Previous pendulum impact tests have shown that knee joint injuries and tibial-fibular fractures may occur when loads are directed against the lower leg rather than directly against the femur in the knee. In order to further improve our understanding of lower extremity restraint mechanics, simulated frontal barrier crash experiments were conducted with unembalmed human cadavers and an anthropomorphic dummy restrained by a two-point shoulder belt. In the first test, an experimental bolster was specifically positioned so that the cadaver's lower leg would strike the bolster, thus inducing restraining loads entirely below the knee joint. The analysis of occupant kinematics showed that the flexed knee rode over and forward of the low-positioned bolster. Restraint induced considerable shearing load across the knee joint. Bolster measurements indicated a peak load of approximately 4.0 kN per leg which resulted in a contralateral central tear of the posterior cruciate ligaments.

Factors influencing the performance of a shoulder belt and knee restraint system were studied in the remaining experiments. Positioning the bolster to provide full knee contact and early femoral loading, and orienting the bolster face to reduce knee joint shear loads (∼ 38.5° from horizontal) were identified as important parameters for improved performance of the knee restraint in these tests. The reaction force vector of the deforming bolster structure (∼ 32.5° from horizontal) represented an additional independent factor in minimizing contact loads and in controlling occupant kinematics. Under these knee restraint conditions, an average peak bolster reaction force of 6.2 kN per leg (compared with a dummy femur load of 5.9 kN) did not result in injury to the lower extremity of four cadavers. The experiments indicated that a two-point shoulder belt and knee restraint system can control occupant kinematics and attenuate potential lower extremity injury during a 13.4 m/s frontal crash simulation.