技术领域technical field
本发明涉及机械再制造领域。The invention relates to the field of mechanical remanufacturing.
背景技术Background technique
随着资源短缺与环境问题的日益加剧，再制造工程受到了广泛的关注。再制造可使废旧资源中蕴含的价值得到最大限度的开发和利用，缓解资源短缺与资源浪费的矛盾，减少大量的失效、报废产品对环境的危害，是废旧机电产品资源化的最佳形式和首选途径，是节约资源的重要手段。With the increasing shortage of resources and environmental problems, remanufacturing engineering has received extensive attention. Remanufacturing can maximize the development and utilization of the value contained in waste resources, alleviate the contradiction between resource shortage and waste of resources, and reduce the environmental hazards of a large number of invalid and scrapped products. It is the best form and resource utilization of waste mechanical and electrical products The preferred way is an important means of saving resources.
然而，当前废旧零部件的再制造过程存在人工参与多，经验依赖性强，修复效率低、可靠性差，修复过程不可逆等问题。However, the current remanufacturing process of waste parts has many problems such as manual participation, strong experience dependence, low repair efficiency, poor reliability, and irreversible repair process.
发明内容Contents of the invention
本发明的目的是解决废旧零部件的再制造过程中，修复方式和参数难以标准化和量化的问题。The purpose of the invention is to solve the problem that repairing methods and parameters are difficult to standardize and quantify in the remanufacturing process of waste parts and components.
为实现本发明目的而采用的技术方案是这样的，一种基于逆向工程的零部件再制造方法，其特征在于，包括以下步骤：The technical scheme adopted for realizing the purpose of the present invention is such that a method for remanufacturing parts based on reverse engineering is characterized in that it comprises the following steps:
1)获取废旧零部件的表面点云数据模型。获取物体表面点云数据模型的方式很多，根据测量探头是否与测量表面接触，可以分为接触式测量和非接触式测量两大类。接触式测量常用设备为三坐标测量机(CMM)，非接触式测量常用设备包括激光扫描仪，结构光扫描仪，工业CT机等。1) Obtain the surface point cloud data model of waste parts. There are many ways to obtain the point cloud data model of the object surface. According to whether the measurement probe is in contact with the measurement surface, it can be divided into two categories: contact measurement and non-contact measurement. The commonly used equipment for contact measurement is coordinate measuring machine (CMM), and the commonly used equipment for non-contact measurement includes laser scanners, structured light scanners, industrial CT machines, etc.
2)获取所述废旧零部件的原始CAD模型；2) obtaining the original CAD model of the waste parts;
3)将步骤1)的表面点云数据模型与步骤2)的原始CAD模型配准；3) the surface point cloud data model of step 1) is registered with the original CAD model of step 2);
4)根据步骤3)的配准的结果，获得所述废旧零部件的最大损伤深度；4) According to the registration result of step 3), obtain the maximum damage depth of the waste parts;
5)若步骤4)获得的最大损伤深度低于阈值，以所述最大损伤深度为进给量，对所述废旧零部件进行减式修复。值得说明的是，减式修复是指在原零部件基体上去除材料的修复方式，即通过车、铣、磨等机械加工方式对零部件损伤表面进行再加工，直至将表面损伤完全去除。5) If the maximum damage depth obtained in step 4) is lower than the threshold value, use the maximum damage depth as the feed rate to perform subtractive repair on the waste parts. It is worth noting that subtractive repair refers to the repair method of removing material on the original part base, that is, reprocessing the damaged surface of the part through turning, milling, grinding and other mechanical processing methods until the surface damage is completely removed.
若步骤4)获得的最大损伤深度高于阈值，对所述废旧零部件进行加式修复。值得说明的是，加式修复是指在废旧零部件基体上添加材料的修复方式，常见加式修复工艺有激光熔敷、热喷涂、堆焊等，激光熔敷工艺因其适用的材料体系广泛、熔覆层与基体结合强度高、基体热变形小及工艺过程易于实现自动化等特点，已越来越多的应用于再制造修复中。If the maximum damage depth obtained in step 4) is higher than the threshold, an additive repair is performed on the waste parts. It is worth noting that additive repair refers to the repair method of adding materials to the matrix of waste parts. Common additive repair processes include laser cladding, thermal spraying, and surfacing welding. Laser cladding process is suitable for a wide range of material systems. , high bonding strength between the cladding layer and the substrate, small thermal deformation of the substrate and easy automation of the process, etc., have been more and more used in remanufacturing and repairing.
本专利从系统的角度提出了基于逆向工程的废旧零部件再制造流程框架，该框架包含加式修复与减式修复两条主线，在分析废旧零部件表面点云数据模型与原始CAD模型间差异的基础上，对废旧零部件展开修复。This patent proposes a framework for the remanufacturing process of waste parts based on reverse engineering from a systematic point of view. This framework includes two main lines of additive repair and subtractive repair. It analyzes the difference between the surface point cloud data model of waste parts and the original CAD model. On the basis of repairing waste parts.
