 percent of the front of the vehicle strikes the barrier.

Once automakers mastered the moderate overlap test, IIHS researchers looked for other ways to raise the bar for frontal crash protection. In a 2009 study of vehicles with good ratings in the moderate overlap test, nearly a quarter of the frontal crashes in which front-seat occupants were seriously or fatally injured involved only a small portion of the vehicle's front end. This finding led to the development of the small overlap front test, in which 25 percent of the vehicle strikes the barrier.

The Institute's side crash test also came about because of a gap in existing evaluations. NHTSA's moving-barrier side test relies on a barrier developed in the early 1980s, when most of the vehicles on the road were cars, not SUVs and pickups. The NHTSA test doesn't account for the much greater risk of head injury from impacts with taller vehicles. In 2003, IIHS began its own testing with a barrier that mimicked the height and shape of the front end of the typical SUV or pickup on the road at the time.

The current side-impact research tests build on a study of real-world side crashes in which people were seriously injured in vehicles with good side ratings from IIHS. In that study, many of the impacts occurred further forward on the vehicle than the spot where the barrier strikes in the current IIHS test. Many also occurred at higher speeds than the 31 mph used in the test.

IIHS engineers have also taken a closer look at the moving barrier and found that it may need to be redesigned to more closely approximate the front ends of SUVs and pickups.

The research tests for rear occupant protection in frontal crashes are looking at what size dummies to use in the rear seat. Engineers are also considering whether to simply add the rear-seat dummies to an existing frontal evaluation or modify other aspects of the test too. The latter route would provide an opportunity to learn something new about front-seat protection at the same time.