基于加工位置不确定的多工步数控铣削工艺参数优化研究*

针对铣削稳定性评价指标极限切削深度随加工位置改变而变化,导致铣削工艺参数优化模型中稳定性约束具有不确定性问题,结合不同加工位置刀具频响函数和切削稳定性理论,建立加工空间极限切削深度广义回归神经网络(GRNN)预测模型,基于该GRNN模型完善铣削稳定性约束条件,进而构建以机床各运动部件位移与粗/精加工切削参数为变量,以粗/精加工总切削时间为目标的多工步数控平面铣削工艺参数优化模型,采用粒子群算法(PSO)求解该优化模型。以某企业加工中心展开实例研究,获取机床加工位置和粗/精加工主轴转速、切削深度、切削宽度、每齿进给量的优化配置,优化后粗/精加工总切削时间比优化前缩短22.47%,并通过该配置下的无颤振铣削加工验证了优化模型的有效性。     

考虑主轴-刀柄结合面特性的机器人铣削系统刀尖频响预测研究

针对机器人铣削系统刀尖频响具有位姿依赖特性,导致机器人变位姿加工时稳定性难以准确预测、加工颤振难以有效控制的问题,提出一种考虑主轴-刀柄结合面接触刚度的机器人铣削系统刀尖频响预测方法。基于欧拉-拉格朗日法与吉村允孝单位面积法,分别构建了机器人本体动力学模型与主轴-刀柄结合面接触刚度模型,进而基于有限元主副自由度理论将机器人本体动力学模型与主轴-刀柄结合面接触刚度模型结合,构建了机器人铣削加工系统刀尖频响预测模型。开展了机器人不同位姿下刀尖频响预测验证实验,结果表明,仿真与实验得到的刀尖频响函数相比,固有频率最大误差为6.63%,对应幅值最大误差为9.80%,验证了所提出的预测模型的准确性,证明了该模型能够实现机器人任意位姿下的频响函数准确预测。     

切削稳定性约束下的铣削参数优化技术研究

目前切削稳定性研究主要集中在不同加工方法及加工条件下的稳定性研究,以无颤振极限切深作为切削参数优化的推荐值,缺少对稳定性与优化模型的深度融合分析。针对这一问题,以材料切除率和刀具寿命构建优化目标函数,提出一种切削稳定性约束下的铣削参数优化模型。通过对铣削稳定性零阶解析算法的分析,论述了切削稳定区域的确定受切削力学模型、刀尖频率响应、及切削参数共同影响关系。在设定机床、工件和刀具的条件下,通过对稳定性叶瓣图形态随切削参数变化规律的研究,得出了极限切深不等效于最大材料去除率;以动态变化的稳定域及机床能效为约束边界,采用变化趋势相反的材料去除率和刀具寿命构建优化目标函数,通过遗传算法获取全局最优解。针对多目标优化中,各分目标权重难以量化设置的问题,提出以材料去除率期望值和刀具寿命期望值作为优化模型设置参数,实现优化参数的量化调节和优化方向的有效控制。在VMC850机床上进行了试验并采用遗传算法对多组参数设定状态进行优化,结果表明切削参数优化结果满足稳定性约束要求,且其优化方向可量化调节。     

一种新的预测铣刀刀尖频响函数的方法

为准确快速地预测铣刀刀尖点频响函数,提出一种基于逆响应耦合子结构分析（IRCSA）法辨识刀柄-刀具结合面参数的刀尖点频响函数预测方法。该方法通过建立计算刀柄末端频响函数矩阵和刀尖点频响函数矩阵的数学模型,利用逆响应耦合子结构分析法求取随频率变化的刀柄-刀具结合面参数。通过Cuckoo search算法及有限元分析确定刀尖点频响函数中对刀柄-刀具结合面复刚度矩阵变化最为敏感的固有频率,取该频率对应的结合面参数为刀柄-刀具结合面复刚度矩阵的辨识结果,由此计算出刀尖点频响函数。通过硬质合金圆柱棒、2刃铣刀和4刃铣刀进行验证,对比了所提预测方法、Cuckoo search优化算法预测的刀尖点频响函数与实测值三者之间的差异,实验结果表明该预测方法预测的刀尖点频响函数的固有频率和实测固有频率的误差在5%以内,所用时间约为Cuckoo search优化算法的1%,达到了较高的预测精度,并且更加省时、简便。     