附图说明Description of drawings
图1为基于逆向工程的废旧零部件再制造流程框架；Figure 1 is the framework of the remanufacturing process of waste parts based on reverse engineering;
图2为废旧零部件表面点云数据模型采集步骤；Figure 2 is the collection steps of the surface point cloud data model of waste parts;
图3为废旧零部件表面点云数据模型采集结果；Figure 3 is the collection result of the surface point cloud data model of waste parts;
图4为废旧零部件表面点云数据模型预处理；Figure 4 is the preprocessing of the surface point cloud data model of waste parts;
图5为废旧零部件表面点云数据模型的损伤边界划分；Figure 5 is the damage boundary division of the surface point cloud data model of waste parts;
图6为废旧零部件原始CAD模型的重构；Fig. 6 is the reconstruction of the original CAD model of waste parts;
图7为废旧零部件缺失部位模型的提取；Figure 7 is the extraction of the missing part model of waste parts;
图8为表面点云数据模型与原始CAD模型配准示意图；Fig. 8 is a schematic diagram of registration of the surface point cloud data model and the original CAD model;
图9为改进ICP配准算法流程；Fig. 9 is the flow chart of the improved ICP registration algorithm;
图10为废旧零部件表面损伤深度示意图；Figure 10 is a schematic diagram of the surface damage depth of waste parts;
图11为废旧模具表面点云数据采集步骤；Fig. 11 is the step of collecting point cloud data on the surface of the waste mold;
图12为废旧模具减式修复流程；Figure 12 is a subtractive repair process for waste molds;
图13为传统ICP与改进ICP算法配准结果对比；Figure 13 is a comparison of registration results between traditional ICP and improved ICP algorithms;
具体实施方式Detailed ways
下面结合附图和实施例对本发明作进一步说明，但不应该理解为本发明上述主题范围仅限于下述实施例。在不脱离本发明上述技术思想的情况下，根据本领域普通技术知识和惯用手段，做出各种替换和变更，均应包括在本发明的保护范围内。The present invention will be further described below in conjunction with the accompanying drawings and embodiments, but it should not be understood that the scope of the subject matter of the present invention is limited to the following embodiments. Without departing from the above-mentioned technical ideas of the present invention, various replacements and changes made according to common technical knowledge and conventional means in this field shall be included in the protection scope of the present invention.
实施例1：Example 1:
一种基于逆向工程的零部件再制造方法，包括以下步骤：A method for remanufacturing components based on reverse engineering, comprising the following steps:
1)获取废旧零部件的表面点云数据模型。参见图2，图中，采用ATOS光学三维扫描仪对废旧零部件进行数据采集，该设备机动性强，数据采集速度快，能够得到各类零部件表面点云数据。实施例中，采集前将被测废旧零部件的表面喷涂白色显影剂，以增强效果。采集的结果如图3。1) Obtain the surface point cloud data model of waste parts. See Figure 2. In the figure, the ATOS optical three-dimensional scanner is used to collect data on waste parts. This equipment has strong mobility and fast data collection speed, and can obtain surface point cloud data of various parts. In the embodiment, the surface of the waste parts to be tested is sprayed with a white developer before collection to enhance the effect. The collected results are shown in Figure 3.
实施例中，优选地，对图3所示的采集结果进行了预处理(过滤、消除噪声、数据精简等)，获得了如图4所示的废旧零部件表面点云数据模型，可以将其作为后续步骤的处理对象。In the embodiment, preferably, the collection result shown in Figure 3 is preprocessed (filtering, noise elimination, data reduction, etc.), and the surface point cloud data model of the waste parts and components as shown in Figure 4 is obtained, which can be as the processing object for subsequent steps.
2)获取所述废旧零部件的原始CAD模型(未磨损的)。本实施例中，所述废旧零部件的原始CAD模型是能够直接获取的，即厂商或设计人员保存了该原始CAD模型。在原始CAD模型丢失的情况下，参见实施例2公开的获取原始CAD模型的方法。2) Obtain the original CAD model (unworn) of the scrap parts. In this embodiment, the original CAD model of the waste parts can be obtained directly, that is, the original CAD model is saved by the manufacturer or designer. In the case that the original CAD model is lost, refer to the method for obtaining the original CAD model disclosed in Embodiment 2.
3)参见图7，将步骤1)的表面点云数据模型与步骤2)的原始CAD模型配准。即，将两个模型统一在(放在)同一个坐标系中，能够比较出两个模型的差异，获知所述表面点云数据模型的缺失部分以及废旧零部件的最大损伤深度。3) Referring to Fig. 7, the surface point cloud data model of step 1) is registered with the original CAD model of step 2). That is, by unifying (putting) the two models in the same coordinate system, the difference between the two models can be compared, and the missing part of the surface point cloud data model and the maximum damage depth of the scrap parts can be known.
4)根据步骤3)的配准的结果，获得所述废旧零部件的最大损伤深度。实施例中，所述废旧零部件是磨损的部件，有若干处磨损位置。图10示出了该部件其中一处磨损位置，其中虚线代表零部件的原始轮廓，圆圈点表示原始轮廓上的点，实线表示零件损伤后的轮廓，十字叉表示损伤轮廓上的点，其磨损深度为H。该部件中，磨损最深的一处磨损位置的磨损深度为下文提到的最大损伤深度。4) According to the result of the registration in step 3), the maximum damage depth of the scrap parts is obtained. In an embodiment, the used component is a worn component with several worn locations. Figure 10 shows one of the worn parts of the component, where the dotted line represents the original contour of the component, the circle point represents the point on the original contour, the solid line represents the contour of the part after damage, and the cross represents the point on the damaged contour. The wear depth is H. In this component, the wear depth of the most worn location is the maximum damage depth mentioned below.
5)若步骤4)获得的最大损伤深度低于阈值(小于整体尺寸的5％)，以所述最大损伤深度为进给量，对所述废旧零部件进行减式修复。即将废旧零部件作为毛坯，以给定的进给量对其进行切削加工。5) If the maximum damage depth obtained in step 4) is lower than the threshold value (less than 5% of the overall size), use the maximum damage depth as the feed rate to perform subtractive repair on the waste parts. That is, waste parts are used as blanks, and they are cut at a given feed rate.
