The Hybrid III dummy is the test device used to ascertain compliance with the dynamic crash test requirements of FMVSS 208. Injury criteria derived from measurements at designated regions of the dummy body must be met and therefore the Hybrid lll must display satisfactory biofidelity . In order to investigate biofidelity, crash simulations were performed using human cadavers and Hybrid lll dummies. Frontal collisions were simulated at an impact velocity ' of 50 km/h and a mean car deceleration of between 14 and 16 g. The surrogates were protected with either a 3- point belt or an air bag - knee bolster system. Two belt systems of different stiffnesses were used, i.e 6% and 16% elongation. Assessment of Hybrid lll biofidelity was based on the kinematic behaviour and thoracic responses.

In the 3-point belt tests, head-neck flexion of the two test subject groups differed in magnitude and phase. Greater flexion was observed with the cadavers compared to the dummy and maximum Hybrid lll neck flexion occurred earlier. With both 3-point belts, 6% and 16% elongation, the Hybrid III experienced higher acceleration maxima and 3ms values at the sternum (in x- direction) and at Th1 (resultant). The resultant acceleration at Th12 was in agreement with the cadaver. X-direction rib accelerations were measured at the medium axillar line of the 8th rib (left and right) in cadavers and 5th rib (left and right) in dummies, and did not show clear differences between the two test subject groups.

Thoracic contours and deflections in the horizontal plane were measured with chest bands. Inter-subject and inter-test variability was observed for thoracic contours and deformations. Chest deflection was generally lower for the dummy than for the cadaver.

The results show that in frontal collision simulations the biofidelity of the dummy neck and thorax should be improved to allow for more valid assessment of occupant protection criteria.