白车身前端结构—材料—性能一体化轻量化多目标协同优化设计

发布时间：2018-06-06 02:42

  本文选题：轿车 + 参数化　； 参考：《吉林大学》2016年博士论文

【摘要】：随着汽车保有量的快速增长,能源过度消耗、环境污染等一系列社会问题随之出现。汽车轻量化是减少能源消耗和污染物排放的重要途径。白车身质量占汽车总质量的30%~40%,制造成本约占整车成本的60%,空载情况下大约70%的燃油被白车身消耗。因此白车身轻量化是汽车轻量化重要的组成部分。正面碰撞无论是发生率还是人员受到伤害和死亡率都较高,正碰被动安全性能是汽车最重要的性能之一。白车身前端结构质量大约为白车身整体质量的30%,吸能量大约为白车身总体吸能量的80%,白车身前端质量对正碰安全性具有非常重要的作用。因此专门针对白车身前端结构的轻量化优化设计显得更加重要。白车身前端结构轻量化设计是一项多参数、多约束系统的复杂工程,涉及到动、静态刚度、NVH、耐撞性和制造成本等多项性能指标。本文基于现有的白车身有限元模型,利用SFE-CONCEPT软件建立了隐式参数化的白车身前端结构模型,并与白车身后端有限元模型耦合在一起。以白车身耦合为基准,在保证白车身静态弯扭刚度、一阶扭转、弯曲模态频率、正面100%碰撞安全性不明显降低和制造成本不显著增加的前提下,综合考虑部件的材料、厚度、断面形状、部件曲率等设计因素对白车身前端进行结构-材料-性能一体化轻量化多目标优化设计。本文主要开展了以下几个方面的研究内容并得出了相关结论:(1)仿真分析白车身有限元模型的静态弯扭刚度、低阶模态,通过与试验结果对比验证白车身有限元模型在静态弯扭刚度、低阶模态性能方面满足建立参数化白车身前端的要求。再将白车身有限元模型与底盘、发动机等有限元模型连接在一起,并进行相关设置,按照新车评价程序(C-NCAP)进行100%正碰、侧碰安全性仿真分析,通过与试验结果对比验证白车身有限元模型在正碰安全性方面满足建立参数化白车身前端的要求。(2)以白车身有限元模型为基准,用隐式参数化建模方法创建白车身前端结构参数化模型,并将其与白车身后端有限元模型连接在一起构成白车身耦合模型。在耦合模型的基础上进行静态弯扭刚度、低阶模态和正碰、侧碰安全性仿真分析。随后将相关性能与有限元模型的性能进行对比从而验证耦合模型的有效性。(3)采用近似模型优化方法对白车身耦合模型的参数化前端结构进行一体化轻量化多目标优化设计,为了减少进行试验优化设计(DOE)过程中大量的重复工作,本文采用模块化方法。根据模块化分类原则将参数化白车身前端、白车身后端有限元模型分别设置为单独的子模块。试验设计中对白车身静态弯扭刚度、低阶模态和正碰安全性仿真分析时,通过改变参数化白车身前端结构,再结合各自的子模块文件,可方便的运行试验设计样本点。(4)在对白车身耦合模型的参数化前端进行一体化轻量化多目标优化设计时,合理的选取设计变量可以减少计算工作量,提高优化效率。因此本文分析常用筛选变量方法之间的联系与区别,在此基础上提出综合灵敏度分析方法,并证实该方法具有较高的优化效率。利用该方法从23个初始设计变量中筛选出14个设计变量。以筛选出的14个设计变量为基准构建二阶响应面(Quadratic Response Surface Methodology,QRSM)、克里格(Kriging)和径向基神经网络(Radial Basis Functions Neural Network,RBF)三种常用的近似模型,在第二代非劣排序遗传算法(NSGA-Ⅱ)的基础上计算出各性能响应,并比较响应与仿真响应之间的相对误差,结果表明Kriging近似模型得到的响应误差最小。(5)为进一步提高耦合模型的参数化前端减重效果,将前防撞梁、吸能盒的钢质材料替换为铝合金。材料替换后对前防撞梁、吸能盒进行拓扑优化并得到相应的主断面结构。随后在SFE-CONCEPT软件中建立前防撞梁、吸能盒的参数化模型,并与参数化白车身前端其他部件连接在一起,构成多种材料的参数化白车身前端模型,铝合金吸能盒与钢质前纵梁之间通过带螺栓的法兰盘实现连接。为得到前防撞梁、吸能盒和法兰盘的尺寸、厚度、材料参数,本文在筛选出的14个设计变量基础上增加表征铝合金前防撞梁等部件尺寸、厚度、材料的12个设计变量。同时本文还提出计算车身部件材料成本的方法,考虑车身前端在一体化轻量化多目标优化设计时带来的材料成本变化。(6)选用Kriging近似模型优化方法对耦合模型的参数化前端进行一体化轻量化多目标优化设计,优化中以前端质量最小、白车身静态扭转刚度最大和材料成本最低为目标,同时约束白车身静态弯曲刚度,一阶扭转、弯曲模态频率和正面100%碰撞安全性能不明显降低,综合考虑部件厚度、断面形状、曲率、不同材料的26个设计变量,利用NSGA-Ⅱ算法在设计空间搜索优化妥协解集。以质量最小的妥协解为参数化前端模型一体化轻量化优化的设计方案,对该方案仿真分析后发现,白车身前端质量减少5.81Kg,减重率达7.01%,且白车身静态弯扭刚度、低阶模态和正碰安全性能没有明显变化,材料成本仅增加3.30%。另外经仿真分析表明优化后的白车身前端对侧碰安全性能也几乎没有影响。
[Abstract]:With the rapid growth of car ownership, a series of social problems such as excessive energy consumption and environmental pollution, automobile lightweight is an important way to reduce energy consumption and pollutant emission. The quality of body white accounts for 30%~40% of the total vehicle quality, and the manufacturing cost accounts for about 60% of the total vehicle cost, and about 70% of the fuel is white in the case of empty load. Light weight is an important part of the vehicle lightweight. The front impact is one of the most important performance of the car, whether it is the occurrence rate, the injury and the death rate, and the positive impact is one of the most important performance of the car. The quality of the front end of the body in white is about 30% of the overall quality of the body in white, and the energy absorption is about the white body. With 80% of the total energy absorption, the front end quality of the body in white has a very important effect on the positive collision safety. Therefore, the lightweight optimization design of the front end structure of the body white is more important. The lightweight design of the front end structure of body white is a complex engineering of a multi parameter, multi constrained system, involving the dynamic, static stiffness, NVH, crashworthiness and Based on the existing finite element model of the body in white, this paper uses the