若步骤4)获得的最大损伤深度高于阈值(大于整体尺寸的5％)，对所述废旧零部件进行加式修复。即通过在磨损位置添加材料的方法，来修复所述废旧零部件。If the maximum damage depth obtained in step 4) is higher than the threshold value (greater than 5% of the overall size), an additive repair is performed on the waste parts. That is, the waste parts are repaired by adding material at the worn position.
实施例2Example 2
废旧零部件的修复过程中，可能会遇到零部件原始CAD文件丢失的情况，此时需要根据采集到的表面点云数据中的残余信息重构其原始CAD模型。与传统逆向建模不同，在重构废旧零部件的原始CAD模型时，由于采集到的点云数据中包含零部件的损伤区域，这部分点云数据与其它完好区域的点云不连续，不能反映零部件的原始表面形貌，故在逆向重构时应避免使用损伤区域的点云数据。During the repair process of waste parts, the original CAD file of the part may be lost. At this time, the original CAD model needs to be reconstructed according to the residual information in the collected surface point cloud data. Different from traditional reverse modeling, when reconstructing the original CAD model of waste parts, because the collected point cloud data contains the damage area of the part, this part of the point cloud data is not continuous with the point cloud of other intact areas, and cannot Reflect the original surface topography of the parts, so the point cloud data of the damaged area should be avoided during reverse reconstruction.
本实施例的主要步骤同实施例1，只是，需要通过以下步骤获取步骤2)所述的原始CAD模型。The main steps of this embodiment are the same as those of Embodiment 1, except that the original CAD model described in step 2) needs to be obtained through the following steps.
实施例中，通过计算点云数据中各点的高斯曲率，提取出点云数据中曲率突变的点，根据这些点构造一条近似的损伤区域边界线，并检查逆向建模中重构的特征轮廓线是否经过损伤区域，确保建模精度。其主要步骤为：In the embodiment, by calculating the Gaussian curvature of each point in the point cloud data, the points where the curvature changes suddenly in the point cloud data are extracted, and an approximate boundary line of the damaged area is constructed based on these points, and the reconstructed feature contour in the reverse modeling is checked Whether the line passes through the damage area to ensure the modeling accuracy. Its main steps are:
1)估算点云曲率1) Estimate point cloud curvature
点云曲率估算的方法很多，常用的方法如：抛物面拟合法，3DShepard曲面拟合法，Gauss–Bonnet法等，由于抛物面拟合法对含噪声点云数据的处理较其他方法更为准确，故采用抛物面拟合法来估算点云网格顶点处的曲率，该方法用一个二阶的解析曲面来逼近给定点及其邻域内的点，用于拟合的二阶曲面表达式由公式(1)所示。There are many methods for point cloud curvature estimation, commonly used methods such as: paraboloid fitting method, 3D Shepard surface fitting method, Gauss–Bonnet method, etc., because paraboloid fitting method is more accurate in processing noise-containing point cloud data than other methods, so paraboloid is used The fitting method is used to estimate the curvature at the vertices of the point cloud grid. This method uses a second-order analytical surface to approximate a given point and the points in its neighborhood. The second-order surface expression for fitting is shown in formula (1) .
z＝f(x,y)＝a0+a1x+a2y+a3xy+a4x2+a5y2 (1)z=f(x,y)=a0 +a1 x+a2 y+a3 xy+a4 x2 +a5 y2 (1)
对点云数据中的某一数据点pi，取该点的k-邻域组成局部点云，对该局部点云内所有的点(xj，yj，zj)，按式(2)做最小二乘拟合，即求解：For a certain data point pi in the point cloud data, take the k-neighborhood of the point to form a local point cloud, and for all points (xj , yj , zj ) in the local point cloud, according to formula (2 ) to do the least squares fitting, that is, to solve:
求得曲面方程系数后，将曲面方程(1)改写为参数方程形式，如公式(3)所示。After obtaining the surface equation coefficients, the surface equation (1) is rewritten into a parametric equation form, as shown in formula (3).
分别求出r(x,y)对x，y，xx，yy，xy的偏微分，记为rx，ry，rxx，ryy，rxy，曲面的单位法向量为则曲面的第一基本形式参数E＝rx·rx，F＝rx·ry，G＝ry·ry，曲面的第二基本形式参数L＝rxx·n，M＝rxy·n，N＝ryy·n。代入曲面高斯曲率的计算公式(4)，即可求出各点处的高斯曲率。Calculate the partial differential of r(x, y) with respect to x, y, xx, yy, xy, denoted as rx , ry , rxx , ryy , rxy , and the unit normal vector of the surface is Then the first basic form parameter of the curved surface is E=rx ·rx , F=rx ·ry , G=ryy ·ry , the second basic form parameter of the curved surface is L=rxx ·n, M=rxy n, N=ryy n. Substituting the calculation formula (4) of the Gaussian curvature of the surface, the Gaussian curvature at each point can be obtained.
2)划分损伤区域边界2) Divide the boundary of the damaged area
根据上步计算得到的各点高斯曲率值Ki，通过设定的阈值Ke，即当Ki＞Ke时，则判断该点为曲率突变点，最后由曲率突变点拟合出损伤区域边界线，将点云数据划分为损伤区域与完好区域，如图5所示，并将损伤区域边界内的点云数据存入集合N中。According to the Gaussian curvature value Ki of each point calculated in the previous step, through the set threshold Ke , that is, when Ki >Ke , the point is judged to be a curvature mutation point, and finally the damage area is fitted by the curvature mutation point The boundary line divides the point cloud data into damaged areas and intact areas, as shown in Figure 5, and stores the point cloud data within the boundary of the damaged area into the set N.
