Abstract:
This study presents the development and validation of an LS-DYNA finite element occupant model for crash analysis of roadside safety features. Traditional crash simulations primarily focus on vehicle kinematics and deformation, but incorporating an occupant model improves the link between vehicle accelerations and potential occupant injury. This approach allows for a more comprehensive evaluation of roadside hardware by assessing occupant kinematics and injury metrics.
The occupant model was developed using a combination of rigid bodies and deformable components, including flexible representations for the neck, spine, and abdomen to enhance biofidelity. The model is computationally efficient, making it suitable for long-duration roadside safety simulations. The validation process involved comparing simulated occupant responses against experimental calibration tests, including head-neck pendulum impact and rigid sled crash testing. These comparisons demonstrated strong correlation in acceleration time histories, head displacement, and thoracic injury criteria.
To optimize mesh interactions, a structured approach to contact modeling was implemented, ensuring stability in simulations involving vehicle interiors, restraints, and crash padding. The model was also parameterized for scalability, allowing for variations in occupant size (50th percentile male, 5th percentile female, and 95th percentile male). Applications of the model include evaluating restraint systems in roadside crashes, investigating injury potential in vehicle-to-barrier impacts, and enhancing the design of protective roadside safety features.
Future improvements will focus on refining joint articulation, expanding validation datasets, and integrating the model into full-scale crash scenarios involving complex oblique and side-impact conditions. The inclusion of this occupant model in LS-DYNA crash analyses will provide engineers and safety researchers with a more accurate tool for predicting and mitigating occupant injuries in vehicle collisions.
