Rapid preparation method of U-shaped paint spraying shielding mold based on wing curved surface characteristics

Abstract

The invention relates to the technical field of shielding mold rapid preparation, in particular to a wing curved surface feature-based U-shaped paint spraying shielding mold rapid preparation method, which comprises the following steps: S1, obtaining a wing curved surface corresponding model; s2, establishing a fragment index for the model corresponding to the wing curved surface, designing a telescopic groove parameter of the mold for each fragment, reconstructing a thickened U-shaped contour, and outputting a three-dimensional solid model of the mold; s3, quickly solving a thermal field and performing thermal-force coupling warping prediction, if a risk area is detected, performing filling through an implantable grid shape or a grid and entity support, performing cyclic correction until residual warping meets paint spraying fitting tolerance, and generating an output G code; s4, mold printing is completed on fused deposition forming equipment; and S5, deviation calibration is conducted on the printed mold.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of rapid preparation of shielding molds, and in particular to a U-shaped paint shielding mold rapid preparation method based on wing surface characteristics. BACKGROUND

[0002] With the rapid evolution of digital twin manufacturing, composite additive forming and intelligent spraying technology, the assembly-painting integrated tooling of large composite wing has become the focus in the field of aviation manufacturing and maintenance. The complex shape of the wing skin, the dense ribbing and the narrow closed angle have higher requirements for the geometric fitting precision, assembly efficiency and stiffness distribution of the paint shielding mold. The existing shielding means still mainly rely on manual cutting of paper templates or segmented metal hard molds, combined with empirical polishing and repeated trial assembly, which generally has the following shortcomings: On the one hand, traditional manual sizing or two-dimensional expansion method cannot achieve high-precision three-dimensional matching in the wing curvature mutation area, fastener steps and closed angle shadow, resulting in gaps or interference between the mold and the skin, frequent paint leakage and assembly rework.

[0003] On the other hand, the existing additive printing mold adopts fixed wall thickness and single field path planning, without considering the coupling effects of printing thermal shrinkage, spraying air pressure impact and mold self-locking force, which is prone to warping, rib suspension or expansion slot failure after forming, and is difficult to meet the on-site use conditions of one-time insertion-automatic locking.

[0004] In addition, the current point cloud reconstruction and virtual assembly process are disconnected, and scanning-modeling-trial assembly usually requires multiple rounds of manual intervention, lacks real-time correction of multiple physical fields and closed-loop verification of assembly mechanics, and has long cycle, high cost and large quality fluctuation.

[0005] Therefore, there is an urgent need for a scalable U-shaped shielding mold preparation technology that integrates high-density curved surface collection, semantic label driven geometry generation, multi-field coupled path correction and on-site rapid verification, to improve paint quality, shorten tooling iteration cycle and reduce assembly labor intensity. SUMMARY

[0006] The technical problem solved by the present application is to provide a U-shaped paint shielding mold rapid preparation method based on wing surface characteristics, which provides an integrated scalable U-shaped shielding mold preparation technology that can improve paint quality, shorten tooling iteration cycle and reduce assembly labor intensity.

[0007] The technical solution of the present application: A U-shaped paint shielding mold rapid preparation method based on wing surface characteristics, comprising the following steps: S1. Mark the unpainted feature areas on the wing surface and make their origin coaxial with the wing root reference hole. Use a laser scanner to scan the wing surface area at a constant speed in a strip manner, calculate the distance measurement index in real time and collect point cloud data, and scan the missing areas to complete noise removal and continuous patching of hole curvature to obtain the corresponding model of the wing surface. S2. Establish a segmented index for the corresponding model of the wing curved surface, design the expansion groove parameters of the mold for each segment, reconstruct the thickened U-shaped contour, and output the three-dimensional solid model of the mold. S3. Perform Z-axis slicing on the three-dimensional solid model of the mold, quickly solve the thermal field and perform thermo-mechanical coupling warpage prediction. If a risk area is detected, it is filled by implantable mesh or mesh plus solid support. The correction is repeated until the residual warpage meets the paint adhesion tolerance and output G code is generated. S4. Mold printing is completed on a fused deposition modeling (FDM) machine; S5. Perform deviation calibration on the printed mold.

