Method for generating non-uniform allowance tool path adaptability of large thin-walled parts

Abstract

The large thin-wall part non-uniform allowance tool path adaptability generation method belongs to the field of aerospace part precision manufacturing, and relates to a large thin-wall part non-uniform allowance tool path adaptability generation method. The method first analyzes the elastic deformation of a large thin-wall cabin sample when horizontally installed, and solves the actual offset of the local support feature mounting surface of the cabin under different stations. After generating an ideal tool path by using a CAM software, the tool path generated by the theoretical model is mapped in position and posture, so that the tool path on the support mounting surface is accurately adjusted. The mapped tool path is arrayed and optimized, and the tool path adapted to the non-uniform machining allowance is quickly and accurately generated. The method accurately controls the thickness of material removed at each layer in the actual machining process, reduces the machining error of the support mounting surface of the large thin-wall part caused by elastic deformation rebound under different stations, and improves the overall machining quality and machining precision of the large thin-wall part.

Description

Technical Field

[0001] This invention belongs to the field of precision manufacturing of aerospace parts, and relates to a method for adaptively generating toolpaths for non-uniform allowances in large thin-walled parts. Background Technology

[0002] Large, thin-walled spacecraft cabin components are critical parts that provide the overall configuration of a spacecraft and directly affect its service performance. They consist of a thin-walled cabin body and over a hundred supports mounted on it. The thin-walled cabin body is large in size and has low rigidity, making it a typical large, thin-walled shell component. The external support mounting surfaces provide installation interfaces for external payloads such as space robotic arms and solar panels. Due to the stringent dimensional and positional tolerance requirements between these interfaces, extremely high precision is required for the design dimensions of the external support mounting surfaces. For the high-precision machining of large cabin external supports, traditional modular assembly manufacturing methods are inefficient and struggle to guarantee consistent machining quality. Gantry machining center-based manufacturing methods are costly and limited in their application. To address these manufacturing challenges, robotic in-situ integrated machining offers a feasible solution for the high-precision manufacturing of cabin support mounting surfaces.

[0003] However, due to their large size and low rigidity, spacecraft cabins are prone to overall elastic deformation during machining. Furthermore, discrepancies exist between actual and designed machining allowances due to overall manufacturing tolerances and bracket installation errors, leading to further problems. Machining solely based on theoretical models and uniform design allowances may result in unstable machining processes, and the precision of the machined brackets may fail to meet design standards. Therefore, when performing robotic in-situ integrated machining of such large spacecraft cabins, precise toolpath planning tailored to the local bracket features plays a crucial role in ensuring the final machining accuracy of the bracket mounting surface.

[0004] The patent publication CN116117597B by Feng Changxi et al., entitled "An Online Measurement and Compensation Processing Method for Lightening Grooves in Thin-Walled Spacecraft Bodies," describes an online measurement method for lightening grooves in thin-walled spacecraft bodies. It calculates compensation values ​​based on contour and thickness data, and then processes the grooves using these compensation values, thus improving the automation of the measurement and processing. However, this method primarily focuses on calculating compensation values ​​to ensure the accuracy of the groove thickness, and the method of calling compensation values ​​has limitations when handling multi-layer, multi-pass removal processes involving complex structural features. The paper "A technology framework for robotic profiling of blade edges based on model reconstruction and trajectory replanning," Journal of Manufacturing Processes, 2023, 94, 214-227, by Dazhuang Tian et al., uses blade model reconstruction and trajectory replanning to generate blade edge profiling trajectories suitable for robotic measurement-machining integrated manufacturing. However, this method relies on complex operations such as point cloud reconstruction and registration to ensure model reconstruction accuracy, resulting in a complex and time-consuming calculation process, thus having certain limitations. Summary of the Invention

[0005] This invention addresses the shortcomings of existing technologies by proposing an adaptive toolpath generation method for non-uniform machining allowances in large, thin-walled parts. The method first analyzes the elastic deformation of a large, thin-walled cabin sample during horizontal installation to determine the actual offset of the mounting surface of the local support features under different work positions. Then, after generating an ideal toolpath using CAM software, the toolpath generated from the theoretical model is pose-mapped to ensure precise adjustment of the toolpath on the support mounting surface, compensating for deviations caused by elastic deformation. The spatial position of the blank surface is determined using actual measurement data to ascertain the spatial distribution of the actual machining allowance. Finally, the mapped toolpath is arrayed and optimized to quickly and accurately generate a toolpath adapted to non-uniform machining allowances. This method is suitable for in-situ integrated machining of local support features in large, thin-walled cabins, avoiding complex allowance modeling and ensuring that the machining accuracy of the mounting surface of the local support features in large, thin-walled cabins meets design requirements while considering elastic deformation and manufacturing errors.

