中国科学院力学研究所机构知识库

铁路运输，是缓解交通压力的有效方式之一。随着列车运行速度的不断提升，气动效应对列车运行安全性产生的影响越来越突出。同时，与无风情况相比，横风作用会使高速列车的气动性能显著恶化，从而带来更大的安全隐患。因此，对横风气动性能进行全面分析并加以优化，是高速列车外形设计过程中必不可少的重要环节。目前，针对高速列车流线头型和局部构件的传统气动性能优化方法几乎没有进步空间，而利用微结构进行非光滑表面设计的新型技术手段正受到国内外学者的广泛关注。虽然表面微结构在高速列车气动减阻、降噪方面的应用已取得不少成果，但还缺乏对微结构表面流动控制机理的研究，且有必要深入剖析微结构表面对列车横向稳定性的影响规律。基于此，本文主要开展了以下工作：

（1）以在车顶加设矩形条带组的方式，对390级Pendolino列车简化模型的头车进行了局部非光滑表面设计；采用数值模拟方法获得了光滑表面列车模型和粗糙表面列车模型的表面压力分布规律，并与风洞试验结果进行对比，验证了所用方法的有效性；结合不同表面列车模型周围的流场结构，揭示了矩形沟槽表面对车体周围气体流动的控制机理；根据关键气动载荷的计算结果，表明此种非光滑表面设计方法确实利于提升列车的横风稳定性。

（2）利用正交试验设计和统计分析相结合的方法，探索了矩形条带四个几何参数与列车两个关键气动载荷间的关系：首先，根据方差分析结果，确定了四个几何参数的主次顺序；接着，假设条带几何参数与关键气动载荷呈线性相关，利用多元线性回归分析对二者的相关方向进行了判断；最后，在保证次要参数不变的情况下，通过极差分析结果得到了条带外形设计的优选方案，并对其作用效果进行了评估。

（3）在灵敏度分析所得结论的基础上，以矩形条带的两个主要几何参数为设计变量、列车的侧向力系数和倾覆力矩系数为优化目标，采用基于Kriging替代模型的多目标遗传算法NSGA Ⅱ，开展了面向高速列车横风气动性能的非光滑表面优化设计。通过全局寻优确定Pareto最优解后，先评估了相应设计方案的优化效果，又结合数值计算结果进一步验证了其有效性，并根据流场特性揭示了列车横风稳定性得以提升的根本原因。

Railway transportation is one of the effective ways to relieve traffic pressure. With the continuous improvement of trains’ speed, the influence of the aerodynamics on the safety of trains becomes more and more prominent. At the same time, compared with no-wind conditions, the effect of cross-wind will significantly deteriorate the aerodynamic performance of high-speed trains, and bringing a potential threat to safety. Therefore, a comprehensive analysis and optimization of cross-wind aerodynamic performance is an indispensable and important link in the process of high-speed train shape design. At present, there is little room for improvement in traditional aerodynamic performance optimization methods for streamlined heads and partial components of high-speed trains. However, a new technical method using microstructures for non-smooth surface design are receiving extensive attention from scholars all over the world. Although the application of surface microstructures in the aerodynamic drag reduction and noise reduction of high-speed trains has achieved some results, there is still a lack of research on the control mechanism of microstructure surface, and it is necessary to deeply analyze the influence of microstructure surfaces on the lateral stability of trains. Based on this, the article mainly carried out the following work:

(1) By adding rectangular strips on the roof, the leading car of the simplified model of the 390 Pendolino train is designed with a local non-smooth surface. The numerical simulation method is used to obtain the surface pressure of the smooth surface train model and the rough surface train model, and the comparison with the wind tunnel experimental results verifies the effectiveness of the method used. Combined with the flow field around the train model with different surfaces, the control mechanism of the rectangular groove surface on the airflow around the train is revealed. According to the numerical results of key aerodynamic loads, it is shown that this non-smooth surface design method does help to improve the cross-wind stability of the train.

(2) Using orthogonal experimental design and statistical analysis, the relationships between four geometric parameters of t rectangular strips and the two key aerodynamic loads of the train are explored. First, according to the results of the analysis of variance, the order of the four geometric parameters is determined. Next, assuming that the geometric parameters of the strips are linearly related to the key aerodynamic loads, the correlation directions of them are judged using multiple linear regression analysis. Finally, under the condition of ensuring that the secondary parameters remain unchanged, the better design of strips is obtained through the results of the range analysis, and its effect is evaluated.

(3) Based on the conclusions of the sensitivity analysis, the two main geometric parameters of the rectangular strips are selected as the design variables, and the side force coefficient and the roll moment coefficient of the train are the optimization objectives. The multi-objective genetic algorithm (NSGA Ⅱ) based on the Kriging surrogate model is adopted. The non-smooth surface optimization design for the cross-wind aerodynamic performance of the high-speed train has been carried out. After determining the pareto optimal solution through global optimization, the effect of the corresponding design scheme is evaluated, and its effectiveness is further verified by the numerical results. Finally, the fundamental reason for the improved crosswind stability of the train is revealed based on the characteristics of the flow field.