基于RCSA与GA的铣刀刀尖点频响函数预测

基于响应耦合子结构分析法建立了铣刀刀尖点的频响函数预测模型,针对其中主轴—刀柄基座系统频响函数矩阵难以理论计算和实验测试的问题,提出了基于实验模态测试与遗传算法寻优相结合的主轴—刀柄基座系统频响函数矩阵的优化拟合方法。以VMC850E型立式加工中心主轴系统为研究对象,基于上述方法对铣刀刀尖点频响函数进行了预测,并与其实测频响函数进行对比。结果表明:刀尖点的预测频响函数曲线与实测频响函数曲线符合较好,预测频响函数的固有频率与实测固有频率在0~4 000Hz范围内的相对误差小于4.37%。     

有限样本下基于迁移学习的铣削稳定性预测方法

传统铣削稳定性分析因采用静态刀尖点频响函数和平均切削力系数而使其在真实工况下的预测精度降低。为此,引入迁移学习提出一种基于少量实验样本的铣削稳定性预测方法。首先,生成静态刀尖点频响函数和平均切削力系数在全转速范围内多个系列的随机值,并在各系列下进行铣削稳定性分析,通过计算少量极限切削深度实验值与对应的预测值之间的误差,确定最优系列并以其构造源域稳定域数据;然后,利用大量源域数据建立极限切削深度的预训练模型,通过少量实验样本全局微调此模型使其适应真实加工场景。以40组颤振实验样本展开实例验证,所提方法比采用少样本建模的预测精度提升32%,并对比不同数据规模下各类模型预测精度,共同验证所提方法的有效性。     

一种改进的基于响应耦合子结构法的刀尖点频响函数预测方法

针对现有的基于响应耦合子结构法(RCSA)的刀尖点频响函数预测方法需要辨识主轴-刀柄、刀柄-刀具结合面参数以及需要自制刀柄模型等引起的预测误差和预测过程复杂等问题,提出一种改进的基于RCSA的铣刀刀尖点频响函数预测方法。该方法首先改进已有的子结构划分方法,将机床-主轴-刀柄-刀具系统划分为机床-主轴-刀柄-部分刀杆、剩余刀杆和刀齿三个子结构;然后改进主轴-刀柄处转动频响函数的计算方法,通过铣刀的模态锤击实验采用反向RCSA和有限差分法计算机床-主轴-刀柄-部分刀杆结构的转动频响函数,并基于Euler梁模型计算出剩余刀杆、刀齿子结构的频响函数;最后将三个子结构的频响函数耦合确定刀尖点的预测频响函数。以一立式加工中心为研究对象,应用所提出的方法对铣刀刀尖点的频响函数进行了预测,并与其实测频响函数进行对比。对比结果表明：刀尖点的预测频响函数与实测频响函数符合程度较高,其预测、实测前三阶固有频率之间的误差在6.9%以内,所提出的方法可行有效、简单方便,且可直接基于铣刀的模态实验计算主轴-刀柄的频响函数,避免了相关结合面参数的辨识和刀柄模型的制作。     

