Cadaveric injuries in simulated collisions are often more severe than those observed in the field for a given barrier equivalent velocity. Unembalmed cadavers are flaccid while all accident victims have muscle tone. Although muscle tension may not be effective during the total period of deceleration, it can be a factor at its initiation and in controlling the subsequent kinematics of the victim.

Nineteen cadavers, restrained with a three-point belt system were subjected to simulated frontal collisions with sled equivalent velocities of 32 km/h to 65 km/h. In one group the knees were flaccid. In the other, braces were used to simulate quadriceps function by providing a constant resistance of 61 N-m. Kinematics were analyzed by using high speed cinephotography and injuries were assessed by roentgenography and necropsy.

The group with no quadriceps function submarined at 64 km/h and sustained severe intra-abdominal injuries which were far more serious than those reported from field studies. In the group with the knee brace there were only a few abdominal injuries. Film analysis showed no significant submarining.

It is concluded that quadriceps function plays a role in the prevention of submarining in frontal collisions and that, in cadavers, muscle action should be simulated when evaluating three-point harness restraint systems.