3)重构原始CAD模型3) Reconstruct the original CAD model
通过“点—线—面—体”的建模思路来实现废旧零部件原始CAD模型的重构，如图6所示，首先在点云模型上构造一组截面轮廓线，通过检验各截面轮廓线上是否有集合N内的点，找出并删除经过点云损伤区域的截面轮廓线，最终得到点云完好区域的特征轮廓线。特征轮廓线经过拉伸、扫掠、蒙皮及裁剪等操作得到零部件的曲面模型，最后对曲面模型加厚或实体化即得到原始CAD模型。The reconstruction of the original CAD model of waste parts is realized through the modeling idea of "point-line-surface-body". As shown in Figure 6, a set of section contour lines is firstly constructed on the point cloud model, and each section contour is checked Whether there are points in the set N on the line, find out and delete the cross-sectional contour line passing through the damaged area of the point cloud, and finally obtain the feature contour line of the intact area of the point cloud. The surface model of the component is obtained by stretching, sweeping, skinning and cutting operations on the feature contour line, and finally the original CAD model is obtained by thickening or solidifying the surface model.
实施例3Example 3
本实施例主要步骤同实施例1，进一步地，公开一种适用于步骤3)的改进配准方法。The main steps of this embodiment are the same as those of Embodiment 1, and further, an improved registration method suitable for step 3) is disclosed.
值得说明的是废旧零部件的表面点云数据模型与原始CAD模型的配准是得到最大损伤深度的关键，两模型间的配准通常需要经过预配准和精配准两个步骤，预配准是将两个模型大体调整到正确的位置，为精确配准提供良好初值，提高精确配准的效率，预配准的方法主要有主成分分析法，力矩主轴法，三点对齐法等。精确配准是在预配准的基础上进一步校正两模型的位置，使两者之间的差异最小。精确配准算法中以迭代最近点(Iterative Clostest Point，ICP)算法最为成熟，该算法的实质是基于最小二乘法的最优匹配方法，算法重复“寻找对应点---对应点之间最优刚体变换”的迭代过程，直到满足设定的收敛准则，其变换关系式与收敛准则，如公式(5)，公式(6)所示。It is worth noting that the registration of the surface point cloud data model of waste parts and the original CAD model is the key to obtain the maximum damage depth. The registration between the two models usually requires two steps of pre-registration and fine registration. Pre-registration Accurate is to roughly adjust the two models to the correct position, provide a good initial value for accurate registration, and improve the efficiency of accurate registration. The methods of pre-registration mainly include principal component analysis, moment axis method, three-point alignment method, etc. . Accurate registration is to further correct the positions of the two models on the basis of pre-registration to minimize the difference between the two models. Among the precise registration algorithms, the Iterative Clostest Point (ICP) algorithm is the most mature. The essence of this algorithm is the optimal matching method based on the least squares method. The iterative process of "rigid body transformation" until the set convergence criterion is met, and the transformation relation and convergence criterion are shown in formula (5) and formula (6).
Qj＝RPi+T (5)Qj =RPi +T (5)
ε＝Σ||qj-(Rpi+T)||2→min (6)ε＝Σ||qj -(Rpi +T)||2 → min (6)
其中Pi和Qi为2个模型数据点集，R为旋转变换矩阵，T为平移变换矩阵，pi为模型Pi中的点，qi为模型Qi中的点，ε最小时满足收敛准则。Among them, Pi and Qi are two model data point sets, R is the rotation transformation matrix, T is the translation transformation matrix, pi is the point in the model Pi , qi is the point in the model Qi , and ε satisfies the minimum Convergence criterion.
但在再制造实际应用中，由于废旧零部件表面存在局部损伤，表面点云数据模型与原始CAD模型相比有一定的差异，若采用传统ICP算法对两模型实施最佳拟合，模型上所有的点都将参与配准运算，则损伤区域的误差会被均匀化，因而得不到准确的最大深度值。现通过图8来说明该问题，图8(a)为某废旧零部件原始CAD模型(粗虚线)与表面点云数据模型(粗实线)配准前的情况，根据该零部件的服役状况得知，零部件损伤集中在上端(圆圈区域内)，其他部位则没有损伤或损伤极小；图8(b)为采用传统ICP算法配准的结果，由于损伤区域的误差分摊给了损伤小的区域，使得原本没有损伤或损伤极小的区域出现了误差；图8(c)则为理想的配准效果。However, in the actual application of remanufacturing, due to the local damage on the surface of waste parts, the surface point cloud data model is different from the original CAD model. If the traditional ICP algorithm is used to perform the best fitting of the two models, all All the points will participate in the registration operation, and the error of the damaged area will be uniformed, so the accurate maximum depth value cannot be obtained. Figure 8 is now used to illustrate this problem. Figure 8(a) shows the situation before the registration of the original CAD model (thick dotted line) and the surface point cloud data model (thick solid line) of a waste part. According to the service status of the part It is known that the damage of the parts is concentrated on the upper end (in the circle area), and the other parts have no damage or the damage is very small; Figure 8(b) is the result of registration using the traditional ICP algorithm, because the error of the damage area is allocated to the small damage , resulting in errors in areas with no damage or minimal damage; Figure 8(c) shows the ideal registration effect.