翻板卧式加工中心铣削颤振稳定性的影响因素研究

以某翻板卧式加工中心为研究对象,基于铣削颤振稳定性叶瓣图理论,通过实验与仿真结合的方法得到机床铣削稳定性的影响因素及变化规律,为工程人员提供参考。首先通过敲击实验获取刀尖点频响函数,再通过ANSYS Workbench中使用虚拟材料轴承的谐响应仿真获取频响函数,对比实验与仿真的结果,得到刀尖频响函数的影响因素。再由刀尖频响函数绘制铣削颤振稳定性叶瓣图,进而得到机床铣削稳定性的影响因素。研究结果表明,机床铣削稳定性几乎不随主轴位置移动而变化,不同刀具的铣削稳定性变化规律及其适宜铣削转速与不宜铣削转速相同,可在工艺设计、机床电主轴选型及电主轴结构设计等方面为工程人员提供帮助。     

高速切削钛合金稳定性模型及极限切深预测

针对TC4钛合金高速加工过程中颤振及切削稳定性问题进行了研究。首先根据闭环加工系统的动力学模型,建立了高速切削过程的稳定性模型,并通过解析计算的方法得到了加工过程中稳定切削极限深度与转速之间的关系。然后采用力锤敲击方法获取了刀具的频响函数,为切削稳定性模型的建立奠定了基础,最后采用切削实验验证了模型的准确性。结果表明:采用高速切削方式加工TC4钛合金能更好的避免加工过程的颤振现象,建立的稳定模型能很好地预测加工过程的稳定性。     

基于迁移学习的变刀具-刀柄夹持状态下数控铣削稳定性预测研究

分析刀具-刀柄结合状态变化时的数控铣削稳定性,其效率因需重复测量刀尖点频响函数而降低。针对此问题,引入迁移学习提出仅需测量目标刀具少量悬伸量下刀尖点频响函数的铣削稳定性预测方法。首先,测量源刀具多个悬伸量和目标刀具少量悬伸量的刀尖点频响函数,采用铣削稳定性解析法获取各铣削参数组合的加工振动状态信息,构建充足的源域数据和少量的目标域数据,并通过源域和目标域的相似匹配筛选源域样本,然后结合神经网络和TrAdaBoost迁移学习算法,自适应更新源域与目标域混合样本权重,建立目标刀具的铣削加工振动状态分类器。以三组刀柄-刀具组合进行实例分析,少样本下采用迁移学习后两组目标刀具的分类器精度分别提升了10.93%、6.25%,并通过铣削实验验证了所提方法的有效性。     

Anisotropic force ellipsoid based multi-axis motion optimization of machine tools

The existing research of the motion optimization of multi-axis machine tools is mainly based on geometric and kinematic constraints, which aim at obtaining minimum-time trajectories and finding obstacle-free paths. In motion optimization, the stiffness characteristics of the whole machining system, including machine tool and cutter, are not considered. The paper presents a new method to establish a general stiffness model of multi-axis machining system. An analytical stiffness model is established by Jacobi and point transformation matrix method. Based on the stiffness model, feed-direction stiffness index is calculated by the intersection of force ellipsoid and the cutting feed direction at the cutter tip. The stiffness index can help analyze the stiffness performance of the whole machining system in the available workspace. Based on the analysis of the stiffness performance, multi-axis motion optimization along tool paths is accomplished by mixed programming using Matlab and Visual C++. The effectiveness of the motion optimization method is verified by the experimental research about the machining performance of a 7-axis 5-linkage machine tool. The proposed research showed that machining stability and production efficiency can be improved by multi-axis motion optimization based on the anisotropic force ellipsoid of the whole machining system.     

基于模态区间的数控机床切削状态监测

随着高速高精数控加工技术的发展,对数控机床切削加工状态的稳定性提出了更高的要求,传统的切削加工状态监测方法中对不确定性处理存在不足。提出了一个基于模态区间的切削状态监测不确定性处理方法,利用模态区间的宽度对传统监测方法中的不确定性加以表述,以解决监测中的不确定性问题。为了验证提出方法的有效性,搭建了切削加工实验平台,通过加速度传感器获取数控机床切削加工信息,由时频分析方法将切削状态划分成稳定、过渡及颤振3个加工阶段,利用基于模态区间的小波包能量百分比方法,提取不同加工阶段的区间特征量,通过Lloyd算法进行编码后作为基于模态区间的广义隐马尔科夫模型的输入特征向量,最后利用广义隐马尔科夫状态辨识方法,对数据机床切削状态进行了识别。实验结果表明,基于模态区间的广义隐马尔科夫模型辨识方法优于传统的隐马尔科夫模型辨识方法。     