[0008] The beneficial effects of this invention are as follows: This invention utilizes core strategies such as curvature difference-driven segmented indexing, rib fitting baseline reconstruction, and elastic back-thrust of the expansion groove to achieve a fully integrated process of close contact, clearance, and springback self-locking between the mold outer wall and the wing ribs. This allows the mold to automatically avoid rib and fastener protrusions during insertion and automatically lock itself after insertion, effectively avoiding the pain points of traditional hard molds being difficult to insert and paper patterns easily becoming suspended. It ensures that there are no visible gaps or hard-impact interference between the skin surface and the shielding wall.

[0009] Based on a dual-loop closed-loop system of coupled thermal, mechanical, and fluid field path correction and risk grid wall thickness fine-tuning, this invention can predict potential warping sources before additive manufacturing and simultaneously adjust extrusion speed, cooling airflow, and local filling stiffness during the manufacturing process. After manufacturing, targeted compensation is made through wall thickness stiffening or weight reduction. This multi-layered protection system of source prevention, process correction, and result feedback significantly suppresses the warping risk in easily deformable areas such as thin walls and closed angles, ensuring that the self-locking groove of the mold maintains a stable locking force after multiple insertions and removals.

[0010] The process iteration scheme, which combines hierarchical locking of local thermal field re-solution with initial compensation vector of knowledge base, enables thermal solution to be recalculated only for the modified area, while existing results are directly reused for other areas, which greatly shortens the time for secondary or even multiple solutions. At the same time, the system automatically records and reuses the compensation experience of similar mold tasks, which enables new tasks to converge quickly and reduces repeated trial and error and manual parameter tuning. Attached Figure Description

[0011] Figure 1 This is a flowchart illustrating a rapid fabrication method for a U-shaped paint masking mold based on the curved surface features of an airfoil. Figure 2 An isometric side view of the U-shaped paint masking mold prepared according to an embodiment of the present invention; Figure 3 A front view of the U-shaped spray paint masking mold prepared according to an embodiment of the present invention; Figure 4 A left view of the U-shaped spray paint masking mold prepared according to an embodiment of the present invention; Figure 5 A top view of the U-shaped spray paint masking mold prepared according to an embodiment of the present invention; Figure 6 A three-dimensional view of the U-shaped spray paint masking mold prepared according to an embodiment of the present invention. Detailed Implementation

[0012] This invention provides a rapid preparation method for a U-shaped paint masking mold based on the curved surface features of an airfoil, such as... Figure 1 As shown, it includes the following steps: S1. Mark the feature area of ​​the wing surface and make its origin coaxial with the reference hole at the wing root. Use a Handyscan handheld laser scanner to scan the wing surface area in a constant speed strip manner, calculate and collect distance in real time, and fill in the missing area to complete the noise removal and continuous patching of hole curvature to obtain a clean surface model.

[0013] S2. Implement curvature difference on the surface constraint vector to establish a piecewise index, design expansion groove parameters according to the rib fit baseline, reconstruct the thickened U-shaped contour, and output a three-dimensional solid model. S3. Perform Z-axis slicing on the solid model, quickly solve the thermal field and perform thermo-mechanical coupling warpage prediction. If a risk area is detected, back-calculate the extrusion speed and cooling air volume and fill it with a mesh or mesh plus solid support. Cyclic correction is performed until the residual warpage meets the paint bonding tolerance and output G-Code is generated. S4. On the fused deposition modeling equipment, complete nozzle and hot bed calibration, G-code import, temperature-flow preheating, first layer adhesion monitoring and airflow-speed coupled printing. After forming, unload the part after uniform cooling and re-inspect key dimensions. S5. After fitting the mold into the wing root reference pin, advance it at a constant speed. Use the thrust curve, thin film pressure plate and structured light point cloud to triple-check the fit. If the maximum deviation exceeds the limit, backtrack to S2 for revision; otherwise, complete the mold preparation.