[0006] The technical solution adopted in this invention is an adaptive toolpath generation method for non-uniform machining allowances of large thin-walled parts. The method is characterized by analyzing the elastic deformation of a large thin-walled cabin sample during horizontal installation to determine the actual offset of the mounting surface of the local support features of the cabin at different workstations. After generating an ideal toolpath using CAM software, the toolpath generated by the theoretical model is pose-mapped to ensure precise adjustment of the toolpath on the support mounting surface. The spatial position of the blank surface is determined using actual measurement data to identify the spatial distribution of the actual machining allowance. Finally, the mapped toolpath is arrayed and optimized to quickly and accurately generate a toolpath adapted to non-uniform machining allowances. This method is suitable for in-situ integrated machining of local support features of large thin-walled cabins, ensuring that the machining accuracy of the mounting surface of the local support features of large thin-walled cabins meets design requirements while considering elastic deformation and manufacturing errors.

[0007] The specific steps of the method are as follows:

[0008] Step 1: Solve for the actual offset of the local support feature mounting surface of the cabin at different work positions.

[0009] A spacecraft thin-walled hull is a typical thin-walled shell structure made of aluminum alloy plates, on which relatively rigid support features are mounted. During the integrated manufacturing process, the hull undergoes elastic deformation due to clamping forces and gravity. Although the local support components are rigid, their orientation changes within the global coordinate system are caused by the overall elastic deformation of the hull.

[0010] First, finite element analysis is used to solve the deformation of the cabin under clamping force and gravity. Then, a local coordinate system is established on the support mounting surface, and the pose changes of the support features are defined in the local coordinate system. Rotation and translation transformations are then used to accurately describe the spatial pose changes of the support mounting surface. Based on the finite element solution results, singular value decomposition is used to analyze the coordinates of multiple nodes on the mounting surface before and after cabin deformation in order to calculate the offset of the support mounting surface. First, the coordinates of the nodes on the support mounting surface before and after cabin deformation are extracted, and the coordinate difference ΔP is calculated. i :

[0011] ΔP i =P i '-P i (1)

[0012] Among them, P i and P i ' represents the coordinates of the i-th node before and after deformation, respectively.

[0013] Furthermore, the center point of the coordinate difference vector between the nodes of the bracket mounting surface before and after deformation is calculated as follows:

[0014]

[0015] Where n represents the total number of grid nodes on the mounting surface.

[0016] Then, the covariance matrix H is constructed and singular value decomposition is performed.

[0017]

[0018] H=UΣV T (4)

[0019] U and V are the left and right singular vectors obtained through singular value decomposition.

[0020] The rotation and translation transformations of the local coordinate system of the support mounting surface before and after the cabin deformation are calculated as follows:

[0021]

[0022] Among them, O L and O' L These are the coordinates of the origin of the local coordinate system before and after the deformation of the cabin.

[0023] The above calculations can determine the pose transformation of the bracket mounting surface, providing accurate bracket mounting surface offset data for the subsequent pose mapping in the toolpath.

[0024] Step 2: Generation and pose mapping of ideal toolpaths for the local support mounting surface of the cabin.

[0025] To reduce machining errors on the bracket mounting surface caused by elastic deformation of the spacecraft cabin, after generating the ideal toolpath using CAM software, it is necessary to perform pose mapping on the toolpath generated from the theoretical model to ensure accurate adjustment of the toolpath on the bracket mounting surface. The toolpath pose mapping process can be described as follows:

[0026]

[0027] Among them, {CL} actual This represents the actual mounting surface toolpath generated after attitude mapping, cl jG and cl' jG These are the ideal toolpath and the toolpath after pose mapping for the j-th bracket mounting surface in the global coordinate system, respectively. jn and S ja Let $\mathbf{j}$ represent the ideal mounting surface and the actual mounting surface of the $j$-th bracket, respectively, where $j = 1, 2, ..., N$, and $N$ is the total number of brackets. The function $f$ is used to map the ideal toolpath to the actual toolpath.