SFE-CONCEPT software to establish an implicit parameterized front end structure model of the body in white and coupled with the finite element model of the back end of the body in white. The static bending stiffness, first torsion, bending die of the body in white body are guaranteed by the coupling of the body in white. On the condition of state frequency, the safety of frontal 100% collisions is not obviously reduced and the cost of manufacturing is not significantly increased, the design factors such as material, thickness, section shape, and component curvature are considered to integrate structure material performance integrated lightweight multi-objective optimization design on the front end of the body in white body. The following aspects are mainly carried out in this paper. The main contents are as follows: (1) simulation and analysis of the static bending and torsion stiffness of the finite element model of the body of white body and the low order mode. Through comparison with the experimental results, it is proved that the finite element model of the body in white can meet the requirements for the establishment of the parameterized front of the body in white, and then the finite element model and the chassis of the body in white. The finite element model of the engine is connected together, and the related setting is carried out. The 100% positive collision and the side collision safety simulation analysis are carried out according to the new car evaluation program (C-NCAP). By comparing with the test results, it is proved that the finite element model of the body in white meets the requirements of the establishment of the parameterized front end of the body in white. (2) the finite element model of the body in white. Based on the implicit parameterized modeling method, the parameterized model of the front end structure of the body in white is created. The model is connected with the finite element model of the back end of the body of the white body to form the coupling model of the body in white. The static bending stiffness, the low order mode and the positive collision and the side collision safety simulation analysis are carried out on the basis of the coupling model. The related performance and the related performance are then carried out. The performance of the finite element model is compared to verify the effectiveness of the coupling model. (3) an approximation model optimization method is used to integrate the parameterized front structure of the car body coupling model to the multi-objective optimization design. In order to reduce the repeated work in the process of experimental optimization design (DOE), the modular method is adopted in this paper. According to the modular classification principle, the front end of the white body is parameterized and the back end finite element model of the body in white is set to separate sub modules respectively. In the experiment design, the static bending and torsion stiffness, the low order mode and the positive collision safety of the body in white body are analyzed. By changing the parameters of the front structure of the body in white, it is convenient to combine the respective sub module files. (4) when integrating the parameterized front-end of the white body coupling model to the integrated lightweight multi-objective optimization design, the reasonable selection of the design variables can reduce the calculation work and improve the optimization efficiency. Therefore, this paper analyzes the relation and difference between the commonly used selection methods and proposes