Three dimensional cutting force analysis in end milling

The analysis of cutting forces plays an important part in the design of machine tool systems as well as in the planning, optimization, and control of machining processes. This paper presents a three-dimensional model of cutting forces in peripheral end milling in terms of material properties, cutting parameters, machining configuration, and tool/work geometry. Based on the relationship of the local cutting force and the chip load, the total cutting force model is established via the angle domain convolution integration of the local forces in the feed, cross feed, and axial directions. The integration is taken along the cutter axis and summarized across the cutting flutes. The convolution integral leads to a periodic function of cutting forces in the angle domain and an explicit expression of the dynamic cutting force components in the frequency domain. The closed-form nature of the expressions allows the prediction and optimization of cutting forces to be performed without the need of numerical iterations. To assess the fidelity of the analytical model, experimental data from end milling tests are presented in the context of three dimensional time waveforms, power spectra, and phase angles, in comparison to the values predicted by the model.     

Integrated simulation method for interaction between manufacturing process and machine tool

The interaction between the machining process and the machine tool (IMPMT) plays an important role on high precision components manufacturing. However, most researches are focused on the machining process or the machine tool separately, and the interaction between them has been always overlooked. In this paper, a novel simplified method is proposed to realize the simulation of IMPMT by combining use the finite element method and state space method. In this method, the transfer function of the machine tool is built as a small state space. The small state space is obtained from the complicated finite element model of the whole machine tool. Furthermore, the control system of the machine tool is integrated with the transfer function of the machine tool to generate the cutting trajectory. Then, the tool tip response under the cutting force is used to predict the machined surface. Finally, a case study is carried out for a fly-cutting machining process, the dynamic response analysis of an ultra-precision fly-cutting machine tool and the machined surface verifies the effectiveness of this method. This research proposes a simplified method to study the IMPMT, the relationships between the machining process and the machine tool are established and the surface generation is obtained.     

Selection of optimum tool geometry and cutting conditions using a surface roughness prediction model for end milling

Influence of tool geometry on the quality of surface produced is well known and hence any attempt to assess the performance of end milling should include the tool geometry. In the present work, experimental studies have been conducted to see the effect of tool geometry (radial rake angle and nose radius) and cutting conditions (cutting speed and feed rate) on the machining performance during end milling of medium carbon steel. The first and second order mathematical models, in terms of machining parameters, were developed for surface roughness prediction using response surface methodology (RSM) on the basis of experimental results. The model selected for optimization has been validated with the Chi square test. The significance of these parameters on surface roughness has been established with analysis of variance. An attempt has also been made to optimize the surface roughness prediction model using genetic algorithms (GA). The GA program gives minimum values of surface roughness and their respective optimal conditions.     

Analysis of the machining stability of a milling machine considering the effect of machine frame structure and spindle bearings: experimental and finite element approaches

Chatter vibration has been a troublesome problem for a machine tool toward the high precision and high-speed machining. Essentially, the machining performance is determined by the dynamic characteristics of the machine tool structure and dynamics of cutting process of spindle tool. Prediction of the dynamic behavior at spindle tool tip is therefore of importance for assessing the machining stability of a machine tool at design stage. This study was aimed to evaluate the machining stability of a vertical milling system under the interactive influence of the spindle unit and the machine frame structure. To this end, a realistic finite element model of a vertical milling tool was generated by incorporating the spindle-bearing model into the head stock mounted on machine frame. The influences of the dynamics of spindle-bearing system and the machine frame structure were investigated respectively. Current results show that the machine tool spindle system demonstrates different dynamic behaviors at different frequency ranges, which are also characterized as structural modes and spindle modes, respectively. In particular, the maximum compliance of spindle tool tip was found to occur at the bending vibrations of spindle shaft and vary with the preload amount of spindle bearing. The machining stabilities were predicted to different extent, depending on the exciting modes which could be related to the influences of machine frame and spindle unit.     

