| There are about 1,240,000 people died in the road traffic accidents each year and 20 million to 50 million people suffered nonfatal injuries according to WHO. Road traffic injuries will cause great economic damage to the victims, their families, and the entire country, and it is expected that the road traffic injury will become the world’s third largest injury burden by 2020. Therefore, the automobile manufacturers and research institutions put a lot of effort into the car crash safety research. But the current occupant injury assessment tools involved in the computer technology, such as the crash dummies and human FE models, generally focus on the protection of the 50 th percentile male occupant, and do not take into account the diversity of the occupant body shape. In the vehicle crash, compared to the 50 th percentile male, the elderly, obese people, children, as well as short women are vulnerable people. Field data analysis shows that the changes of human characteristics, including age, sex, height, as well as BMI(Body Mass Index) has an important effect on occupant injury. For example, with the increase of age, the risk of MAIS 3+ for all body parts of the occupant will increase, especially for the chest, head, lower extremities and spine. In the front vehicle collision accident, obese occupants had 26% and 33% higher risks of AIS 2+ and AIS 3+ thoracic injuries than the normal-weight occupant, and while maintaining collision conditions, women had 47% and 71% higher risks of MAIS 3+ and MAIS 2+ than men. If taking into account the interaction between different parameters, such as the obese elderly passenger, obese children and elderly women, and the elderly female passengers, the injury will be more serious. However, at present, all of the occupant injury assessment tools cannot be carried out on the above types of occupants. Therefore, a method was developed in this study to build a parametric FE(finite element) human model based on the statistical analysis technique and the mesh morphing technique, and the injuries were studied for the obese occupants in the rear seat and the occupants with different gender, age and height in the front seat according to the established parametric human body model. The focus and innovation of this research are as follows:1. A parametric FE model of human skeleton was established accounting for the age, height, gender and BMI based on the shape statistical analysis. Using the CT scans of different subjects, the parametric FE models of ribcage, pelvis, femur and tibia were built through statistical shape analysis techniques, including the image threshold segmentation, landmark matching selection, generalized Procrustes analysis, principal component analysis and linear regression analysis. These parametric models could predict the subject-specific skeletons FE model according to any reasonable parameter combination of age, sex, height and BMI.2. A parametric FE model of human body surface was established accounting for the age, height, gender and BMI based on the shape statistical analysis. The baseline body surface mesh was morphed into the target geometry presented with the point cloud through the mesh morphing technique. And then according to the human body characteristics, the morphed body surface mesh was adjusted into a symmetrical model. Statistical analysis results showed that: with the increase of BMI and height, all the occupant body dimensions increased, for example the obese occupant tended to have a large chest and waist size and the high occupant had a larger sitting height. The age had no effect on the occupant chest width and hip width, but the elderly waist size increased and kyphosis occurred for the elderly.3. A method was developed to build a parametric human body model based on the statistical analysis technique and the mesh morphing technique. Using the established parametric skeleton FE model(chest, pelvis, femur and tibia) and parametric human body surface FE model, the target human geometric model was obtained by putting the skeleton into the human body surface with least squares method. The landmarks for positioning the skeleton were the joint locations predicted by the human body surface. Then the separated mesh morphing method was applied to map the baseline human body FE model into the target geometry. This method could quickly generate a human body model based on the occupant characteristics in 30 minutes, and at the same time the model maintained good mesh quality. By contrasting the skeletal shape differences, occupant dynamic response and occupant injury between the model and cadaver test, it was found that the occupant FE model generated through our parametric human body model could simulate the cadaver response more accurately than the baseline model in the collision. Moreover, many human body models based on the individual characteristics could be rapidly generated by this method, which can be applied into the parametric optimization for the vehicle restraint system.4. The injuries of occupants with different BMI levels in the rear seat were studied and the method was put forward to protect the obese occupants. Four human body models with different BMI levels(25/30/35/40kg/m2) were generated according to the established parametric human body model. A parametric study was performed through a validated rear seat model to analyze the effect of BMI, force levels of load limiter, anchor pretensioner and belt position on the occupant injury. The simulation results showed that the injury of the head or lower extremities of the rear seat occupant was inversely proportional to the chest injury, and load limiter and anchor pretensioner could decrease the occupant injury caused by obesity.5. The injuries of occupants with different gender, age and height in the front seat were studied. Four FE occupant models with different gender, age and height were generated according to the established parametric human body model, then the corresponding material properties were selected respectively for the four FE occupant models from the literatures. Furthermore, different occupant FE models were positioned into the front seat constraint system and placed with the belt through the statistical model of human body posture and statistical model of belt position. Four simulations were performed to study the effect of age, sex and height on the occupant injury. The results of the study showed that: with the increase of the age, the fracture risk increased significantly for the human ribcage and lower extremities, and the injury for the female is higher than the male, and the age had a larger effect on the female injury than the male.The developed parametric human body model has very important significance for the safety research of occupants with different characteristics. Through this method a large number of human body models based on different human characteristics can be quickly generated, which broadens the protected population as far as possible in the vehicle collision. At the same time the parameter optimization design for the vehicle restraint system can be carried out according to the parametric human body model, which is impossible for the current injury assessment tools. |
PDF Full Text Request