Postural and Muscular Responses of Car Occupants under Pre-Crash Conditions

Abstract: Advanced integrated safety technologies in modern cars such as collision avoidance intervention and pre-crash activated restraint systems require comprehensive research on how vehicle occupants respond to these systems. The aim of this thesis is to provide insights into car passengers’ body kinematics and muscle activations in representative pre-crash circumstances with respect to two belt configurations (i.e., standard versus pre-pretensioner). Another objective is to explore the influence of occupants’ individual characteristics — namely age, stature, and sex — on their body kinematics. A complementary objective is to provide validation data for human body models (HBMs). A set of in-vehicle experiments was carried out in which front-row passengers were traveling at 73 km/h and subjected to autonomous lane changes and lane changes combined with braking, each with two belt configurations: standard and reversible pre-pretensioner belts. Volunteer muscle activations were measured by the surface electromyography (EMG) technique. Transformation of coordinates corresponding to several film targets attached to the head and upper torso was used to calculate the kinematics in 3-D. The volunteers’ EMG and kinematics were processed, and the quantified kinematics were statistically explored using principal component analysis and linear mixed model. Compared to the standard belt, pre-tensioning the seat belt prior to the maneuvers reduced lateral and forward displacements of the head and upper torso significantly. Seat belt pre-tensioning was also associated with earlier muscle activation onset and significantly lower activation amplitude for specific muscles. The influence of sex, stature, and their interaction on the head and upper torso kinematics were found statistically significant but accounted for a small amount of variance. A statistical model was developed which can predict head and upper torso kinematics of occupants with different stature and sex. The data provided in this thesis can be used for further enhancement and validation of HBMs. Consequently, the design of integrated safety systems in modern cars can benefit from more biofidelic models representing a wide range of population more accurately. Further statistical investigations for other types of omnidirectional loading scenarios and, preferably with a larger and more diverse sample space, are required to establish more accurate statistical models that can be generalized to the whole population.