[0028] Using CAM software, a local machining coordinate system is established on the bracket mounting surface, and an ideal toolpath {cl} is generated within this coordinate system based on the theoretical model. L In this local machining coordinate system, the tool contact point coordinates and tool axis vector of the ideal toolpath are as follows:

[0029]

[0030] Among them, P L This is a dataset of tool contact point coordinates in the local support coordinate system, x i y i z i These are the x, y, and z coordinates of the i-th tool contact point, respectively, and V L V is the tool axis vector in the local support coordinate system. x V y V z These are the tool axis vector components in three directions.

[0031] The rotation and translation components from the local coordinate system to the global coordinate system are denoted as follows: G R L and G D L In the global coordinate system, the tool contact point coordinates and tool axis vector are represented as follows:

[0032]

[0033] Since the support structure is a locally rigid body, by transforming the local machining coordinate system, the pose mapping of the ideal toolpath on the support mounting surface can be achieved in the global coordinate system. The rotation and translation components of the local machining coordinate system caused by the elastic deformation of the entire cabin are R1 and R2, respectively. Trans and D Trans After the entire cabin undergoes elastic deformation, the rotation matrix and translation vector obtained from the global coordinate system to the local coordinate system are as follows:

[0034]

[0035] The mapped toolpath is determined by the tool contact point coordinates P in the global coordinate system. G 'and tool axis vector V G 'Composition.' Furthermore, in planar feature end milling, the tool axis vector remains constant. Therefore, the tool contact point coordinates and tool axis vector after pose mapping are calculated as follows:

[0036]

[0037] Combining formulas (7) to (10), the theoretical toolpath {CL} from the bracket mounting surface can be derived. faceThe actual toolpath {CL'} after offset face The mapping relationship.

[0038]

[0039] The offset of the support mounting surface caused by the overall deformation of the cabin is calculated, and the rotation component R of the local coordinate system is obtained. Trans Translation component D Trans It can generate theoretical toolpaths {CL} based on ideal design models. face The actual toolpath {CL'} on the bracket mounting surface after the overall deformation of the cabin face The mapping.

[0040] Step 3: Solve for the actual spatial position of the feature blank surface of the local support of the cabin.

[0041] Due to overall machining errors in the spacecraft cabin and installation errors in the support structure, there is a deviation between the actual machining allowance and the design allowance of the support features during the in-situ integrated machining process of the cabin. To solve this problem, structured light measurement technology is first used to scan and measure the actual blank surface of the support features to accurately determine the spatial position of the blank surface.

[0042] To match the ideal machining trajectory with the coordinate system of the actual measured point cloud data, the coordinate data Q of the measuring points on the surface of the support blank in the global coordinate system are first... G (x i ,y i ,z i Transform to the actual local machining coordinate system to obtain the coordinate data Q of the measuring point in the local machining coordinate system. L .

[0043] Q L =( G R L ) -1 (Q G - G D L (12)

[0044] Then, the plane equation of the support blank surface in space is defined as follows:

[0045] z = ax + by + c (13)

[0046] Among them, a, b, and c are the parameters of the plane equation of the support blank to be determined.

[0047] In actual measurement operations, the collected measurement point data may be affected by measurement errors. To accurately determine the optimal planar position on the surface of the blank, the least squares method is used to fit and solve the measurement point data. Specifically, for the coordinate data Q of the measurement points on the surface of the support blank... L Construct the following objective function for plane fitting:

[0048]

[0049] Taking partial derivatives with respect to a, b, and c, and setting these partial derivatives to zero, we obtain:

[0050]

[0051] Rewrite the above equation in matrix form:

[0052]

[0053] By solving the above matrix equations, the optimal values ​​of a, b, and c can be obtained, thereby determining the optimal position of the actual blank surface of the support. For the actual support features, the envelope space between the target surface and the blank in the local support coordinate system is the actual machining allowance space. This method ensures the accuracy of the blank surface positioning and provides a reliable basis for subsequent toolpath planning.

[0054] Step 4: Adaptive generation of toolpath for non-uniform allowance of local support features in the cabin.

