【作者】 陈建岭；

【导师】 李剑峰；

【作者基本信息】 山东大学 ， 机械制造及其自动化， 2009， 博士

【摘要】 钛合金在航空航天领域有着广泛的应用,被誉为是一种使人类走向太空时代的战略性金属材料。然而,由于具有导热性系数低、高温化学活性高和弹性模量小特点,钛合金又是一种典型的难加工材料。在切削过程中,容易产生较高的切削温度,使刀具磨损加快,表面质量难以控制,实际切削速度难以提高。同时,由于航空钛合金零部件的结构设计特点,大量的材料需要从整块坯料中去除,使得切削成本较高。因此,低效的加工能力和加工质量与不断增长的加工需求已构成矛盾,成为制约航空航天制造业发展的瓶颈之一。针对航空钛合金Ti6Al4V的铣削加工,本课题从铣削力、铣削温度、表面完整性以及刀具寿命方面对高速铣削机理进行了系统研究,为建立钛合金高效加工工艺规范提供理论依据。以此为基础,以加工效率最大和刀具消耗最小为优化目标,对铣削参数进行优化,为实际加工提供工艺支持。切削力是研究切削过程的重要物理量之一,其大小和变化对工件表面完整性、刀具磨损和寿命等具有影响。准确地预测切削力对于优选切削用量、刀具结构参数以及提高加工精度具有积极的指导意义。针对圆柱螺旋铣刀,将切削刃沿轴向切削深度离散为一系列微元,而微元铣削力大小可表示为剪切力和犁耕力的和,然后将微元沿刀具轴向切削深度进行积分,从而建立了考虑刃口犁耕效应的三维铣削力学模型。与以往求解铣削力系数方法不同,本课题提出了应用瞬时铣削力识别铣削力系数的方法,具有更快速、更经济特点。考虑铣削力系数随被加工材料力学性能变化特点,通过试验和回归分析方法,建立了适合高速铣削范围的铣削力系数经验预测模型。经验证,所建立的铣削力学模型具有很高的预测精度。从控制铣削力角度,铣削参数应选择适中的铣削速度、较大的铣削径向深度、较小的进给速度和轴向铣削深度。对切削热控制的好坏一定程度上决定了钛合金零部件的生产效率。因此,有必要研究切削过程中热量的产生和传递的规律,了解刀具和工件中温度的分布状态,从而优化工艺措施,提高刀具寿命,进一步提高加工效率和加工质量。利用红外辐射测温方法,研究了铣削参数对铣削温度的影响。通过试验发现,随着铣削速度和每齿进给量增加,铣削温度呈增加趋势。以温度测量试验为参照,利用有限元方法,建立三维铣削仿真模型,对工件变形区和刀具的温度场进行了分析。工件上最高温度始终产生在切屑与切削刃接触的根部区域,该区域具有高温、大应变和高应变率特点。刀具的最高温度产生于前刀面主切削刃附近,由于具有良好的导热性,最高温度低于切屑的最高温度。当铣削速度增大,刀刃与切屑根部接触区域温度上升,并且高温热源区域沿轴向铣削深度扩大,同时静水压力也增大。航空钛合金零部件因其特殊的工作环境,可靠性要求较高,相应地对表面完整性也有严格的要求。深入研究铣削加工对钛合金表面完整性的影响,实现表面质量预报和控制,可为钛合金零部件设计以及加工工艺选择提供参考依据。根据铣削试验发现,使用涂层刀片和MQL能够获得较好的表面粗糙度。在采用较高速度铣削钛合金时,配合低进给、较低的轴向铣削深度以及适中的径向铣削深度,可以获得较好的表面粗糙度。综合工件与刀具相对运动的几何特征和工件材料弹塑性变形特点,建立了钛合金铣削加工表面粗糙度的理论模型,揭示了表面粗糙度的形成机理。通过对已加工表层微观组织进行SEM观察,发现低速和小进给情况下表层晶粒扭曲变形的程度和深度比较显著,而高速铣削减轻了表层晶粒扭曲的程度,并且减小了变质层的厚度,有利于提高零件表面完整性。通过显微硬度测量,研究了铣削速度和每齿进给对加工硬化的影响。加工硬化程度随铣削速度增加而降低,然而当铣削速度超过一定界限,加工硬化程度将会增加。随着每齿进给增加,加工硬化程度降低。无论高速铣削与低速铣削,硬度沿距离表面深度变化的趋势基本一致,都经历了表面强化-热软化-再度强化-趋于稳定的过程。高速铣削条件下软化层具有较低的硬度值,软化层深度略小于低速情况。根据位错理论,建立了加工硬化动力学模型,揭示了加工硬化与微观组织、流动应力的联系。利用铣削试验和X射线衍射应力测量方法,研究了铣削参数对已加工表面残余应力的影响。每齿进给对残余应力大小影响最为显著,而径向铣削深度影响最小。通过铣削力和铣削温度分析,认为残余应力与机械应力、热应力产生的塑性变形有关。切削力使工件表面层产生不均匀塑性变形,具体来自于两个方面,一方面是刀具接触点前方区域的“塑性凸出”效应,另一方面是刀具后刀面对工件表面的“挤光效应”。前者使已加工表面产生残余拉应力,后者则产生残余压应力。切削热使工件表层材料产生塑性凸出,从而产生残余拉应力。从X射线应力测量结果看,已加工表面残余应力表现为压应力性质。由于钛合金弹性模量小,已加工表面容易出现较大回弹,从而与刀刃后刀面产生强烈的挤光作用。因而可以认为,挤光作用是表面压应力的形成原因。钛合金切削加工容易引起刀具过早磨损,严重影响着加工效率和成本。通过磨损试验、SEM观察和能谱分析手段,对钛合金铣削加工刀具的磨损形式和磨损机理进行了分析。钛合金铣削加工刀具磨损和失效是在多种机制下共同作用的结果,主要表现为粘结磨损、氧化磨损、磨粒磨损、扩散磨损以及裂纹等形式。在高温高压下,钛合金工件表面材料和刀具后刀面容易产生粘结,当粘结层厚度达到一定程度,粘结层出现开裂和脱落现象,从而磨损产生。通过能谱分析,在干切削条件下,刀具材料以及粘结的钛合金材料均发生了氧化,而氧化产物的硬度较高,为磨粒磨损提供硬质点,并可能会使刀具发生塑性变形。在很高的切削温度下,接触表面发生大的塑性变形,不断产生的新生表面,促进了刀具材料与被加工材料之间的固体溶解,从而导致扩散磨损。根据磨损试验结果,建立了刀具寿命经验预报模型,发现铣削速度、每齿进给以及轴向铣削深度对刀具寿命具有显著性影响,而径向铣削深度影响较弱。针对钛合金铣削加工存在的高成本、低效率问题,建立了以最大生产效率和最小刀具寿命消耗为目标的铣削参数优化模型。对于多目标优化问题,由于目标的冲突性,在改善其中一个目标的同时,往往会削弱其他目标。因此,多目标优化问题要找的并不是所有子目标的最优解,而是Pareto解。经典的多目标问题求解方法一般是通过归一化处理将多目标问题转化为单目标问题。经典方法的不足主要表现为两个方面,一是很难建立合适的归一化准则,再就是每次只能得到一个Pareto解。进化算法具有很强的并行搜索能力,可以一次得到多个Pareo解,为决策者提供多样化的选择。本课题在NSGA-Ⅱ算法的基础上进行改进,将非支配概念拓展到约束空间,使得多目标多约束问题的处理更具有鲁棒性和有效性。根据钛合金铣削试验获得的数据,建立铣削力、粗糙度以及刀具寿命经验模型,代入铣削参数优化模型,对铣削参数进行优化,得到了Pareto前沿。工艺人员可根据优化目标的侧重灵活选择相应的铣削参数,比用传统方法优化钛合金切削参数更经济和方便。将该优化方法程序化,嵌入到工艺数据库,可为实际加工提供有利支持。更多还原

【Abstract】 With the reputation of space age metal, Titanium alloys have been used extensively in aerospace industry. However, it is difficult to machine due to their poor thermal conductivity, low elastic modulus and high chemical activation. During machining of Titanium alloys, rapid tool wear and deterioration of surface integrity make the cutting speed very low. Simultaneously, due to the characteristics of component design, a great deal of stock has to be removed from primary forms, resulting in a larger machining cost. Therefore, the contradiction between increasing machining demand and poor machining performance has been one of the bottlenecks blocking the development of aerospace industry. In this study, the high speed milling mechanism of Titanium