汽车在为人们带来便利的同时，也带来了交通事故和人员伤亡，在正面碰撞事故中，下肢是最容易受到伤害的部位。有限元仿真分析被广泛用于乘员保护研究，现有人体有限元模型没有考虑性别、年龄、身高和身体质量系数等人体特征的多样性，且下肢模型几何尺寸来自美国人体，而中美人体下肢差异尚未得到充分研究。本文提出了参数化下肢建模流程和方法，旨在基于中国人体CT数据建立反映中国人体特征的参数化下肢有限元模型，研究人体特征参数对下肢几何形态和碰撞损伤响应的影响，并同步建立美国人体下肢模型，探究中美人体下肢几何形态和碰撞损伤响应差异。本文提出了适用于参数化下肢建模的表面投影方法和基于CT扫描的下肢密质骨厚度估算方法，并开展了尸体试验标定并验证了估算方法中的关键参数。基于中美人体下肢CT扫描数据，本文通过标志点识别、网格变换与表面投影、密质骨厚度估计和主成分与回归分析等主要步骤，建立了以人体特征参数为自变量的包含骨骼形状尺寸和密质骨厚度分布的下肢统计学模型，并研究了人体特征参数对上述骨骼几何形态的影响。结果表明，人体特征参数对下肢骨骼形状尺寸具有显著影响，但影响方式在中美人体之间存在差异，人体特征参数对密质骨厚度分布也具有显著影响，且中美人体下肢骨骼形状尺寸和密质骨厚度分布存在显著差异。基于统计学模型，本文提出了下肢有限元模型密质骨厚度调整方法和单元质量自动检查与修正方法，建立了参数化下肢有限元模型。依据尸体试验数据，验证了股骨模型的生物仿真度和建模方法有效性。结合其他人体部位模型，本文进一步建立了完整的参数化人体模型，并依据尸体试验数据验证了膝盖-大腿-臀部模型的生物仿真度，同时生成了一系列具有不同人体特征参数的中美乘员有限元模型。最后，本文建立了面向多体征人体有限元模型的碰撞仿真分析流程，并开展了正面偏置碰撞的仿真分析，重点研究了乘员下肢损伤。结果表明，老年、肥胖和女性乘员的下肢损伤风险更高，相同碰撞工况下，具有相同人体特征参数的中美乘员下肢碰撞载荷没有明显差异，但中国乘员股骨的高应力区域大于美国乘员股骨，当考虑人体特征参数分布在中美人群中的差异时，中国人群的下肢损伤风险比美国人群低12%。本文建立的参数化模型可为优化乘员约束系统、研究不同碰撞工况下乘员损伤响应和易损伤乘员保护提供参考和帮助，并为提出适合中国人体特征的损伤指标提供理论和实践基础。

Injuries and casualties caused by traffic accidents and the convenience provided by vehicles are just two sides of the same coin. In frontal motor vehicle crashes, lower-extremity is still the most injured body part. Finite element analyses have been widely used in the study on occupant protection, but current human body finite element models do not account for the diversity of sex, age, stature and BMI among people. In addition, the geometry target of the lower-extremity model is from American people, and the differences between lower-extremities of Chinese and American population have not been fully studied. In this study, a systematic procedure of developing parametric lower-extremity models was proposed to build a parametric lower-extremity finite element model representing Chinese population using CT scans and study the effects of subject characteristics on the geoemtry and injury responeses of the lower-extremity during impacts. Another lower-extremity model representing American population was built at the same time to explore possible differences between lower-extremities of Chinese and American population in terms of geometry and injury responses during imapcts.In this study, a surface projection method adapted for parametric lower-extremity modeling and a method to estimate cortical bone thickness for lower-extremity using clinical CT scans were proposed, and a cadaver test was performed to calibrate and validate the key parameter in the proposed method. Using lower-extremity CT scans of Chinese and American subjects, through a modeling process including landmark identification, mesh morphing and surface projection, cortical bone thickness estimation, principal component analysis and multivariable regression, a statistical lower-extremity model was built to predict bone geometry and cortical bone thickness distribution using subject characteristics as inputs. Studies were performed to address the effects of subject characteristics on the aforementioned bone geometry. Results indicated that subject characteristis had significant effects on bone size and shape, and that such effects were different between Chinese and American populations. Subject characteristics also showed significant effects on the distribution of cortical bone thickness. It was also revealed that there was a significant difference between Chinese and American population in terms of the size, shape and cortical bone thickness distribution of the bone in the lower-extremity.Based on the statistical lower-extremity model, a thickness adjustment method and an automatic mesh quality checking and fixing technique were proposed to build parametric lower-extremity finite element models, which were validated against femur cadaver tests in literature. By integrating the lower-extremity models developed in this study with other models of different body parts, a set of whole human body finite element models of various subject characteristics representing Chinese and American population were built. In addition, the knee-thigh-hip part was validated against cadaver tests in literature. In the end, a crash simulation procedure targeting parametric human body models was established. Frontal oblique crash simulations were performed using the human body models representing various subject characteristics and the occupant lower-extremity injuries were carefully studied. The results indicated that aged, obese and female occupants suffered a higher lower-extremity injury risk. Under the same impact condition, there was no significant difference between the impact loads sustained by Chinese and American lower-extremities, but Chinese femurs had a larger area of high stress values. Taking the variation of the distribution of subject characteristics in both populations into consideration, the lower-extremity injury risk of Chinese population was 12% lower than that of American population. The parametric models developed in this study can offer meaningful reference and help in the study of occupant injury responses under various impact conditions and the protection of vulnerable occupants, and the optimization of occupant restraint systems. The parametric lower-extremity model can also lay a foundation for establishing new injury criteria suitable for Chinese population.