【作者】 吴俊；

【导师】 曹立波；

【作者基本信息】 湖南大学 ， 车辆工程， 2013， 博士

【摘要】 道路交通事故是造成儿童伤亡的重要因素之一。每年有26万名儿童死于交通事故，1000万名儿童在交通事故中受伤。由于我国实行了计划生育政策，交通事故导致的儿童伤亡给家庭带来了巨大的精神伤害和物质损失，儿童乘车安全问题在我国显得尤为突出。随着汽车安全技术的发展，以及儿童乘车安全问题逐步受到重视，儿童安全已经成为汽车安全研究的重要课题。本文在文献研究的基础上，使用统计学方法、计算机仿真方法和优化方法，分别对后排儿童乘员的损伤流行病学和损伤防护、后排其他乘员（成人和婴幼儿）的损伤防护以及儿童假人生物逼真度改进等进行了研究。1.后排儿童乘员损伤流行病学研究。本研究基于NASS-CDS数据库，使用描述性统计和多元逻辑回归的方法，计算了4至18岁儿童/青少年乘员的约束系统类型和损伤风险，并以儿童年龄作为安全带保护效果变更的判定因子，评定了适合佩带成人安全带的年龄。结果显示，对于大于3岁的后排儿童乘员，成人安全带是最主要的约束形式。在4至18岁的儿童/青少年中，8至12岁的儿童的损伤风险最高。对于4至15岁的儿童/青少年，头部为最常见的损伤部位。小个儿童（4至7岁）很少发生面部损伤，但是8至18岁儿童/青少年的面部损伤风险很高。16至18岁青少年的胸部损伤风险最高（0.54%），但是其脊柱的损伤风险最低（0.03%）。对于8至18岁的儿童/青少年，腹部损伤风险随着儿童年龄的增长而降低。对于4至8岁的儿童，使用成人安全带会增加在正碰中的损伤几率（OR=2.51；95%CI:1.55-4.04）；对于9至18岁的儿童/青少年，使用成人安全带会降低在正碰中的损伤几率（OR=0.93；95%CI:0.81-1.07）。该研究为后续后排儿童乘员损伤防护研究的开展提供了现实依据。2.参数化儿童假人模型的建立和验证。本研究基于MADYMO软件提供的Hybrid III6岁儿童假人模型，引入了更真实的髋部和腹部几何、优化的铰链特性和接触特性，建立了可代表6至12岁儿童的参数化假人模型。该模型经过严格的试验验证，通过多目标优化方法与12种不同工况的台车试验进行了对标，如采用了不同座垫长度的后排座椅、不同的安全带固定点位置以及两种身体尺寸的儿童假人。使用统计评估的方法客观的评价了模型的有效性。结果显示，模型预测的假人头部、胸部、髋部加速度以及安全带力都与试验结果较好的吻合。该参数化假人模型可自动生成6至12岁之间任意年龄的儿童假人模型，为研究汽车乘坐环境对大龄儿童乘车安全性的影响，及其约束系统设计标准的建立，提供了有效的工具。3.后排儿童乘员约束系统参数分析和优化。本研究基于前文建立的参数化儿童假人模型，分析了身体尺寸、后排座椅参数和安全带参数对6至12岁儿童乘员损伤的影响，并通过优化方法得到该年龄段儿童的最佳约束系统设计。集合MADYMO、Modefrontier和Scilab软件，开发了一套自定义程序，可实现乘员定位、安全带匹配、约束系统参数设置和模型运算的全自动化。参数研究结果表明，身体尺寸（或年龄）对儿童在正面碰撞中的运动影响最为显著。为了平衡碰撞中的头部前移量和膝部前移量，并且防止出现“下潜”，本研究采用优化方法改进约束系统设计。总的来说，较低且靠后的上挂点位置，较高且靠前的锚点和带扣点位置，较短、薄且刚度大的座垫可以较好的保护6至12岁儿童。对于小个儿童乘员，上挂点、锚点和带扣点位置应靠前，以避免出现“下潜”。本研究得到的最优约束系统参数取值范围，可为大龄儿童乘员约束系统的设计提供指导。4.后排其他乘员（成人和后向式儿童座椅中的婴幼儿）的损伤防护研究。为分析前文得到的大龄儿童最佳约束系统配置对其他乘员损伤的影响，本研究建立了后排成人和婴幼儿后向式儿童座椅约束系统模型。模型与12个不同工况的台车试验进行对标，统计评估数据显示，建立的成人和婴幼儿后向式儿童座椅模型具有较高的仿真度。对两个模型中安全带和座椅参数进行优化，得到的优化参数与6岁儿童最佳约束系统配置进行对比，结果显示，三个年龄组乘员的最佳约束系统（座椅安全带固定点位置和座椅座垫长度）配置并不相同。比较靠前的锚点和带扣点位置可减少大龄儿童的“下潜”风险，却降低对成人和后向式儿童座椅中婴幼儿的保护；较短座垫长度可为成人和大龄儿童提供最佳的保护，却会显著增加后向式儿童座椅的旋转，增加婴幼儿的损伤风险。因此，有必要采用自适应式或智能约束系统，提高所有后排乘员的乘车安全性。5. Hybrid-III6岁儿童假人脊柱生物逼真度改进。由于Hybrid-III儿童假人继承了成人假人的刚性胸椎，因此不能准确模拟儿童在碰撞中的响应。本研究以经过改进的Hybrid-III6岁MADYMO假人模型作为基础模型，采用儿童志愿者试验、尸体试验和真实事故数据作为儿童假人改进设计的目标，探寻提高Hybrid-III6岁儿童假人脊柱生物逼真度的可行设计方案。使用优化技术，实现了模型与试验/事故数据的对标。假人的改进包括增加胸椎铰链和调整颈椎和腰椎的铰链特性。研究结果表明，为得到更真实的脊柱运动特性，应降低目前儿童假人的颈椎和腰椎在身高方向的移动刚度，并且增加包含曲伸和拉伸自由度的胸椎铰链。研究得到的脊柱铰链参数范围可为今后儿童假人的改进提供指导，并且建立的儿童假人模型可用于评价儿童约束系统的有效性。更多还原

【Abstract】 Motor vehicle crashes (MVCs) are one of the leading causes of injuries and deathsamong children. There are approximately260,000child fatalities and10,000,000injured in traffic accidents worldwide every year. Due to the one child policy in China,the safety of children in cars has become the spotlight in China. Along with thedevelopment of vehicle-safety-technology and the increasing attention on the issue ofchild safety in cars, child occupant safety has become an important research subject inthe filed of vehicle crash safety. Based on the literature review, this paper focused onthe study of injury epidemiology and injury prevention of child occupant seated in therear seat, injury prevention for the other occupants (adults and infants) seated in therear seat and improvement of spine biofidelity of child dummy through using statisticanalysis, computer simulation and