为此本实施例的解决思路是：在传统ICP配准方法中增加对两模型对应点的筛选过程，通过检查对应点的距离及方向向量夹角是否在设定的范围内，来判断对应点是否为损伤区域的点，若判断为损伤区域的点，则剔除该组对应点并重新生成对应点集，再进行配准运算，改进后的配准算法流程图如图9所示。For this reason, the solution idea of this embodiment is: in the traditional ICP registration method, increase the screening process for the corresponding points of the two models, and judge the corresponding points by checking whether the distance between the corresponding points and the angle between the direction vectors are within the set range Whether it is a point in the damaged area, if it is judged to be a point in the damaged area, then the group of corresponding points is eliminated and the corresponding point set is regenerated, and then the registration operation is performed. The improved registration algorithm flow chart is shown in Figure 9.
其主要步骤为：Its main steps are:
1)预配准：采用三点对齐的方法，以原始CAD模型点集Q为参考基准，对表面点云数据模型点集P进行预配准初始变换，原始CAD模型点集Q及预配准后的表面点云数据模型点集P0可表示为：1) Pre-registration: Using the method of three-point alignment, the point set P of the surface point cloud data model is pre-registered for initial transformation with the original CAD model point set Q as the reference, and the original CAD model point set Q and pre-registration The point setP0 of the surface point cloud data model can be expressed as:
P0＝{pi0|pi0∈R3，i＝1，2，…n}，Q0＝{qj0|qj0∈R3，j＝1，2，…m}。P0 ={pi0 |pi0 ∈R3 , i=1, 2,...n}, Q0 ={qj0 |qj0 ∈R3 , j=1, 2,...m}.
2)寻找对应点：对表面点云数据模型点集Pk中的任意一点pik，寻找pik到原始模型点集Q中距离最近的点，记原始模型点集Q中与pik距离最近的点为qik，组成对应点集合Qk＝{qik|qik∈R3，i＝1，2，…n}，距离计算公式为dik＝||pik-qik||→min，k为迭代次数。2) Find the corresponding point: For any point pik in the point set Pk of the surface point cloud data model, find the point with the closest distance from pik to the original model point set Q, and remember that the distance between the original model point set Q and pik is the closest The point is qik , which constitutes the corresponding point set Qk ={qik |qik ∈R3 , i=1, 2,...n}, and the distance calculation formula is dik =||pik -qik ||→ min, k is the number of iterations.
3)求解变换矩阵：对寻找得到的对应点集Pk与Qk，采用最优化解析方法计算Σ||Rkpik+Tk-qik||2→min，求得第k次迭代时的旋转变换矩阵Rk和平移变换矩阵Tk。3) Solve the transformation matrix: For the found corresponding point sets Pk and Qk , use the optimal analytical method to calculate Σ||Rk pik +Tk -qik ||2 →min, and obtain the kth iteration Rotation transformation matrix Rk and translation transformation matrix Tk when .
4)更新模型间相对位置：用步骤3中得到的变换矩阵，对表面点云数据模型点集P进行旋转与平移变换，得到表面点云数据模型新的位置，即Pk+1＝RkPk+Tk。4) Update the relative position between models: use the transformation matrix obtained in step 3 to perform rotation and translation transformation on the point set P of the surface point cloud data model, and obtain the new position of the surface point cloud data model, that is, Pk+1 = Rk Pk + Tk .
5)检查对应点距离及方向一致性：检查对应点间距离是否小于设定的阈值de，即dik＜de，并计算各组对应点的方向向量，检查其夹角θ是否小于设定的阈值θe，即θ＜θe。5) Check the distance and direction consistency of corresponding points: check whether the distance between corresponding points is less than the set threshold de , that is, dik <de , and calculate the direction vector of each group of corresponding points, and check whether the included angle θ is less than The set threshold θe , that is, θ<θe .
6)剔除损伤区域对应点：若对应点间距离dik＞de或θik＞θe，则判断对应点为损伤区域点，将损伤区域点剔除并生成新的表面点云数据模型点集P’与原始CAD模型点集Q’。6) Eliminate the corresponding points in the damaged area: if the distance between corresponding points dik >de or θik >θe , then judge that the corresponding point is a damaged area point, remove the damaged area point and generate a new surface point cloud data model point set P' and original CAD model point set Q'.
7)迭代终止判定：对应点间平均距离小于给定的阈值，即其中则迭代终止。7) Iteration termination judgment: the average distance between corresponding points is less than a given threshold, that is in then the iteration terminates.
实施例4Example 4
本实施例应用了实施例1、2或3所述的方法，对磨损的滚齿机挂轮架进行修复处理。In this embodiment, the method described in Embodiment 1, 2 or 3 is applied to repair the worn rack of the gear hobbing machine.
滚齿机挂轮架是滚齿机的关键零部件之一，用于安装差动齿轮，实现直齿加工传动链与斜齿加工传动链间的切换。挂轮架磨损后，将直接影响差动挂轮间的啮合，最终影响斜齿轮的加工精度。图2(a)为从某废旧滚齿机上拆卸得到的挂轮架，挂轮架的弧形槽区域出现明显磨损，其他部位状况良好，现对其实施再制造修复。Gear hobbing machine hanger frame is one of the key parts of gear hobbing machine, which is used to install differential gear and realize the switching between straight gear processing transmission chain and helical gear processing transmission chain. After the hanger frame is worn out, it will directly affect the meshing between the differential hanger wheels, and finally affect the machining accuracy of the helical gear. Figure 2(a) shows the hanger frame disassembled from a waste gear hobbing machine. The arc groove area of the hanger frame is obviously worn, and other parts are in good condition. Now it is remanufactured and repaired.