基于动力学不确定性的重型切削工艺参数优化

针对数控重型切削加工过程的切削稳定性具有不确定性的特点,提出了在切削稳定性和机床工作能力的约束下,获得最大材料去除率的工艺参数优化方法。根据重型切削加工的工艺特点建立三维动力学模型,以机床的固有频率、阻尼比、刚度和切削力系数作为不确定因素,结合排零定理和边理论对其进行不确定性分析,获得稳健的切削稳定性叶瓣图,结合切削深度、刀具直径和刀具齿数的关系,为加工过程选择能获得最大切削深度的刀具。在此基础上,建立工艺参数优化模型,选择最佳的轴向切削深度、径向切削深度和主轴转速的组合,最后以一台加工中心上某型号发动机缸体表面的粗加工过程为例进行了验证。     

A method for off-line compensation of cutting force-induced errors in robotic machining by tool path modification

Industrial robots are promising cost-effective and flexible alternatives for multi-axis milling applications in machining of complex parts of light materials with lower tolerances, having freeform surfaces. As it is well-known, the poor accuracy, stiffness, and the complexity of programming are the most important limiting factors for wider adoption of robotic machining in machine shops. The paper presents the developed method for off-line compensation of machining robot tool tip static displacements as a dominant part of cutting force-induced errors. The developed method is based on modification of programmed trajectory in G-code. Off-line modification of programmed trajectory is performed according to the predicted static tool tip displacements calculated based on developed robot compliance model and cutting forces predicted by mechanistic model. The obtained experimental results show the relevance of developed method since the machining errors could be significantly reduced. This allows the desired accuracy of robot machining to converge towards nominal specifications.     

An integrated prediction model for the dynamics of machine tool spindles

ABSTRACT

The predictility of dynamics of the machine tool spindles is essential for machining precision. During machining, the machine tool components and the cutting process interact with each other. Accordingly, it is necessary to take the process-machine interaction effects into account in order to predict the spindle's dynamics accurately. This paper presents an integrated model for the prediction of a spindle's dynamics. The model synthesizes the interactive influence between machine dynamics and forces in grinding process. The thermo-mechanical model of the spindle with angular contact ball bearings was built by using the finite-element method. The analytical model was used to calculate the process forces. A coupled simulation was adopted to accomplish the interactive process between the two models. Basing on the integrated model, the bearing stiffness, the natrual frequency, the spindle tip stiffness and deformations of a grinder's spindle were investigated. The prediction of the deformation fluctuations at the spindle tip due to process-machine interaction was also achieved.     

A manufacturing model of helical groove on rotary burr and a universal post processing method

A systematic machining theory and precision method to determine cutter location in a grinding system is presented for rotary burr. First, the helical cutting edge on various kinds of revolving surfaces is built. Then, based on the geometry model of the helical cutting edge, the smooth spiral rake surface with constant normal rake angle and flank surface can been formed during the one-pass grinding process by this method. No interference between the grinding wheel and workpiece happens by the wheel special rotation. The method has the characteristic of detaching the grinding wheel path solution from specified machining conditions. The grinding wheel path is suitable for different NC machine tools through post processing. Meanwhile, a mechanism kinematic model of the NC machine tool is built, and a generalized algorithm for post-processing of multi-axis NC machine tools is presented. This model is applied to arbitrary configuration of NC machine tool, and the motion value for each axis will be generated by the inputting structure and motion parameters of the machine tool. The model, together with the machining method mentioned in this paper, make the calculation and generation of the grinding wheel path simpler and universal. At last, the validity of the method given in the paper is identified by an example of grinding.