Development, Calibration, and Validation of a Head–Neck Complex of THOR Mod Kit Finite Element Model

<b>Introduction/Objective:</b> In an effort to continually improve upon the design of the test device for human occupant restraint (THOR) dummy, a series of modifications have recently been applied. The first objective of this study was to update the THOR head–neck finite element (FE) model to the specifications of the latest dummy modifications. The second objective was to develop and apply a new optimization-based methodology to calibrate the FE head–neck model based on experimental test data. The calibrated head–neck model was validated against both frontal and lateral impact test data. Finally, the sensitivities of the model, in terms of head and neck injury criteria, to pretest positioning conditions were evaluated in a frontal crash test simulation.<b>Methods:</b> The updated parts of the head–neck THOR FE model were remeshed from CAD geometries of the modified parts. In addition, further model modifications were made to improve the effectiveness of the model (e.g., model stability). A novel calibration methodology, which incorporates the CORA (CORelation and Analysis) rating system with an optimization algorithm implemented in Isight software, was developed to improve both kinematic and kinetic responses of the model in various THOR dummy certification and biomechanical response tests. A parametric study was performed to evaluate head and neck injury criteria values in the calibrated head–neck model during a 40 km/h frontal crash test with respect to variation in the THOR model upper body and belt pretest position.<b>Results:</b> Material parameter optimization was shown to greatly improve the updated model response by increasing the average rating score from 0.794 ± 0.073 to 0.964 ± 0.019. The calibrated neck showed the biggest improvement in the pendulum flexion simulation from 0.681 in the original model up to 0.96 in the calibrated model. The fully calibrated model proved to be effective at predicting dummy response in frontal and lateral loading conditions during the validation phase (0.942 average score). Upper body position was shown to have a greater effect on head–neck response than belt position. The pretest positioning variation resulted in a 10 percent maximum change in HIC₃₆ values and 14 percent maximum change in <i>N</i>_IJ values.<b>Conclusion:</b> The optimization-based calibration methodology was effective as it markedly improved model performance. The calibrated head–neck model demonstrated application in a crash safety analysis, showing slight head–neck injury sensitivity to pretest positioning in a frontal crash impact scenario.

<b>引言/研究目标：</b> 为持续优化人体乘员约束测试假人（Test Device for Human Occupant Restraint，简称THOR）的设计，近期已针对该假人开展一系列改进工作。本研究的首要目标是将THOR头颈有限元（Finite Element，简称FE）模型更新至最新假人改进后的规格参数；第二项目标是开发并应用一种基于优化的新型方法，基于实验测试数据对头颈有限元模型进行校准。校准后的头颈模型已通过正面与侧面碰撞测试数据完成验证。最后，在正面碰撞仿真试验中，评估了该模型针对头颈损伤指标对预试验定位条件的敏感性。<b>研究方法：</b> 更新后的THOR头颈FE模型部件，基于改进部件的计算机辅助设计（Computer Aided Design，简称CAD）几何模型重新划分网格。此外，为提升模型有效性（如模型稳定性），还对模型进行了进一步优化调整。本研究开发了一种新型校准方法，该方法将CORA（CORrelation and Analysis，相关性与分析）评分系统与Isight软件中集成的优化算法相结合，以改善THOR假人各类认证试验及生物力学响应试验中模型的运动学与动力学响应性能。同时开展参数化研究，以评估在40km/h正面碰撞试验中，校准后头颈模型的头颈损伤指标随THOR模型上半身及安全带预试验位置变化的情况。<b>研究结果：</b> 材料参数优化可显著提升更新后模型的响应性能，使平均评分从0.794±0.073提升至0.964±0.019。校准后的颈部模型在摆锤屈曲仿真中的表现提升最为显著，评分从原始模型的0.681提升至校准后模型的0.96。经全面校准后的模型在验证阶段可有效预测正面及侧面载荷条件下的假人响应，平均评分为0.942。研究表明，上半身位置对头颈响应的影响程度大于安全带位置。预试验定位偏差可导致HIC₃₆值最大变化幅度达10%，N_IJ值最大变化幅度达14%。<b>研究结论：</b> 基于优化的校准方法效果显著，可大幅提升模型性能。经校准后的头颈模型可应用于碰撞安全分析，结果表明在正面碰撞场景中，头颈损伤指标对预试验定位条件仅存在轻微敏感性。