【作者】 张勇；

【导师】 李光耀；

【作者基本信息】 湖南大学 ， 车辆工程， 2009， 博士

【摘要】 随着汽车工业在国内的蓬勃发展,使得能源日趋紧张、环境压力加剧的问题日益凸显,这必将成为汽车工业可持续发展的绊脚石。作为有效节能手段的汽车轻量化技术已成为汽车产业发展中的一项关键性研究课题,在不久的将来,它必将成为汽车公司的核心竞争力之一。近年来国内外很多学者在汽车轻量化相关方面开展了较为广泛的研究,并已取得诸多成果,但是汽车轻量化的相关研究仍然存在较多问题,研究方法仍然集中在对汽车车身的一些零部件用低密度材料进行替换,如铝、镁合金、复合材料等,这些方法大多数仍是凭经验设计。本文在总结前人的研究基础上,采用了近似模型技术结合数值优化方法对汽车轻量化设计进行了较为系统和深入的研究,在确保汽车综合性能指标的前提下,尽可能的降低汽车产品自身重量,达到减重、降耗、安全的综合指标。围绕汽车轻量化这一主题,本论文主要开展了以下几方面的研究内容:1.建立了一种基于移动最小二乘的汽车正面碰撞响应面近似模型,提出了针对不同的优化设计响应采用不同的近似模型技术来达到提高优化效率的目的。通过与目前广泛运用的多项式响应面与Kriging响应面的对比研究可知,多项式响应面只是对空间试验点的逼近,拟合精度不高,故一般适用于替代非线性程度不高的响应,而Kriging模型准确的通过了空间拟合的采样点,拟合精度高,但它容易受到数值噪声的影响且计算成本高。而移动最小二乘拟合的响应面近似模型恰恰兼顾了两者的优点,故它非常适合于具有高度非线性且具有数值噪声的汽车碰撞安全的优化研究。最终把该模型成功运用到汽车轻量化优化设计中来验证了此模型的有效性。2.提出了一种基于近似模型和多目标遗传算法的高强度钢板选材方法。虽然汽车制造工业中铝合金等轻质材料在逐渐增加,但是由于铝合金等加工难度高,成形性能较差,又由于铝导热性好,导致铝合金的焊接性能差,关键是它的成本还比钢材高。故目前在汽车制造中仅能部分代替钢铁材料。而厚度变薄的高强度钢相对于普通钢来说则具有高强度、质量轻等优点,而且还能提高车辆的安全性能,所以高强度钢已成为普遍采用的汽车原材料。采用高强度钢板代替普通钢板制造的汽车车身,其重量可以得到较大幅度的降低。但是传统的选材设计往往都是根据工程人员的经验来选择车辆各个部分的材料配置,导致车辆各个部分的材料选择具有盲目性而不能最优化,文中采用近似模型和多目标遗传算法相结合的选材方法对整车的正面碰撞和侧面碰撞安全性进行了深入研究,得到了汽车车身驾驶室前部关键部件与侧围的高强度钢板及其板厚的合理配置,并达到使车辆的安全性能提高和车身轻量化的目的。3.提出了一种基于近似模型和多学科设计优化相结合的整车轻量化优化设计方法。汽车轻量化的首要前提是应保持汽车原有的性能不受影响,既要有目的地减轻汽车自身的重量,又要保证汽车行驶的安全性、耐撞性、抗振性及舒适性等。但是传统的汽车设计模式往往采取的是一种串行设计模式,在不同的设计阶段设计人员将选择不同的重点来对汽车进行设计和优化,这就把影响汽车性能的空气动力学、振动、噪声、耐撞性、安全性以及舒适性等学科人为的割裂开来,并没有利用和考虑到各个学科之间存在的相互耦合而可能产生的协同效应,这样将导致得到的优化结果往往不是系统整体的最优解。同时,如果轻量化设计中一起考虑到的汽车安全性、振动、刚度等各学科的相互作用来进行优化设计,这样势必遭遇优化设计的瓶颈(单次CAE计算仿真时间过长而引起的时间问题)。正是基于上述传统串行设计与CAE仿真时间过长的缺陷,文中提出了汽车的轻量化优化设计过程也必是一个多学科设计优化的过程,并建立了将响应面和多学科设计优化方法相结合的汽车轻量化优化设计近似模型,成功解决了单次优化时间过长的优化瓶颈问题,并将该方法成功运用到某车型的轻量化优化设计中。4.建立了基于可靠性设计理论的汽车薄壁纵梁轻量化优化设计模型。车身薄壁纵梁不仅是汽车车身设计的基本承载部件,而且也是车身发生正面碰撞时的主要吸能部件。同时,薄壁纵梁的变形模式和吸能特性直接决定了车体在碰撞时的加速度和力的响应,也就直接影响到车内乘员的安全性。此外,由于它的使用环境较为恶劣,这就对它的设计提出了更高的可靠性要求。传统的优化过程往往忽略了工程设计中的不确定性,如几何尺寸、材料属性、载荷和边界条件以及操作环境等相关的多种不确定因素的影响,而使设计目标超出设计界限而失效,从而降低了产品的可靠性。文中提出将薄壁梁设计看着是一个需要考虑安全性、静态承载特性与一阶振动模态等方面的多学科可靠性优化设计过程,并将该方法成功运用到帽型薄壁纵梁的轻量化优化设计中。更多还原

