Research And Application Of VRU Axonal Fiber Injury Dynamic Mechanism Based On Accident Reconstruction

The brain injury of vulnerable road user in traffic accidents is one of the main causes of their casualties.About 35% of people who die from brain injuries caused by traffic accidents have brain nerve injuries.Currently,most of the research based on finite element reconstruction of road traffic accidents only explores brain injury focus on the brain tissue level,and few studies have explored brain injury at the microscopic level.Therefore,it is of great significance to explore the brain injury from a microscopic level of human head during collision based on real road traffic accident data for developing personnel protective equipment,further promoting the development of automobile safety technology and improving the relevant laws and regulations of automobile safety.The objective of this study is to explore the correlation mechanism between brain nerve injury and head load at the microscopic level of the human head during collision based on traffic accident reconstruction with upgraded THUMS ver6.1 head-fiber coupling head finite element model.A convolutional neural network model for rapidly predicting the risk of human brain fiber injury is developed with an established criteria for brain fiber injury.Firstly,based on the existing methods of our team,the THUMS ver6.1 head-fiber coupling head finite element model was established.Based on the experimental data of head biomechanics in the existing literature research,using the fiber-coupled FE head model to develop eight experiments FE reconstructions.The correlation evaluation method was used to verify the biological fidelity of the fiber-coupled FE head model.Secondly,the head-fiber coupling head finite element model is used for accident reconstruction,and the relationship between intracranial dynamic response parameters and human head injury was explored.The evaluation criteria of strain distribution of brain fibers were constructed in the whole time domain of the collision process which is based on the dynamic response of brain fibers,and its prediction ability for severe brain injury in collision accidents was evaluated.The risk curve of fiber injury is established based on human injury information and the evaluation criteria of fiber injury.The relationship between fiber injury and head motion load was explored with the combination of the head load boundary condition,head motion response and fiber dynamic response during the collision.Finally,the head boundary conditions and the evaluation criteria for brain fiber injury during the collision process are applied to the development of brain fiber injury risk prediction models.The training and verification of the convolutional neural network model is completed,which is based on an evaluation criteria of fiber injury in the reconstruction of accident.In a conclusion,the THUMS ver6.1 head-fiber coupling head finite element model was an effective tool to investigating the injury of brain.Based on the reconstruction of accident cases,the evaluation criteria of brain fiber injury are proposed which can better predict the severe brain injury to the human head.The correlation mechanism between head impacts and fiber injury were analyzed: Compared to the peak linear acceleration of the head,the peak rotational angular acceleration is highly correlated with brain nerve injury.And the constructed convolution neural network model can better predict the risk of brain fiber injury.