[0055] Furthermore, to address the unevenly distributed machining allowance on the spacecraft cabin support, an adaptive toolpath generation strategy was developed to control the removal thickness and ensure that the robotic machining system can accurately remove the actual allowance from the support mounting surface. The specific steps are as follows:

[0056] 1) Construct the coordinate array of the knife contact points on the mounting surface of the bracket.

[0057] Calculate the surface measurement data Q of the bracket blank L The maximum distance H between the actual mounting surface after pose mapping and the actual mounting surface max And this is used as the maximum spacing for toolpath array replication. Since the support blank surface is a plane and the tool axis vector is parallel to the Z-axis of the local machining coordinate system, H max This corresponds to the maximum Z-coordinate value in the measurement data of the blank surface.

[0058] H max =max(z) i (17)

[0059] Then, using the single-layer removal thickness h of the bracket mounting surface finishing as the step size, the coordinates of the tool contact points on the bracket mounting surface are arrayed along the Z direction of the local machining coordinate system.

[0060] P ij =P0+i·(0 0 h) T (18)

[0061] Where i represents the sequence number of each layer in the toolpath array, i = 1, 2, ..., n, and j represents the machining layer number, j = 1, 2, ..., H. max / h. P ij This represents the coordinates of the knife contact point after the array is created.

[0062] 2) Filter the valid tool contact coordinates from the tool contact coordinates after arraying.

[0063] For the coordinates P of the tool contact point after each array ij The positional relationship of the tool contact point relative to the blank surface B and the target surface T must be calculated to determine whether the tool contact point is located between these two surfaces. The tool contact point position is determined by the following formula:

[0064] f effective =(B(P) ij )≤0)∩(T(P ij )≥0) (19)

[0065] Among them, B(P) ij ) indicates the coordinates of the tool contact point P ij The directed distance to the blank surface B (if B (P) ij If )≤0, then the knife contact point P ij Located on or below the surface of the blank), T(P) ij ) indicates the coordinates of the tool contact point P ij The directed distance to the target surface T (if T(P) ij If )≥0, then the knife contact point P ij Located on or above the target surface.

[0066] Based on the judgment of formula (19), the effective tool positions can be filtered into the set {CL}. body middle:

[0067] {CL} body ={P ij |f effective} (20)

[0068] This ensures that the final toolpath only includes those tool contact points located between the workpiece surface and the target surface, thereby improving machining accuracy and efficiency.

[0069] 3) Tool contact connection and toolpath generation

[0070] The effective tool contact coordinates of each machining layer are sorted in ascending order and then connected sequentially. Appropriate transition points are then inserted between layers to form a continuous toolpath that can accommodate the actual non-uniform machining allowance of the support. Through these operations, a toolpath that completely covers the non-uniform allowance can be generated.

[0071] The significant effect and benefit of this invention is that it addresses the problem of discrepancies between the actual machining allowance and the pre-designed allowance caused by overall cabin manufacturing errors and bracket feature installation errors during the integrated machining of local support features in large thin-walled cabins. This method first analyzes the elastic deformation of the large thin-walled cabin sample during horizontal installation to determine the actual offset of the local support feature mounting surface at different workstations. After generating an ideal toolpath using CAM software, the toolpath generated from the theoretical model is pose-mapped to ensure precise adjustment of the toolpath on the bracket mounting surface. The mapped toolpath is then arrayed and optimized to quickly and accurately generate toolpaths that adapt to non-uniform machining allowances. This method accurately controls the thickness of material removed at each layer during actual machining, reduces machining errors on the bracket mounting surface of large thin-walled parts caused by elastic deformation rebound at different workstations, and improves machining stability and the overall machining quality and accuracy of large thin-walled parts. Attached Figure Description

[0072] Figure 1 —Overall flowchart of the method.

[0073] Figure 2 —Scene diagram of in-situ integrated robotic machining of large thin-walled cabin parts. In the diagram: 1—Large thin-walled cabin, 2—Mobile robot measurement and machining integrated system, 3—Support device, 4—Rotating fixture.

[0074] Figure 3 —Diagram of a mobile robot measurement and processing integrated system. In the diagram: 5—Industrial robot, 6—High-speed electric spindle, 7—Line laser scanner, 8—Mobile platform.