【Abstract】 Along with the fast development of automotive industry, the energy tense and environment pressure becomes increasingly serious, which will prevent the development of automotive industry. As an effective energy-saving method, automotive lightweight has already become a crucial research topic. In the near future, it must be one of core competence among car companies. In recent years, many research achievements have been made in the automotive lightweight field. However, many lightweight-related issues need to be studied. Research methods still concentrated on replacing some automobile body parts with low-density materials, such as aluminum, magnesium alloy, composite materials, and so on. These methods almost depend on the empirical design. Therefore, in this paper, based on the predecessor’s literature research, an in-depth study about the lightweight, using approximate model technology and optimization method, was conducted, so as to reduce the weight of automobile and achieve aggregative objective (reducing weight and consumes and improving vehicle safety) , under the guarantee of comprehensive performances of automobiles.Main research content and innovation of the paper are as follows:1. Based on moving least square method, a response surface approximate model of automotive front impact was established, and used the different approximate model technology for the different optimized design response to achieve the goal of enhancing optimization efficiency. By contrasting to Kriging response surface, the polynomial response surface is approaching the sampling points in design space, and the fitting accuracy is not high, so it is generally applied to the responses of low non-linear level, while the Kriging model whose fitting accuracy is high, can pass the spatial sampling point accurately, but it is vulnerable to numerical noise which raise the cost of calculation. However, the moving least square response surface model has both merits above, so it is very suitable for the car crash safety optimization of highly non-linear, and numerical noise. And, the model’s validity is confirmed by utilizing the model in the automotive lightweight.2. A method of high strength steel plate selection which based on the multi-objective genetic algorithm and approximate model was proposed. Although the application of the light quality materials such as aluminum alloy increases gradually in automotive manufacturing industry, due to the difficulty of processing, forming, and welding results from the good thermal conductivity, and even the higher cost than the steel, aluminum can only be used as vehicle body instead of only part of the iron and steel in the automobile manufacture at present. While the thin high-strength steel has merits of lighter quality, higher strength, as compared to the traditional steels. And it also can improve the performance of vehicle safety. Therefore, the high-strength steel has been widely used as automotive raw materials. Using the high-strength steel plate instead of traditional steel plate, the automobile body weight can be reduced to a greater degree. However, the traditional selection of materials often based on the experiences of engineers, which made the choice of materials in every part of vehicles blind and not optimal. In the paper, the safety of front impact and side impact of vehicles are conducted in deep study by combining the approximate model and the multi-objective genetic algorithm, make the high strength steel plate of front key absorbing parts of the automobile cab and the side parts and its thickness reasonable in disposition. And, the safety of vehicles was improved, and the weight of automotive body was reduced.3. A method of the entire vehicle lightweight optimization design basing on the approximate model and the multi-disciplinary design optimization was proposed. With the most important premise which is to maintain the original performance, the automobile lightweight, not only destinate to reduce the weight, but also ensure the safety, crashworthiness, vibration, comfort, and so on. However, because the traditional model of car design was often taken as a serial design patterns, in which engineers chosed different key points to design and optimize in different stages, the influence of automobile performance was separated about the aerodynamic performance, vibration, noise, crashworthiness, safety, comfort, and the coordination effect which be produced by coupling of various subjects was not used and taken into account. So it always turned out the optimization results were not the overall optimal solution. At the same time, if the lightweight design of automobile takes the effects of safety, vibration, stiffness into account, this is bound to encounter bottleneck of the optimal design (CAE time of single calculation and simulation is too long). It is based on the flaw above-mentioned traditional serial design and CAE simulation time, and the paper propose that the process of the lightweight of automobile is a multi-disciplinary design optimization process, and establish the approximate model of automobile lightweight optimization design by combining the response surface method and multi-disciplinary design optimization method, and solve the problem that the time of optimization is too long one-time. The method was utilized in lightweight design of a certain vehicle successfully.4. A design model of thin-walled beam for lightweight optimization was established based on the reliability of design theory. Thin-walled beam of the body is not only a basic loading part in the body design, but also a main energy-absorption part in frontal impact. Meantime, impact responses of acceleration and impact force of a vehicle and occupant safety at collision depend on the pattern of deformation and energy absorption characteristic of the thin-walled beam. In addition, the poor environment in use leads to higher design requirement of reliability . The traditional process of optimization often ignore the effects of uncertainty of the engineering design, such as geometry size, material properties, loads and boundary conditions and operating environment and so on, which leads to the design failure beyond the constraints and reduces the reliability of the product. In the paper, the thin-walled beam design was regarded as a multi-disciplinary reliability optimization design process of considering the safety, the static load-bearing characteristics and the first-order vibration mode. And this method has been applied to the lightweight optimization design of cap-thin-walled successfully.更多还原