alloys Ti6A14V was systematically researched from the sides of milling force, milling temperature, surface integrity and tool wear and life, in order to provide theoretical support for the built-up of Titanium alloys’ high performance machining technique. According to the understanding of machining mechanism, the milling parameters are optimized under the objectives of both maximized production efficiency and minimized exhaustion of tool life.Cutting force is one of the important physical variables in cutting process, which shows strong influence on workpiece’s surface integrity and tool wear. Therefore, it is encouraging to accurately predict cutting force in order to optimize cutting parameters and tool design. For end milling cutters, a three dimensional milling force model was proposed, in which the plough effect of tool edge’s radius was incorporated. The cutting edge was discretized into a series of small elements. The cutting forces of each small element includes shear force and plough force. Through integrating the cutting forces of small elements along axial depth of cut, the cutting forces of tool can be obtained. Unlike many traditional methods about identifying milling force coefficients, a more rapid and economical method was advised, which uses the instantaneous milling forces to identify milling force coefficients. Because milling force coefficients are influenced by workpiece’s material mechanical properties, an empirical model was built up with experiments and regress method, which is tenable under high speed condition. From the view of controlling of milling forces, medium milling speed, large radial depth of cut, small feed and axial depth of cut are preferred.The machining efficiency of Titanium alloys depends on the effect of heat control. Through studying the generation of heat and the distribution of temperature in workpiece and tool, the cutting process and tool life can be improved, further more, the machining efficiency and quality may be enhanced. The influences of milling parameters on milling temperature were investigated through infra-radiation measurement method. It was found that, the milling temperature increases with milling speed and feed per tooth. A three dimensional finite element model of milling was modeled to analyze the temperature distributions in workpiece’s deform zones and tool. The highest temperature in workpiece locates near the root zone of the chip, where high strain rate and large strain also occur. The highest temperature in the tool is near the primary cutting edge in the rake face, which is lower relative to that in workpiece due to good thermal conductivity. When milling speed increases, the temperature in the contact zone between the chip and the tool rises up simultaneously, the heat source is expanded along the axial depth of cut, and pressure also increases.Surface integrity directly determines the performance of components, which is the criterion in production design and selection of technique. According to the observations of experiments, coated inserts and MQL conditions can get lower roughness. When milling under higher speed condition, a combination with smaller feed per tooth, small axial depth of cut and medium radial depth of cut can lead to good roughness. According to the relative movement features between tool and workpiece and the mechanical properties of workpiece, a surface residual height model for horizontal machined surface was proposed. According to the observation of the microstructure in machined surface layer with SEM, the grains show a larger deformation depth and magnitude under low milling speed or small feed per tooth