[0075] Figure 4 —A typical bracket and its actual machining allowance distribution diagram.

[0076] Figure 5 —A typical bracket mounting surface toolpath result diagram generated based on theoretical design allowance.

[0077] Figure 6 —A typical toolpath result diagram of the bracket mounting surface generated based on the actual machining allowance.

[0078] Figure 7 —Generate a typical bracket mounting surface deviation distribution diagram based on the theoretical toolpath machining obtained from the design model.

[0079] Figure 8 —A typical bracket mounting surface deviation distribution diagram obtained by machining the tool path after pose mapping.

[0080] Figure 9 —Comparison chart of absolute values ​​of the dimensional error of the mounting surface of the bracket obtained by different processing methods. Detailed Implementation

[0081] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the specific implementation methods of the present invention will be described in detail below with reference to the technical solutions and accompanying drawings.

[0082] Due to their large size and low rigidity, spacecraft cabins are prone to overall elastic deformation during machining. Furthermore, discrepancies exist between actual and designed machining allowances due to overall manufacturing tolerances and bracket installation errors, leading to further problems. Machining solely based on theoretical models and uniform design allowances may result in unstable machining processes, and the precision of the machined brackets may not meet design standards. To overcome these issues, an adaptive toolpath generation method for non-uniform allowances in large, thin-walled parts has been invented. The overall process is shown in the attached figure. Figure 1 As shown.

[0083] Taking the in-situ integrated machining of a 7075 aluminum alloy thin-walled cabin prototype with an outer diameter of 3600mm and a length of 6200mm as an example, the implementation process of this invention is described in detail. (Appendix) Figure 2 In the scenario of in-situ integrated robotic machining of large thin-walled cabin parts, the large thin-walled cabin parts are stably installed on the support device 3. The circumferential rotation and displacement of the cabin parts are realized with the help of the rotating fixture 4. Furthermore, the scanning and machining of the actual blank surface of the local support features of the cabin is completed by the mobile robot measurement and machining integrated system 2.

[0084] First, according to the appendix Figure 2 The actual clamping layout and standard gravity load are shown. Static analysis of the thin-walled cabin is performed using ANSYS software. The material elastic modulus of the large thin-walled cabin is E = 71.7 GPa, Poisson's ratio ν = 0.33, and density ρ = 2.81 g / cm³. 2 The mesh element type is defined as tetrahedral element with a mesh size of 5 mm. Post-processing yields the topological relationships between the finite element mesh, elements, and nodes of the large thin-walled cabin, as well as the coordinate values ​​and deformation displacement vectors of the finite element mesh nodes before and after elastic deformation of the large thin-walled cabin under gravity and clamping. The rotation component R of the local machining coordinate system caused by the elastic deformation of the entire cabin is calculated according to equations (1)-(5). Trans Translation component D Trans .

[0085] Secondly, taking four typical local support features as examples, UG software was used to plan the machining path for the target surfaces of the local support features in the design model cabin. During machining, the spindle speed was 5000 r / min and the feed rate was 200 mm / min. The theoretical toolpath {CL} generated based on the ideal design model was obtained through post-processing. face And according to equations (7)-(11), the theoretical toolpath {CL} generated based on the ideal design model is completed. face The actual toolpath {CL'} on the bracket mounting surface after the overall deformation of the cabin face The mapping.

[0086] Then, using the attached Figure 3 The mobile robot measurement and processing integrated system shown scans and measures the blank surface of a local support feature of a large thin-walled cabin in a horizontal installation state, obtains the point cloud data of the actual blank surface of the local support feature, and converts it to the actual local processing coordinate system. The actual spatial position of the support blank surface is solved according to equations (13)-(16). Based on the spatial position of the support mounting surface and the actual blank surface after pose mapping, the actual processing allowance distribution of four typical local support features is obtained as shown in the appendix. Figure 4 As shown. The machining allowance reserved during the design process for the local support feature is a uniformly distributed 5mm thickness removal amount. However, the actual machining allowance of the local support feature is different from the ideal allowance state. Its thickness distribution is not uniform, and a targeted machining path must be generated for its actual non-uniform allowance.