对挂轮架部件拆卸清洗后，在其表面喷涂白色显影剂以增强扫描效果，利用ATOS光学三维扫描仪采集表面点云数据，采集过程如图2所示。After disassembling and cleaning the hanger frame parts, spray white developer on its surface to enhance the scanning effect, and use the ATOS optical 3D scanner to collect surface point cloud data. The collection process is shown in Figure 2.
由于挂轮架的原始CAD模型已丢失，需要根据扫描点云数据重构其原始CAD模型，重构过程见图5和图6。根据配准结果分析得知，挂轮架损伤区域集中且磨损较深，大于整体尺寸的5％，故选择加式修复方案，加式修复过程的如图7(最下部分)所示，将挂轮架原始CAD模型与表面点云数据模型按实施例3所述的配准方法进行配准，再对配准后的两模型实施布尔操作，得到缺损部位的模型。最后，对缺损部位模型进行切片，生成缺损部位模型截面轮廓数据，模拟激光涂覆路径。Since the original CAD model of the wheel frame has been lost, it is necessary to reconstruct the original CAD model based on the scanned point cloud data. The reconstruction process is shown in Figure 5 and Figure 6. According to the analysis of the registration results, it is known that the damage area of the hanging wheel frame is concentrated and the wear is deep, which is greater than 5% of the overall size. Therefore, the additive repair scheme is selected. The additive repair process is shown in Figure 7 (the bottom part). The original CAD model of the wheel frame and the surface point cloud data model are registered according to the registration method described in Example 3, and then the Boolean operation is performed on the two registered models to obtain the model of the defect part. Finally, slice the defect part model to generate the cross-sectional profile data of the defect part model and simulate the laser coating path.
实施例5Example 5
本实施例应用了实施例1、2或3所述的方法，对磨损的锤锻模具进行修复处理。In this embodiment, the method described in Embodiment 1, 2 or 3 is applied to repair the worn hammer forging die.
锤锻模具在使用过程中，模腔内部承受巨大的冲击载荷，型腔局部产生塑性变形，同时，坯料金属与型腔表面产生剧烈摩擦，型腔表面会出现磨损甚至剥落。对失效锤锻模具进行再制造修复，能够延长模具使用寿命，降低生产成本。图11(a)所示为报废的汽轮机叶片锤锻模具，现对其开展再制造修复。During the use of the hammer forging die, the inside of the die cavity is subjected to a huge impact load, and the cavity is partially plastically deformed. At the same time, the blank metal and the surface of the cavity generate severe friction, and the surface of the cavity will be worn or even peeled off. Remanufacturing and repairing the failed hammer forging die can prolong the service life of the die and reduce the production cost. Figure 11(a) shows the scrapped steam turbine blade hammer forging die, which is now being remanufactured and repaired.
模具表面点云数据的采集过程如图11所示，因模具上表面为主要工作面，故只对模具上表面进行数据采集，提高修复效率。The acquisition process of point cloud data on the mold surface is shown in Figure 11. Since the upper surface of the mold is the main working surface, only the upper surface of the mold is collected to improve the repair efficiency.
由于模具损伤区域大，磨损较浅，小于整体尺寸的5％，且加工余量充足，故选择减式修复方案，将模具上表面按照原始设计形状重新铣形。减式修复过程的如图12所示，将模具原始CAD模型与表面点云数据模型按实施例3描述的配准方法进行配准，图13是按照传统ICP算法配准与改进ICP算法配准的结果对比，通过测量分析点云模型上同一点，分别在这两种配准方法下的磨损量，如表1所示。Due to the large damage area of the mold, shallow wear, less than 5% of the overall size, and sufficient machining allowance, the subtractive repair plan was chosen to re-mill the upper surface of the mold according to the original design shape. The subtractive repair process is shown in Figure 12. The original CAD model of the mold and the surface point cloud data model are registered according to the registration method described in Example 3. Figure 13 is based on the traditional ICP algorithm registration and the improved ICP algorithm registration. Comparison of the results, by measuring and analyzing the same point on the point cloud model, the amount of wear under these two registration methods, as shown in Table 1.