the comprehensive spirit on this basis. The method of sensitivity analysis proves that the method has high optimization efficiency. Using this method, 14 design variables are selected from 23 initial design variables. The two order response surface (Quadratic Response Surface Methodology, QRSM), Craig (Kriging) and radial basis neural network (Radial Basis) are constructed with selected 14 design variables. Functions Neural Network, RBF) three common approximate models, calculated the performance response on the basis of the second generation non inferior sorting genetic algorithm (NSGA- II), and compared the relative error between the response and the simulation response. The results show that the response error of the Kriging approximation model is the least. (5) to further improve the parameterization of the coupling model. The front end weight reduction effect is made by replacing the steel material of the front collision beam and the energy absorption box with the aluminum alloy. After the material is replaced, the front anti-collision beam, the energy absorption box is topologically optimized and the corresponding main section structure is obtained. Then, the pre collision beam, the energy absorption box's parameterized model is established in the SFE-CONCEPT software and connected with the other parts of the parameterized front end of the body of the body of the white body. Together, the parameterized front end model of the body in white is made up of a variety of materials. The aluminum alloy energy absorption box and the steel front longitudinal beam are connected through the flange plate with a bolt. In order to obtain the size, thickness and material parameters of the front collision avoidance beam, the energy absorption box and the flange plate, this paper increases the anticollision beam of the aluminum alloy on the basis of the selected 14 set variables. 12 design variables of component size, thickness and material, and the method of calculating the material cost of body parts is also proposed in this paper. The change of material cost brought by the body front end in integrated lightweight and multi-objective optimization design is considered. (6) the Kriging approximation model optimization method is used to integrate the parameterized front end of the coupled model. The objective optimization design is to minimize the front end quality, the maximum static torsional stiffness and the lowest material cost in the body of the body, while restraining the static bending stiffness of the body in white, first order torsion, the bending modal frequency and the 100% collision safety performance of the front face, and considering the thickness of the parts, the shape of the section, the curvature, and the 26 different materials. Design variables, using NSGA- II algorithm to optimize the compromise solution set in the design space search. With the minimum quality of the compromise solution to the parameterized front-end model integrated light quantization optimization design. After the simulation analysis, it is found that the front end quality of the body in white is reduced by 5.81Kg, the weight loss rate is 7.01%, and the static bending stiffness of the body in white, the lower order mode and the positive. There is no obvious change in the safety performance, the material cost is only increased by 3.30%. and the simulation analysis shows that the optimized front end of the body has little effect on the safety performance of the side impact.
【学位授予单位】：吉林大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：U463.82
，

本文编号：1984686