[0087] Combining the actual tool path and actual blank surface calculation results of the bracket mounting surface obtained from the above process after pose mapping, the machining path of the actual non-uniform allowance of the cabin bracket feature is calculated and generated according to equations (17)-(20). Finally, through post-processing, a machining file that the robot can recognize is output, realizing the adaptive planning of the machining path of the actual non-uniform allowance of the local bracket feature of the large thin-walled cabin part, and ensuring machining accuracy while achieving stable machining of local features.

[0088] To verify the effectiveness of the proposed method in controlling the thickness of single-layer material removal, a single-layer material removal thickness of 2 mm was used. The processing paths planned according to the theoretical design margin and those generated by the proposed method were compared. The processing paths planned by the different methods are shown in the appendix. Figure 5 and attached Figure 6 As shown. (From the appendix) Figure 5It is evident that all machining paths planned directly according to the ideally reserved 5mm thickness machining allowance are three-layer machining paths. This results in insufficient machining path layers. The thickness of the material to be removed is significant between the first machining path and the actual blank surface. Machining along this path leads to noticeable variations in the material removal thickness, with the actual maximum removal thickness even exceeding 5mm. Therefore, the toolpaths for typical bracket mounting surfaces generated based on the theoretical design allowance cannot be directly used in robotic machining systems. (See attached...) Figure 6 It can be seen that the processing path generated by this method can be generated layer by layer with equal thickness between the actual blank surface and the target mounting surface, which can effectively fill all the remaining space and ensure that the thickness of a single layer of material removed during the actual processing does not exceed 2mm, thus effectively ensuring the stable and controllable removal of the actual allowance.

[0089] Meanwhile, to verify the effectiveness of the method in improving machining accuracy, four typical bracket samples were used as examples. The machining results obtained by using toolpaths generated according to the theoretical design model and toolpaths planned by the proposed method were compared. The actual deviation distribution of the mounting surface of the four typical brackets machined by different methods was calculated, and the absolute value of the final dimensional error of the bracket mounting surface was used as the criterion for judging the quality of the machining result. Among them, the distribution of the deviation value of the bracket mounting surface obtained by machining according to the toolpath generated by the theoretical design model is shown in the attached figure. Figure 7 As shown in the attached figure; the toolpaths planned according to the proposed method are used for machining, and the distribution of the deviation values ​​of the bracket mounting surface is shown in the attached figure. Figure 8 As shown. (From the appendix) Figure 7 and attached Figure 8 The deviation distribution results show that, when machining using the toolpath generated by the theoretical design model, the minimum deviation of the four typical bracket mounting surfaces is greater than 0.9 mm, and the maximum deviation can even reach more than 1.6 mm. However, the absolute values ​​of the deviations of the four typical bracket mounting surfaces obtained by machining using the proposed method are all below 0.15 mm, indicating that the proposed method can effectively reduce the machining deviation of the bracket mounting surfaces.

[0090] To further evaluate the machining accuracy of different methods, the absolute value of the final dimensional error of the machining bracket mounting surface was calculated. The absolute value of the final dimensional error is shown in the attached figure. Figure 9 As shown. (Attached) Figure 9 The calculation results show that the absolute value of the final dimensional error of the bracket mounting surface is significantly reduced when the tool path planned by the proposed method is used for machining. Compared with machining directly according to the tool path planned by the theoretical design model, the absolute value of the dimensional error of the four typical bracket samples is reduced by more than 90%.

[0091] The comparative results show that the non-uniform allowance tool path adaptive generation method for large thin-walled parts of the present invention can accurately control the thickness of material removed from each layer during the actual machining process. At the same time, it can reduce the machining error of the mounting surface of the bracket of large thin-walled parts caused by elastic deformation and springback at different work stations, improve the overall machining quality and machining accuracy of large thin-walled parts, and has important guiding significance for the in-situ integrated machining of large thin-walled components in engineering practice.