表1 失效部位磨损量的测量Table 1 Measurement of the amount of wear at the failure site
表1的结果表明，在使用传统ICP算法配准时，平面区域的磨损量在0.49～0.82mm，使用改进ICP算法配准时，平面区域的磨损量为0.02mm左右，而该模具平面区域几乎不存在磨损，故改进ICP配准结果更为精确。配准后测量两模型对应点间的距离，得出对应点间最大距离为4.53mm，即模具的最大磨损深度H为4.53mm，则减式修复加工余量取4.53mm即可。最后，由原始CAD模型及减式修复加工余量值，生成机械加工刀具路径。The results in Table 1 show that when the traditional ICP algorithm is used for registration, the wear amount of the plane area is 0.49-0.82mm, and when the improved ICP algorithm is used for registration, the wear amount of the plane area is about 0.02mm, and the mold plane area is almost non-existent wear, so the improved ICP registration results are more accurate. After registration, the distance between the corresponding points of the two models is measured, and the maximum distance between the corresponding points is 4.53mm, that is, the maximum wear depth H of the mold is 4.53mm, so the subtractive repair machining allowance is 4.53mm. Finally, the machining tool path is generated from the original CAD model and the subtractive repair machining allowance value.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410691799.3ACN104484507B (en) | 2014-11-26 | 2014-11-26 | A kind of spare parts remanufacture method based on reverse-engineering |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410691799.3ACN104484507B (en) | 2014-11-26 | 2014-11-26 | A kind of spare parts remanufacture method based on reverse-engineering |
| Publication Number | Publication Date |
|---|---|
| CN104484507Atrue CN104484507A (en) | 2015-04-01 |
| CN104484507B CN104484507B (en) | 2019-01-18 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410691799.3AActiveCN104484507B (en) | 2014-11-26 | 2014-11-26 | A kind of spare parts remanufacture method based on reverse-engineering |
| Country | Link |
|---|---|
| CN (1) | CN104484507B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105260989A (en)* | 2015-09-10 | 2016-01-20 | 安徽工业大学 | Method for restoring three-dimensional image based on multi-image registration |
| CN105499904A (en)* | 2016-01-04 | 2016-04-20 | 湘潭大学 | Part repairing device based on additive and decreasing manufacturing and application method of part repairing device |
| CN106339543A (en)* | 2016-08-26 | 2017-01-18 | 云南能投水电新能源技术工程有限公司 | Construction method for hydraulic turbine flow passage component morphology repairing model |
| CN107019922A (en)* | 2016-01-31 | 2017-08-08 | 浙江智源办公设备制造有限公司 | A kind of method for maintaining of toy |
| CN107134012A (en)* | 2017-05-03 | 2017-09-05 | 金陵科技学院 | The method that former CAD model is accurately changed based on Boolean calculation and reverse-engineering |
| CN109345523A (en)* | 2018-09-21 | 2019-02-15 | 中国科学院苏州生物医学工程技术研究所 | Surface defect detection and 3D modeling methods |
| CN110276790A (en)* | 2019-06-28 | 2019-09-24 | 易思维（杭州）科技有限公司 | Point cloud registration method based on shape constraining |
| CN111077844A (en)* | 2019-12-12 | 2020-04-28 | 南京航空航天大学 | Part accurate machining method based on measured data feature guidance |
| CN111590265A (en)* | 2020-05-25 | 2020-08-28 | 哈尔滨鑫润工业有限公司 | Mould repairing method based on laser scanning technology |
| CN111692991A (en)* | 2020-06-02 | 2020-09-22 | 哈尔滨工程大学 | Point cloud data acquisition method for measuring batten bonding surface based on white light interference |
| CN112231850A (en)* | 2020-08-04 | 2021-01-15 | 沈阳工业大学 | A feature reconstruction method of gear failure parts based on RE |
| CN112365500A (en)* | 2021-01-12 | 2021-02-12 | 深兰人工智能芯片研究院（江苏）有限公司 | Contour data completion method and device, electronic equipment and storage medium |
| CN114626166A (en)* | 2022-03-30 | 2022-06-14 | 长安大学 | A method, system, equipment and storage medium for remanufacturing and repairing waste parts |
| CN117058199A (en)* | 2023-08-15 | 2023-11-14 | 厦门大学 | Prediction method for judging removal function of flexible air bag tool based on abrasion of polishing pad |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100702794B1 (en)* | 2005-12-14 | 2007-04-03 | 동부일렉트로닉스 주식회사 | How to correct mask layout through OPC verification |
| CN101634544A (en)* | 2009-09-02 | 2010-01-27 | 郑州辰维科技有限公司 | Water turbine blade blank profile measuring and machining allowance analyzing method |
| CN101927391A (en)* | 2010-08-27 | 2010-12-29 | 大连海事大学 | Method for automatic overlay welding repair of damaged metal parts |
| CN102798362A (en)* | 2012-06-20 | 2012-11-28 | 北京航空航天大学 | Point cloud data-based method for estimating working allowance of casting |
| US20130291908A1 (en)* | 2012-05-01 | 2013-11-07 | General Electric Company | Spray arm assembly for a dishwasher appliance |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100702794B1 (en)* | 2005-12-14 | 2007-04-03 | 동부일렉트로닉스 주식회사 | How to correct mask layout through OPC verification |
| CN101634544A (en)* | 2009-09-02 | 2010-01-27 | 郑州辰维科技有限公司 | Water turbine blade blank profile measuring and machining allowance analyzing method |
| CN101927391A (en)* | 2010-08-27 | 2010-12-29 | 大连海事大学 | Method for automatic overlay welding repair of damaged metal parts |
| US20130291908A1 (en)* | 2012-05-01 | 2013-11-07 | General Electric Company | Spray arm assembly for a dishwasher appliance |
| CN102798362A (en)* | 2012-06-20 | 2012-11-28 | 北京航空航天大学 | Point cloud data-based method for estimating working allowance of casting |
| Title |
|---|
| 李文姬: "曲面重构中散乱点云数据曲率估算算法的研究", 《机械设计与制造》* |
| 江志刚 等: "基于协同效应的再制造工艺技术决策模型及应用", 《机械设计与制造》* |
| 沈灿铎 等: "应用逆向工程技术实现再制造毛坯的建模分析方法研究", 《世界维修大会》* |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105260989B (en)* | 2015-09-10 | 2019-01-29 | 安徽工业大学 | The method of 3-D image is restored based on more figure registrations |
