ARLINGTON, VA

Four of eight midsize cars earned good ratings in side impact crash tests recently conducted by the Insurance Institute for Highway Safety. None of the cars received a poor rating. The 2004 Saab 9-3, 2004 Acura TL, 2004 Lexus ES 330, and 2005 Mitsubishi Galant are rated good for side impact protection. The 2004 Saab 9-5, 2005 Mercedes C class, and 2005 Volvo S40 earned acceptable ratings. The 2004 Jaguar X-Type is rated marginal.

The better performers among vehicles that earn good ratings are designated "best picks." The Saab 9-3, which also earned a "best pick" designation in the Institute's frontal offset crash test, is the first car to earn this designation in both front and side crash tests. It's a "double best pick."

"Earning a 'double best pick' is a rare achievement," says Institute chief operating officer Adrian Lund. "The only other vehicle to have done this is the Toyota RAV4, a small SUV, and then only when it was equipped with optional side airbags."

The Institute didn't test 11 other midsize inexpensive to luxury car models because the manufacturers are making design changes to improve side impact performance. The Institute will test these models in the near future.

Institute's side impact test is more challenging than federal test: In the Institute's side impact test, a moving deformable barrier strikes the driver side of a passenger vehicle at 31 mph. The barrier weighs 3,300 pounds and has a front end shaped to simulate the front end of a typical pickup or SUV. In each sidestruck vehicle are two instrumented dummies the size of a short (5th percentile) female, one positioned in the driver seat and one in the rear seat behind the driver.

"The Institute's test is more challenging because the top of the barrier is at the same level as the heads of the test dummies in the car," Lund says. "This is the scenario in real-world side impact crashes where occupants' heads are often struck by the intruding hood of an SUV or pickup."

The federal government's side impact test uses a barrier representing a car's front end. In this test, there's no chance that the heads of the dummies in a struck vehicle will be hit by the intruding barrier.

Saab 9-3 is top performer: All eight cars in this round of crash tests are equipped with standard side airbags with head protection. The Saab 9-3 has side curtain airbags that deploy from the roof as well as airbags that deploy from the side of the car's front seats to protect occupants' chests. All 9-3s built after December 2003 have been modified to improve occupant protection in side impacts.

"The curtain airbag did a good job of cushioning the heads of both the driver and rear passenger dummies," Lund says. "Plus injury measures recorded on the neck, torso, pelvis, and left leg of the dummies all were low. Saab can still make improvements to the 9-3's structure, but in all other areas this was very good performance."

Galant goes from poor to good: This is the second time the Institute has tested the Galant. The 2004 model earned a poor rating for side impact protection when it was tested earlier this year. While side airbags that protect only the torso were available as an option, head protection airbags weren't available. If side airbags are optional, Institute policy is to test the vehicle without the option. The manufacturer may request an additional test with the optional airbags, but in this case Mitsubishi did not request the second test.

"The structure of the 2004 Galant performed very well, but without side airbags the driver dummy's head was struck by the intruding barrier," Lund explains. "There was a possibility of a serious skull fracture or brain injury. Rib fractures and other internal injuries also were possible."

For the 2005 Galant, Mitsubishi added combination head and chest protection side airbags for front-seat occupants as standard equipment and asked the Institute to test the new model.

"The difference in the two tests is dramatic," Lund says. "This time the driver dummy's head was protected from impact with any hard structures, including the intruding barrier."

Even though there was still the possibility of rib fractures or internal organ injuries, the forces recorded in these areas were lower than in the 2004 Galant. In the tests of both the 2004 and 2005 models, the rear dummy's head hit the window frame and sill of the rear door, but these hits didn't produce high head injury measures.

The Galant improved from a poor performer to one of the better performers in the side impact test.

"Mitsubishi should be commended for making side airbags with head protection standard in this relatively inexpensive car. The results for the Galant show that you don't have to spend a lot of money to get good protection in side impact crashes," Lund points out.

Side airbags reduce risks in real-world crashes: Institute research shows that side airbags with head protection are reducing deaths by about 45 percent among drivers of cars struck on the driver side. Side airbags that protect the chest and abdomen, but not the head, also are reducing deaths but are less effective (about a 10 percent reduction in deaths). Before the availability of head-protecting airbags, there was virtually nothing to prevent people's heads from being struck by intruding vehicles or rigid objects like trees or poles in serious side impacts.

"These crash test results confirm what the Institute found is happening in real-world crashes," Lund says. "Side airbags designed to protect people's heads can prevent very serious head injuries. This test is resulting in more manufacturers adding side airbags as standard equipment."

Jaguar is rated marginal: Side airbags with head protection do a good job of protecting people in side impact crashes, but the side structures of passenger vehicles also need to be strong to prevent major intrusion into the occupant compartment.

In serious side crashes some intrusion is inevitable, but it should be minimized. The Jaguar's structure was marginal because the sill below the doors and the B-pillar between the doors were severely crushed.

"This contributed to high forces recorded on the torso and pelvis of the driver dummy," Lund points out. "Head protection in the Jaguar was good, but a real driver might have sustained serious rib fractures, internal organ injuries, and possible pelvic fractures."

How vehicles are evaluated: Each vehicle's overall side evaluation is based on injury measures recorded on two instrumented SID-IIs dummies, assessment of head protection countermeasures, and the vehicle's structural performance during the impact. Injury measures obtained from the two dummies, one in the driver seat and the other in the rear seat behind the driver, are used to determine the likelihood that the driver and/or passenger would have sustained serious injury to various body regions. The movements and contacts of the dummies' heads during the crash also are evaluated. This assessment is more important for seating positions without head-protecting airbags which, assuming they perform as intended, should prevent injurious head contacts. Structural performance is based on measurements indicating the amount of B-pillar intrusion into the occupant compartment. Some intrusion into the compartment is inevitable in serious side impacts, but any intrusion that does occur should be uniform both horizontally and vertically and shouldn't seriously compromise the driver or passenger space.

For more information go to www.iihs.org