With the introduction of fully-automated vehicles, new seating configurations of the passenger compartment has been proposed. Rearward facing front seats are considered to provide so-called living room seating. At least as long as conventional and fully-automated vehicles share the same roads in mixed traffic, crashes may occur. Occupant protection on a rearward facing seat must therefore be on the same level as on a forward facing seat to comply with legal requirements. In order to assess dummy response on a rearward facing seat in a 56 km/h full frontal impact, sled tests were performed, analysed, and discussed.

A total of 23 sled tests in three series with a Hybrid III 50 th percentile adult male dummy were performed to simulate a vehicle frontal impact against a rigid barrier at impact speeds up to 56 km/h. In the first test series, a serial vehicle seat was used, but it showed already considerable deformation at an impact speed of 40 km/h. Therefore, a generic concept seat was developed. In the second test series, the concept seat was tested and tuned to enable it to perform tests at the target impact speed of 56 km/h. In the third series, tests to investigate repeatability were performed. Dummy loadings at 56 km/h were compared with reference values from legislation and literature. Focus was set on thorax and lumbar spine loadings.

For a qualified interpretation of dummy loadings and the performance of the restraint system, the crash was divided into three phases:​ (1) impact phase until the maximum dummy rearward displacement, (2) dummy rebound before interaction with the seat belt, and (3) dummy in rebound and interaction with the seat belt. The impact phase (1) is characterized by the highest 3 ms chest acceleration, close to 60 g in 56 km/h tests. Notably, this was the loading closest to the injury assessment reference value (IARV). The lumbar spine was mainly loaded in compression with forces rising up to 5.8 kN. Chest deflection of about 8 mm was caused by inertia of the dummy rib cage. The rebound phase before interaction (2) did not show any substantial dummy loading. The rebound interaction phase (3) was influenced by the seat belt system, chest deflection ranged from 5 mm in the test with lap belts to 19 mm in the test with two crossed shoulder belts (crisscross belt). The viscous criterion was below 0.1 m/s in all tests. Overall, the tests showed good repeatability and the ability of the generic concept seat to control dummy kinematics.

A limitation of our study is, that only full frontal loading directions were studied, dummy kinematics of oblique impact direction, simulating e.g. +30° impacts to the barrier, were not included. The head rest was not in focus of our investigation and the head was fixed to the head rest without any gap in between.