The objective of the present work is to determine the potential injury-reducing benefits to car occupants in accidents of seat and seat-belt design. The main problems addressed are the risk of injury to head, neck, and upper torso in frontal impacts, and the risk of minor neck-injuries - whiplash-injuries - in rear-end collisions. The approach is to combine mathematical and mechanical modelling of representative frontal and rear-end crash-situations, thereby enabling general-purpose parametric studies of relevant design factors. Thus the crash situations are first modelled mechanically because this enables realistic occupant environments. The next step is to transfer the scenario into the mathematical model MADYMO, where the seat and seat-belt design are further elaborated on in order to resolve the detailed effects of changes to the occupant environment.
Since designs for increased restraining effect of the seat-belt may adversely affect the risk of occupant submarining, a means of assessing the tendency for submarining in MADYMO-simulations of frontal impacts is developed. This method can predict lap-belt-slip off the pelvis, which is the first step in the submarining event. The risk of submarining is then predicted in subsequent simulations. A means of reducing the belt-force while still maintaining the full protective capacity of the belt, is also studied. This principle allows for an increase in impact velocity of up to 25 % without increased risk for belt-induced injury.
The coupling between design parameters of a seat-belt system which has all anchorages incorporated in the seat structure, i. e. a belt-in-seat system, is analysed. The results show system response to be more sensitive to lap-belt geometry and seat properties than to shoulder-belt geometry. In fact, the small differences in occupant response obtained when the shoulderbelt geometry has been changed indicate most occupant sizes may be accommodated without elaborate devices for adjustment of the upper point of belt-presentation.
A number of parameters with possible effects on the risk of injury in rear-end collisions are also analysed. The results show the following 1) For a given impact velocity, the risk of injury increases when crash-pulse duration decreases. 2) Seatbelt usage only influences occupant response during rearward motion in severe impacts. 3) For strengths and stiffnesses below a threshold level, occupant protection decreases with decreasing stiffness and strength of the seat-back recliner. For strengths and stiffnesses above the threshold, occupant response is less sensitive to changes of these properties. 4) Recliner energy absorption is important for reducing occupant rebound, as well as for the protection of out-of-position occupants in severe impacts.
MADYMO proved effective for evaluating relations between car design parameters and occupant responses, although its algorithm for rigid body contacts was considered inadequate in applications involving analysis of contacts between the occupant and its environment.