| CN105260989A (en)* | 2015-09-10 | 2016-01-20 | 安徽工业大学 | Method for restoring three-dimensional image based on multi-image registration |
| CN105499904A (en)* | 2016-01-04 | 2016-04-20 | 湘潭大学 | Part repairing device based on additive and decreasing manufacturing and application method of part repairing device |
| CN105499904B (en)* | 2016-01-04 | 2017-12-08 | 湘潭大学 | A kind of part prosthetic device and its application method based on increase and decrease material manufacture |
| CN107019922A (en)* | 2016-01-31 | 2017-08-08 | 浙江智源办公设备制造有限公司 | A kind of method for maintaining of toy |
| CN106339543A (en)* | 2016-08-26 | 2017-01-18 | 云南能投水电新能源技术工程有限公司 | Construction method for hydraulic turbine flow passage component morphology repairing model |
| CN107134012A (en)* | 2017-05-03 | 2017-09-05 | 金陵科技学院 | The method that former CAD model is accurately changed based on Boolean calculation and reverse-engineering |
| CN107134012B (en)* | 2017-05-03 | 2020-02-18 | 金陵科技学院 | A method of accurately modifying the original CAD model based on Boolean operation and reverse engineering |
| CN109345523B (en)* | 2018-09-21 | 2022-08-16 | 中国科学院苏州生物医学工程技术研究所 | Surface defect detection and three-dimensional modeling method |
| CN109345523A (en)* | 2018-09-21 | 2019-02-15 | 中国科学院苏州生物医学工程技术研究所 | Surface defect detection and 3D modeling methods |
| CN110276790A (en)* | 2019-06-28 | 2019-09-24 | 易思维（杭州）科技有限公司 | Point cloud registration method based on shape constraining |
| CN111077844A (en)* | 2019-12-12 | 2020-04-28 | 南京航空航天大学 | Part accurate machining method based on measured data feature guidance |
| CN111077844B (en)* | 2019-12-12 | 2020-09-22 | 南京航空航天大学 | Part accurate machining method based on measured data feature guidance |
| CN111590265A (en)* | 2020-05-25 | 2020-08-28 | 哈尔滨鑫润工业有限公司 | Mould repairing method based on laser scanning technology |
| CN111692991A (en)* | 2020-06-02 | 2020-09-22 | 哈尔滨工程大学 | Point cloud data acquisition method for measuring batten bonding surface based on white light interference |
| CN111692991B (en)* | 2020-06-02 | 2021-09-10 | 哈尔滨工程大学 | Point cloud data acquisition method for measuring batten bonding surface based on white light interference |
| CN112231850A (en)* | 2020-08-04 | 2021-01-15 | 沈阳工业大学 | A feature reconstruction method of gear failure parts based on RE |
| CN112365500A (en)* | 2021-01-12 | 2021-02-12 | 深兰人工智能芯片研究院（江苏）有限公司 | Contour data completion method and device, electronic equipment and storage medium |
| CN112365500B (en)* | 2021-01-12 | 2022-02-08 | 深兰人工智能芯片研究院（江苏）有限公司 | Contour data completion method and device, electronic equipment and storage medium |
| CN114626166A (en)* | 2022-03-30 | 2022-06-14 | 长安大学 | A method, system, equipment and storage medium for remanufacturing and repairing waste parts |
| CN117058199A (en)* | 2023-08-15 | 2023-11-14 | 厦门大学 | Prediction method for judging removal function of flexible air bag tool based on abrasion of polishing pad |
| Publication number | Publication date |
|---|---|
| CN104484507B (en) | 2019-01-18 |
| Publication | Publication Date | Title |
|---|---|---|
| CN104484507A (en) | Part remanufacturing method based on reverse engineering | |
| CN103488832B (en) | A kind of geometry restorative procedure of complex curved surface parts damaged area | |
| Yilmaz et al. | A repair and overhaul methodology for aeroengine components | |
| CN104598675B (en) | A kind of assembly simulation method of the spacecraft parts based on measured data | |
| EP3366883A1 (en) | Methods and system for turbine component repair with additive manufacturing | |
| CN104674210A (en) | Workpiece laser automatic repair method | |
| Wang et al. | Point cloud 3D parent surface reconstruction and weld seam feature extraction for robotic grinding path planning | |
| Zheng et al. | Algorithm for remanufacturing of damaged parts with hybrid 3D printing and machining process | |
| CN105447910B (en) | A kind of aero-engine compressor blade tip defect three-dimensional rebuilding method | |
| CN104439884A (en) | Mould three-dimensional rapid forming and repairing method based on reverse engineering | |
| CN103093065A (en) | Three-dimensional model rebuilding method of aviation engine blade tip missing parts | |
| CN105739440A (en) | Adaptive machining method of wide-chord hollow fan blade | |
| CN103955927A (en) | Fillet weld automatic tracking method based on laser vision | |
| Hascoët et al. | Automated identification of defect geometry for metallic part repair by an additive manufacturing process | |
| Praniewicz et al. | Adaptive geometry transformation and repair for hybrid manufacturing | |
| CN103942837A (en) | Method for direct construction of blade point cloud model section curve based on sequence linear programming | |
| Wu et al. | A review of geometric reconstruction algorithm and repairing methodologies for gas turbine components | |
| Zhang et al. | 3D curvature grinding path planning based on point cloud data | |
| CN109967878A (en) | Laser crack repair system and laser crack repair method | |
| Ghorbani et al. | Accurate registration of point clouds of damaged aeroengine blades | |
| CN105867300A (en) | Reverse remachining method for large forming welded part with complex contour | |
| Hou et al. | An adaptive repair surface modeling approach for worn blades | |
| CN110405259A (en) | Machining system of free-form surface parts based on multi-sensor integrated measurement | |
| Aprilia et al. | Towards automated remanufacturing process with additive manufacturing | |
| CN100369067C (en) | Curve construction method of intercept cloud in 3D scanning system |
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |