Occupant Neck Muscle Modelling in Rear-End Crashes

Doctoral thesis, 2022

The ultimate goal of the present research is to incorporate active and passive neck muscle effects in a female Finite Element (FE) Human Body Model (HBM). The application of interest is Whiplash Associated Disorders (WAD), which can occur in a low-speed rear-end impact. Two reflex mechanisms, the Vestibulocollic reflex (VCR) and the Cervicocollic reflex (CCR), are integral to maintaining head orientation. Therefore, active muscle modelling in HBMs should address the behaviour of these reflex mechanisms. Female FE HBMs are the focus of the present thesis because of their higher risk of sustaining WAD than males. This model should reproduce kinematics that can be used for global and local tissue injury prediction of WAD. The present thesis was arranged to address the main objective systematically and consists of six studies addressing five research questions. Two human body models representing the 50th percentile population, the VIVA OpenHBM and VIVA+ HBM, were used. The model was developed with and benchmarked against volunteer test data. Based on the collective studies in this thesis, the isolated head-neck model can be used to develop an active muscle controller. A simple, single-link approach was used to design a Proportional-Derivative (PD) controller called Angular Positioned Feedback (APF). This simple controller was convenient to implement and calibrate with available experimental data. Furthermore, reliable parameter identification, such as active muscle controller gains, were obtained via optimization using both head and cervical vertebral kinematics as objectives. A parameter study of different control strategies confirmed that the APF control strategy, combined with parallel damping elements (PDE), was the most effective for recreating volunteer kinematic responses compared to the model with only passive elements, particularly when impact severity was varied. Real-world collision data was used to evaluate the model’s usefulness using injury outcome data for known collision severities. The inclusion of neck muscle responses considerably influenced the cervical vertebral kinematics but only slightly influenced head kinematics before the rebound phase, depending on the head-to-headrest offset. Consequently, a slight difference in global kinematic-based injury criteria such as Neck Injury Criteria (NIC) was observed between a model with and without neck muscle responses. In contrast, significant differences between the two groups were observed for local, tissue-based, whiplash injury prediction. Hypotheses, such as Aldman pressure, require cervical spine kinematics and place higher requirements on the model’s performance. This analysis revealed the need for both global-based injury criteria and local, tissue level analysis to understand how WAD occur. Therefore, whiplash injury prediction would be more reliable using a model with the APF control strategy combined with PDE developed herein, than a model without active neck muscle responses. The FE HBMs with neck muscle responses have been developed and validated for low-speed rear-end impact and WAD analyses. The models have been shown to be robust and able to replicate volunteer head-neck kinematics.