condition. High speed milling may weaken the deformation of grains and reduce the thickness of deformation layer, which is beneficial to component’s surface integrity. Through the measurement and analysis of microhardness, the influences of milling speed and feed per tooth were investigated. With the increase of milling speed while keep feed unchanged, the microhardness firstly decreases, then increases due to tool edge’s ploughing effect. The microhardness decreases with feed per tooth. A process of hardening-softening-hardening-leveling off occurs under both high speed milling condition and low speed condition. The softening layer has smaller mircohardness and depth under high speed milling condition. According to theory of dislocation, a dynamic model of work hardening was suggested, which discloses the relation between work hardening, microstructure and flow stress. By means of X radiation diffraction method, the influences of milling parameters on residual stress were investigated. The feed per tooth shows a strong effect on residual stress, while the radial depth of cut has a minor effect. Through analysis of milling forces and temperature, it can be concluded that the surface residual stresses are only relevant with the deformation induced by mechanical and thermal stress, and the burnishing effect between machining formation surface and tool’s flank surface results in the generation of compressive residual stress.During machining of Titanium alloys, rapid tool wear shortens tool’s life, which reduces the machining efficiency and increases machining cost. According to experiments, the types and mechanisms of tool wear during milling of Titanium alloys Ti6A14V were investigated, the empirical model tool life was proposed through regress method and the factors influencing tool wear and tool life were concluded. During milling of Titanium alloy Ti6A14V, the wear and failure of tool were a combination of many mechanisms such as adhesion, oxidization, attrition, diffusion-expansion and crack. Milling speed, feed per tooth and axial depth of cut have a strong effect on tool life, while the radial depth of cut has a minor effect.With an attempt to relief condition of the high machining cost and low machining efficiency existed in milling of Titanium alloys, a cutting process optimization model was proposed, in which the maximal production rate and minimal exhaustion of tool life were considered. For the multi-objective optimization problems, due to the conflict or lack of comparability between objectives, much encouraging efforts are to find the Pareto-optimal solutions. Classical optimization methods usually scalarize the vector of objectives into one objective by using some knowledge of the problem being solved. The main drawbacks of these algorithms are their sensitivity towards weights or demand levels and single point solution. Evolution algorithms have strong parallel search ability and can find multiple Pareto optimal solutions in a single run. In this study, the NSGA-II was improved with the extension of non-dominance concept from solution space to constraints space. The improvement makes the handling of problems with multiple objectives and multiple constraints more robust and effective. By virtue of milling force, roughness and tool life’s empirical models, an optimization process was implemented and the Pareto front was found. The technicians can choose milling parameters flexibly according to the preference to the objectives. It is more economical and convenient when compared with weight methods. It will provide benifical support for milling of Titanium alloys if it is programmed into machining database.更多还原