Passenger kinematics in evasive maneuvers
Doctoral thesis, 2023
In situations that might lead to a vehicle crash, drivers often perform an evasive maneuver, such as braking or steering, in an attempt to avoid a crash. If a crash was not avoided, the maneuver could influence the injury outcome by altering the occupant’s position. Occupants use their muscles in response to a maneuver, and because the typical accelerations are low during maneuvers, the muscle activity can influence the kinematics. Thus, it is important to include the response to these potential maneuvers before the crash when predicting occupant injuries in a crash. The response to maneuvers could be evaluated by adding active musculature to existing evaluation tools, such as human body models. Furthermore, in volunteer studies, the head and torso displacements during maneuvers vary between occupants, but the cause for this variability remains to be identified. Two aims were defined for this thesis, addressed in two parts. The first aim was to advance the active neck and lumbar muscle controllers in the SAFER HBM to predict average response to maneuvers. The second aim was to further understand why such variability is seen in occupant response to evasive maneuvers.
Three muscle controller concepts were evaluated in this thesis, two of which were aimed at emulating the reflexes responding to input from the vestibular system that control the head position in space, and one controller that emulated reflexes that respond to lengthening of muscles. For the first aim, the active muscle controllers in the SAFER HBM were updated to allow for simulations with large vehicle yaw rotations, and the predictive capabilities were evaluated in braking, steering, and combinations. In a subsequent study, the updated controllers were tuned to volunteer kinematics in braking and steering, and the model performance was evaluated in the same conditions. It was concluded that the SAFER HBM, with the updated and tuned controllers, could predict passenger head kinematics in braking and steering with good to excellent results.
The occupant variability was addressed by statistical analysis of volunteer kinematics in six different vehicle maneuvers. In two subsequent studies, the Active Human Body Model developed within the first aim was used to analyze the model sensitivity to Human Body Model and boundary condition characteristics in braking. From the analysis of volunteer kinematics, it was concluded that the belt system was the most influential predictor for head and torso displacements across all maneuvers, while other characteristics such as sex, stature, age, and body mass index were less influential. In the subsequent studies, the seat forward/rearward position and spinal curvature were found to be most influential in braking.
Three muscle controller concepts were evaluated in this thesis, two of which were aimed at emulating the reflexes responding to input from the vestibular system that control the head position in space, and one controller that emulated reflexes that respond to lengthening of muscles. For the first aim, the active muscle controllers in the SAFER HBM were updated to allow for simulations with large vehicle yaw rotations, and the predictive capabilities were evaluated in braking, steering, and combinations. In a subsequent study, the updated controllers were tuned to volunteer kinematics in braking and steering, and the model performance was evaluated in the same conditions. It was concluded that the SAFER HBM, with the updated and tuned controllers, could predict passenger head kinematics in braking and steering with good to excellent results.
The occupant variability was addressed by statistical analysis of volunteer kinematics in six different vehicle maneuvers. In two subsequent studies, the Active Human Body Model developed within the first aim was used to analyze the model sensitivity to Human Body Model and boundary condition characteristics in braking. From the analysis of volunteer kinematics, it was concluded that the belt system was the most influential predictor for head and torso displacements across all maneuvers, while other characteristics such as sex, stature, age, and body mass index were less influential. In the subsequent studies, the seat forward/rearward position and spinal curvature were found to be most influential in braking.
volunteer kinematics
Active Human Body Model
pre-crash
active muscles
occupant variability
evasive maneuvers
Author
Emma Larsson
Chalmers, Mechanics and Maritime Sciences (M2), Vehicle Safety
Active Human Body Model Predictions Compared to Volunteer Response in Experiments with Braking, Lane Change, and Combined Manoeuvres
Conference proceedings International Research Council on the Biomechanics of Injury, IRCOBI,; (2019)p. 349-369
Paper in proceeding
Passenger Kinematics Variance in Different Vehicle Manoeuvres - Biomechanical Response Corridors Based on Principal Component Analysis
Conference proceedings International Research Council on the Biomechanics of Injury, IRCOBI,; Vol. 2022-September(2022)p. 793-843
Paper in proceeding
Larsson E., Iraeus J., Pipkorn B., Östh J., Forbes P. A, Davidsson J., Predicting occupant head displacements in evasive maneuvers; tuning and comparison of a rotational based and a translational based neck muscle controller,
Larsson E., Iraeus J., Davidsson J., Investigating sources for variability in volunteer kinematics in a braking maneuver, a sensitivity analysis with an active Human Body Model.
Larsson E., Iraeus J., Davidsson J., Synthetic experiments to investigate occupant variability in braking maneuvers, a simulation study using Active Human Body Models.,
Most drivers that notice an impending crash try to avoid it by performing an evasive maneuver, such as braking or steering. In response to the maneuver, the people inside the car move relative to the interior. In cases where the crash was not avoided despite the evasive maneuver, this movement relative to the vehicle could change the injury outcome of the crash, compared to if no maneuver had been performed. Because the maneuver could influence the injuries in the crash, it is important to consider maneuvers that occur before the crash to develop safer cars.
Simulation models of human bodies, vehicles, seat belts, and airbags are used typically to evaluate safety performance during vehicle development. Compared to physical testing, where relatively few tests can be performed, simulation models allow for thousands of virtual crash tests of the car and human during car development. These comparably inexpensive virtual tests allow for evaluation of more scenarios and braking or steering prior to the crash can be included.
In contrast to in crashes, humans can counteract the movement during maneuvers with muscle force. Thus, simulation models that evaluate the relative movement during maneuvers need to consider the muscle forces produced to counteract the movement. In this thesis, existing simulation models of car passengers that account for the muscle forces have been further developed to allow for more accurate modelling of movement during these maneuvers. The control systems that regulate muscle force production in the simulation model have been updated, and the parameters of the control system have been tuned such that the simulation model's movement during braking and steering simulations is as close as possible to how an average human moves during braking and steering.
Additionally, physical tests with human volunteers in a vehicle have been analyzed to understand how humans move during braking and steering. In that analysis, it was found that there is a large difference between how different people move during braking or steering. A statistical analysis was conducted to investigate if there were systematic differences between how males and females, shorter and taller people, lighter and heavier people, and younger and older people moved. It was found that only small systematic differences were present in the tests, and even after accounting for these differences, large variability remained. Therefore, the simulation models of humans were used to identify if other systematic differences could explain the large differences. With these simulations, it was found that some effects could be seen when letting the human sit more upright or slouched in the seat and when moving the seat backwards or forwards. The findings in this thesis suggest that future physical tests with human volunteers should investigate if sitting more upright or more slouched prior to a maneuver influence how people move during a maneuver.
Simulation models of human bodies, vehicles, seat belts, and airbags are used typically to evaluate safety performance during vehicle development. Compared to physical testing, where relatively few tests can be performed, simulation models allow for thousands of virtual crash tests of the car and human during car development. These comparably inexpensive virtual tests allow for evaluation of more scenarios and braking or steering prior to the crash can be included.
In contrast to in crashes, humans can counteract the movement during maneuvers with muscle force. Thus, simulation models that evaluate the relative movement during maneuvers need to consider the muscle forces produced to counteract the movement. In this thesis, existing simulation models of car passengers that account for the muscle forces have been further developed to allow for more accurate modelling of movement during these maneuvers. The control systems that regulate muscle force production in the simulation model have been updated, and the parameters of the control system have been tuned such that the simulation model's movement during braking and steering simulations is as close as possible to how an average human moves during braking and steering.
Additionally, physical tests with human volunteers in a vehicle have been analyzed to understand how humans move during braking and steering. In that analysis, it was found that there is a large difference between how different people move during braking or steering. A statistical analysis was conducted to investigate if there were systematic differences between how males and females, shorter and taller people, lighter and heavier people, and younger and older people moved. It was found that only small systematic differences were present in the tests, and even after accounting for these differences, large variability remained. Therefore, the simulation models of humans were used to identify if other systematic differences could explain the large differences. With these simulations, it was found that some effects could be seen when letting the human sit more upright or slouched in the seat and when moving the seat backwards or forwards. The findings in this thesis suggest that future physical tests with human volunteers should investigate if sitting more upright or more slouched prior to a maneuver influence how people move during a maneuver.
Active human body model for virtual occupant response, step 5
VINNOVA (2020-05155), 2021-04-01 -- 2023-03-31.
Active human body models for virtual occupant response (A-HBM), step 4
VINNOVA (2017-05516), 2018-04-01 -- 2021-03-31.
Areas of Advance
Transport
Subject Categories
Transport Systems and Logistics
Vehicle Engineering
Infrastructure
C3SE (Chalmers Centre for Computational Science and Engineering)
ISBN
978-91-7905-889-0
Doktorsavhandlingar vid Chalmers tekniska högskola. Ny serie: 5355
Publisher
Chalmers