Pre-Crash and In-Crash Car Occupant Safety Assessment
Licentiate thesis, 2021

Tens of millions are annually injured in Road Traffic Accidents (RTAs) worldwide, while the estimated number of RTA fatalities amounted to 1.35 million in 2016. In Europe, car occupants hold the largest share (48%) of fatalities among all road users. The high fatality and injury numbers motivate the work of enhancing road traffic safety. A holistic safety assessment approach, considering both the pre- and the in-crash phase of a crash, has the potential to enhance real-world occupant protection evaluation, as well as facilitate the development of effective countermeasures.

In standardized car occupant safety assessments, occupant surrogates of standardized anthropometries are employed in standardized postures, with the seat adjusted to a single predefined position. The vehicle is then subjected to predefined crash configurations with meticulously described impact points and angles. In contrast, real-world traffic crashes involve occupants of different shapes and sizes, who adjust the position of the seat and their posture on the seat differently, and the vehicles are subjected to diverse crash configurations (multiple impact locations, impact directions, and speed combinations). The overall aim of this thesis is to develop and apply methods, spanning from the pre-crash to the in-crash phase, capable of evaluating and enhancing the real-world occupant protection of future vehicles.

The introduction of crash-avoidance systems has the potential to alter the crash configurations that future vehicles will be exposed to. A method for predicting crash configurations has been developed in this thesis and applied to highway driving, and urban intersection crashes. Performing counterfactual simulations of digitized real-world crashes, with and without the addition of a conceptual Automatic Emergency Braking system, provides a prediction of the remaining crashes. The use of a novel crash configuration definition, along with a purpose-designed clustering method, facilitates the reduction of the number of predicted crash configurations without sacrificing coverage of the diverse real-world situations. Three predicted crash configurations, representative of urban intersection crashes, were further analyzed during the in-crash phase. A Human Body Model was positioned in a wide range of occupant postures identified from the literature. The findings suggest that the lower extremity postures had the largest overall influence on the lower extremities, pelvis, and whole-body responses for all crash configurations. In the evaluated side-impacts, leaning the torso in the coronal plane affected the torso and head kinematics by changing the interaction with the vehicle’s interior. Additionally, in far-side impacts supporting the occupant’s arm on the center console resulted in increased torso excursions. Moreover, the upper extremity responses were consistently sensitive to posture variations of all body regions.

Crash Configurations

Advanced Driver Assistance Systems (ADAS)

Human Body Model

Crashworthiness

Intersection crashes

Occupant postures

Real-world safety

Zoom, Password: 034178
Opponent: Adjungerad Professor Anders Kullgren, Folksam Research, Sweden

Author

Alexandros Leledakis

Chalmers, Mechanics and Maritime Sciences, Vehicle Safety

A method for predicting crash configurations using counterfactual simulations and real-world data

Accident Analysis and Prevention,; Vol. 150(2021)

Journal article

Leledakis, A., Östh, J., Davidsson, J., Jakobsson, L., 2021. The influence of car passengers’ postures in intersection crashes.

Future Occupant Safety for Crashes in Cars (OSCCAR)

European Commission (EC), 2018-06-01 -- 2021-05-31.

Open Access Virtual Testing Protocols for Enhanced Road User Safety (VIRTUAL)

European Commission (EC), 2018-06-01 -- 2022-05-31.

Areas of Advance

Transport

Subject Categories

Other Medical Sciences not elsewhere specified

Infrastructure Engineering

Vehicle Engineering

Publisher

Chalmers University of Technology

Zoom, Password: 034178

Online

Opponent: Adjungerad Professor Anders Kullgren, Folksam Research, Sweden

More information

Latest update

3/17/2021