Claims

1. A method for adaptively generating toolpaths for non-uniform allowances in large, thin-walled parts, characterized in that, This method analyzes the elastic deformation of a large thin-walled cabin sample during horizontal installation to determine the actual offset of the mounting surface of the local support features under different work positions. After generating an ideal toolpath using CAM software, the toolpath generated from the theoretical model is pose-mapped to ensure precise adjustment of the toolpath on the support mounting surface. The spatial position of the blank surface is determined using actual measurement data to identify the spatial distribution of the actual machining allowance. Finally, the mapped toolpath is arrayed and optimized to quickly and accurately generate a toolpath suitable for non-uniform machining allowances. This method is applicable to the in-situ integrated machining of local support features in large thin-walled cabins, ensuring that the machining accuracy of the mounting surface of the local support features meets design requirements while considering elastic deformation and manufacturing errors. The specific steps of the method are as follows: Step 1: Solve for the actual offset of the local support feature mounting surface of the cabin at different work positions; A spacecraft thin-walled cabin is a typical thin-walled shell structure made of aluminum alloy plates, on which a support structure is installed. During the integrated manufacturing process, the cabin will undergo elastic deformation due to clamping forces and gravity. Although the local support components are rigid, they will undergo pose changes in the global coordinate system due to the elastic deformation of the cabin as a whole. First, finite element analysis is used to solve the deformation of the cabin under clamping force and gravity. Then, a local coordinate system is established on the support mounting surface, and the pose changes of the support features are defined in the local coordinate system. Rotation and translation transformations are used to accurately describe the spatial pose changes of the support mounting surface. Based on the finite element solution results, singular value decomposition is used to analyze the coordinates of multiple nodes on the mounting surface before and after cabin deformation in order to calculate the offset of the support mounting surface. First, the coordinates of the nodes on the support mounting surface before and after cabin deformation are extracted, and the difference in node coordinates is calculated. : (1) in, and These are the coordinates of the i-th node before and after deformation, respectively; Furthermore, the center point of the coordinate difference vector between the nodes of the bracket mounting surface before and after deformation is calculated as follows: (2) Where n represents the total number of grid nodes on the mounting surface; Then, the covariance matrix H is constructed and singular value decomposition is performed; (3) (4) U and V are the left singular vector and the right singular vector obtained through singular value decomposition; The rotation and translation transformations of the local coordinate system of the support mounting surface before and after the cabin deformation are calculated as follows: (5) in, and These are the coordinates of the origin of the local coordinate system before and after the deformation of the cabin; The above calculations can determine the pose transformation of the bracket mounting surface, providing accurate bracket mounting surface offset data for the subsequent pose mapping in the toolpath. Step 2: Generation and pose mapping of ideal toolpaths for the local support mounting surface of the cabin; To reduce machining errors on the bracket mounting surface caused by elastic deformation of the spacecraft cabin, after generating the ideal toolpath using CAM software, it is necessary to perform pose mapping on the toolpath generated by the theoretical model to ensure accurate adjustment of the toolpath on the bracket mounting surface. The toolpath pose mapping process is described below: (6) in, This represents the actual toolpath on the mounting surface generated after attitude mapping. and These are the ideal toolpath and the toolpath after pose mapping for the j-th bracket mounting surface in the global coordinate system, respectively. and Let $\mathbf{j}$ represent the ideal mounting surface and the actual mounting surface of the $j$-th bracket, respectively. N is the total number of supports; the f function is used to map the ideal toolpath to the actual toolpath; Using CAM software, a local machining coordinate system is established on the bracket mounting surface, and an ideal toolpath is generated within this coordinate system based on the theoretical model. In this local machining coordinate system, the tool contact point coordinates and tool axis vector of the ideal toolpath are as follows: (7) in, This is a dataset of tool contact point coordinates in the local support coordinate system. , , These are the x, y, and z coordinates of the i-th blade contact point. The tool axis vector in the local support coordinate system. , , These are the tool axis vector components in three directions; The rotation and translation components from the local coordinate system to the global coordinate system are denoted as follows: and In the global coordinate system, the tool contact point coordinates and tool axis vector are represented as follows: (8) Since the support is a locally rigid body, by transforming the local machining coordinate system, the pose mapping of the ideal toolpath on the support mounting surface can be achieved in the global coordinate system; the rotation and translation components of the local machining coordinate system caused by the elastic deformation of the entire cabin are respectively and After the entire cabin undergoes elastic deformation, the rotation matrix and translation vector obtained from the global coordinate system to the local coordinate system are as follows: (9) The mapped toolpath is