optimization.1. Investigation of injury epidemiology of rear-seated child occupants. In thisstudy, NASS-CDS database was used to calculate the restraint types and injury risksfor rear-seated children/youth aged4-18and evaluate specific cutoff points for age aseffect modifiers of the association between the using of vehicle seat belts andsignificant injury, through applying the method of descriptive statistics andmultivariate logistic regression. The results showed that vehicle seatbelt was theprimary form of restraint for children older than3years-old. Among Children/youthaged4to18, children aged8to12had the highest injury risk. Head injury was themost common injury for children/youth aged4-15. Face injury rarely happened inyounger children (4-7years old), but had a high risk of injury for children/youth aged8-18. Chest injury risk was the highest for youth aged16to18(0.54%), but the risk ofspinal injury was the lowest (0.03%). Abdomen injury risk decreased when ageincreased, except for children aged4to7. Children in4to8seemed to be at aincreasing risk of significant injury when using seat belts (OR:2.51;95%CI:1.55-4.04). Among children/youth in9to18, there was a protective effect on significantinjury from using seat belt (OR:0.93;95%CI:0.81-1.07). This research can providea realistic basis for future studies on injury prevention for child occupants.2. Development and validation of the parametric child ATD model. In this study, aparametric ATD model capable of representing6–12YO children was developedbased on Hybrid-III6year old child ATD model provided by MADYMO. A more realistic representation of pelvis and abdomen geometry, modified joint stiffness andimproved contact characteristics were added to the baseline model. The newparametric ATD model was validated against results from12sled tests using realsecond-row vehicle seats with Hybrid III6YO and10YO ATDs under differentrestraint configurations, using a multi-objective optimization method. The modelvalidity was evaluated by statistical assessments of output measurements between thetests and simulations. Results showed that the model-predicted ATD head, chest andpelvis accelerations as well as the seatbelt forces were in good agreement with thosefrom the tests. This validated parametric child ATD model provides a useful tool toinvestigate the body size effects and to develop restraint system design guidelines for6–12YO child occupants.3. Parametric analysis and optimization of restraint system for rear seated childoccupant. In this study, a parametric analysis was conducted to investigate the effectsof body size, seat belt anchorage locations, and rear seat design parameters on theinjury risks in frontal crashes of children aged6to12by using the parametric childATD model developed above. An automated program was developed to integrate thedummy positioning procedure, belt fitting and varying seat and seat belt systems intocrash simulations by means of MADYMO, Modefrontier and Scilab software. Theresults of parametric analysis showed that child body size was the dominant factoraffecting outcome measures. In order to improve restraint system design, optimizationapproach was adopted to reduce head and knee excursions and prevent submarining.In general, lower and more rearward D-rings (upper belt anchorages), higher and moreforward lap belt anchorages, and shorter, stiffer, and thinner seat cushions wereassociated with improved restraint performance. In these simulations, children withsmaller body sizes require more-forward D-ring, inboard anchor, and outboard anchorlocations to avoid submarining. The range of optimum restraint system configurationobtained from this study can guide the future design of restraint system for older childoccupant.4. Investigation of injury prevention for other rear-seated occupants (adult andinfant in the rear-facing child restraint system（RF-CRS）). In order to investigate howthe optimum restraint system configuration for older children affects the injuries ofother occupants, rear-seated adult and infant in RF-CRS models were developed inthis study. A series of12sled tests were used to validate the computational models.The validity of two models were evaluated by statistical assessments, and the resultsshowed good agreement between tests and simulations. The optimum seat and seat belt configurations were obtained by optimization techniques and were compared withthose of the6year old child. Results showed that the optimal belt anchorage locationsand the seat cushion length for older children, adults, and infants in RF-CRS areconflicting to each other. In particular, more forward lap belt anchorage locationsthat prevent submarining for older children would reduce the protection to both adultsand CRS-seated infants. Shorter seat cushion could provide optimal protection toolder children and adults, but would significantly increase the CRS rotation. Thefindings of this study suggested that adaptive or adjustable restraint systems arenecessary to improve the rear seat occupant protection for different age groups.5. Improve spine biofidelity of Hybrid-III6year old ATD. Because Hybrid-IIIchild ATDs inherit a rigid thoracic spine from the adult HIII ATDs, so they can notcorrectly reflect child responses in MVCs. A previously developed and validated HIII6-year-old MADYMO ATD model was used as the baseline model to investigate thepossible design modifications on the spine biofidelity of current ATD. Several sets ofchild volunteer, cadaver test and motor vehicle crash data were considered as thedesign targets. Optimization techniques were used to match simulation results to eachset of test results. ATD design modifications include adding an additional joint to thethoracic spine region and changing the joint characteristics at the cervical and lumbarspine regions. The results indicated that, to achieve a realistic spine flexibility, thetranslational characteristics of the cervical and lumbar spine in the current child ATDneed to be reduced, and an additional joint at the thoracic spine region with degree offreedom in both flexion/extension and tension should be added. The child ATD modeldeveloped in this study can be used as an important tool to improve child ATDbiofidelity and child restraint system design in motor-vehicle crashes.更多还原