derived from the tool contact point coordinates in the global coordinate system. and tool axis vector The composition; moreover, in planar feature end milling, the tool axis vector remains unchanged. Therefore, the tool contact point coordinates and tool axis vector after pose mapping are calculated as follows: (10) Combining formulas (7) to (10), the theoretical toolpath from the bracket mounting surface can be derived. Actual toolpath after offset The mapping relationship; (11) The offset of the support mounting surface caused by the overall deformation of the cabin is calculated, and the rotation component of the local coordinate system is obtained. Translation components It can generate theoretical toolpaths based on ideal design models. Actual toolpath to the bracket mounting surface after overall cabin deformation Mapping; Step 3: Solve for the actual spatial position of the feature blank surface of the local support of the cabin; Due to the overall machining error of the spacecraft cabin and the installation error of the bracket, there is a deviation between the actual machining allowance and the design allowance of the bracket features during the in-situ integrated machining of the cabin. To solve this problem, structured light measurement technology is first used to scan and measure the actual blank surface of the bracket features to accurately determine the spatial position of the blank surface. To match the ideal machining trajectory with the coordinate system of the actual measured point cloud data, the coordinate data of the measuring points on the surface of the support blank in the global coordinate system are first... Transform to the actual local machining coordinate system to obtain the coordinate data of the measuring points in the local machining coordinate system. ; (12) Then, the plane equation of the support blank surface in space is defined as follows: (13) Among them, a, b, and c are the parameters of the plane equation of the support blank to be determined; In actual measurement operations, the collected measurement point data will be affected by measurement errors. To accurately determine the optimal planar position on the surface of the blank, the least squares method is used to fit and solve the measurement point data. Specifically, for the coordinate data of the measurement points on the surface of the support blank... Construct the following objective function for plane fitting: (14) Taking partial derivatives with respect to a, b, and c, and setting these partial derivatives to zero, we obtain: (15) Rewrite the above equation in matrix form: (16) By solving the above matrix equations, the optimal values ​​of a, b, and c can be obtained, thereby determining the optimal position of the actual blank surface of the support. For the actual support features, the envelope space between the target surface and the blank in the local support coordinate system is the actual machining allowance space. This method ensures the accuracy of the blank surface positioning and provides a reliable basis for subsequent tool path planning. Step 4: Adaptive generation of toolpaths for non-uniform allowances in local support features of the cabin. Furthermore, to address the unevenly distributed machining allowance on the spacecraft cabin support, an adaptive toolpath generation strategy was developed to control the removal thickness and ensure that the robotic machining system can accurately remove the actual allowance from the support mounting surface. The specific steps are as follows: 1) Construct the coordinate array of the knife contact points on the mounting surface of the bracket; Calculate the surface measurement data of the support blank. The maximum distance between the actual mounting surface after pose mapping and the actual mounting surface This is used as the maximum spacing for toolpath array replication; since the support blank surface is a plane and the tool axis vector is parallel to the Z-axis of the local machining coordinate system, Corresponding to the maximum Z-coordinate value in the measurement data of the blank surface; (17) Then, using the single-layer removal thickness h of the bracket mounting surface finishing as the step size, the coordinates of the tool contact points on the bracket mounting surface are arrayed along the Z direction of the local machining coordinate system; (18) Where i represents the sequence number of each layer in the toolpath array. j represents the processing layer number. ; Represents the coordinates of the tool contact point after arraying; 2) Filter the valid tool contact coordinates from the arrayed tool contact coordinates; For the tool contact point coordinates after each array The positional relationship of the tool contact point relative to the blank surface B and the target surface T must be calculated to determine whether the tool contact point is located between these two surfaces; the position of the tool contact point is determined by the following formula: (19) in, Indicates the coordinates of the tool contact point The directed distance to the surface B of the blank, if Then the knife contact point Located on or below the surface of the blank, Indicates the coordinates of the tool contact point The directed distance to the target surface T, if Then the knife contact point Located on or above the target surface; Based on the judgment of formula (19), the effective tool positions can be filtered into the set. middle: (20) This ensures that the final toolpath only contains those tool contact points located between the workpiece surface and the target surface, thereby improving machining accuracy and efficiency; 3) Tool contact connection and toolpath generation The effective tool contact coordinates of each machining layer are sorted in ascending order and then connected sequentially. Appropriate transition points are then inserted between the layers to form a continuous toolpath that can adapt to the actual non-uniform machining allowance of the support. Through the above operations, a toolpath that can completely cover the non-uniform allowance can be generated.

Images