[0014] Optionally, S1 includes the following steps: S11, Calibration and Coordinate System Calibration points should be affixed to the characteristic areas of the wing's curved surface, and these points should be affixed randomly. The origin of the calibration points should be fixed coaxially with the wing root reference hole.

[0015] We continued to use the Handyscan handheld laser scanner for optical tracking, solved the homogeneous transformation matrix from the scanner coordinate system to the wing assembly coordinate system in real time, and locked it throughout the scanning process.

[0016] S12, Strip Division and Constant Speed ​​Scanning Based on the spacing between adjacent ribs, the target is divided into several longitudinal strips along the wingspan. Maintaining a constant linear velocity, the flanges of the ribs within each strip are covered, and then the grooves of the ribs in the same strip are scanned back to ensure that the spacing between points within the strip is evenly distributed, avoiding measurement blind spots caused by rib obstruction.

[0017] S13. Data Acquisition, Ranging, Monitoring, and Real-time Rescanning The scanning software calculates and collects distance measurement parameters for each sliding window in real time. During normal scanning, the scanner indicator light is green. When a yellow (too close) or blue (too far) indicator light appears on the scanner, it returns to S12 to perform a supplementary scan on the corresponding area to avoid feature loss during the surface reconstruction stage.

[0018] S14. Noise Removal and Isolated Node Clustering The original point cloud is subjected to dual screening based on a normal change rate threshold and a reflected light intensity threshold to remove unstable noise points. A clustering method based on local point density is used to merge or delete floating isolated points, outputting a clean point cloud that retains only the geometric features of the wing surface region.

[0019] S15, Hole Patch and Curvature Continuity Constraint The curvature of the hole edge is detected in the data processing software, and a patch is generated by least squares fitting of the hole edge neighborhood, so that the curvature of the patch is continuous with the curvature of the periphery, and the shape of the wing surface region is completely restored.

[0020] S16 Mesh-Spline Hybrid Reconstruction Using the refined point cloud output from the previous sub-step as input, the Poisson surface reconstruction algorithm is called to generate a closed triangular mesh in one go; the strip data with guaranteed coverage density is directly used to perform normal unification and adaptive resampling on the mesh, and automatic subdivision is performed at the ribs and fastener protrusions to preserve local curvature.

[0021] After completing the triangular mesh, smoothing is performed based on moving least squares (MLS), and only the high-frequency noise components are softened to ensure that the geometric features of ribs and closed angles are not excessively weakened.

[0022] Optionally, S2 includes the following steps: S21. Creation of Curvature-Driven Fragment Index In the decoding layer, the curvature values ​​of each band are read sequentially along the wingspan direction, forming a curvature sequence ordered in order.

[0023] Compare this sequence before and after: subtract the curvature value of the current band from the curvature value of the previous band and observe the magnitude of the change. If the magnitude of the change at any point reaches or exceeds a pre-set threshold, that location is considered a curvature abrupt change point, and it is also identified as the boundary between two segments. Record all such abrupt changes in the order they are discovered, and compile them into a segment index table. The table's arrangement directly corresponds to the future segment generation order, ensuring that the segmentation order of the digital model is completely consistent with the actual assembly process.

[0024] S22, Rib-fitting baseline reconstruction For each segment, the corresponding rib centerline is obtained by first retrieving the primary label, and the hard collision area is removed by using a curvature expansion strategy, and finally the cross-sectional baseline is reconstructed.

[0025] The baseline is collinear with the rib axis in the longitudinal direction and avoids the step area of ​​the fastener seat in the transverse direction, ensuring that the mold wall and the main rib fit together without gaps and without weakening the structural stress.

[0026] S23, Solving for expansion joint parameters Further in this step, all protruding labels and fastener labels are imported, and the assembly gap required for insertion is calculated for each piece.

[0027] Find the highest protrusion that the segment will encounter during the advancement process, add a small margin for safe entry and exit, and obtain the total dimension that the expansion groove must give way to in the lateral direction. The dimension determines the width of the groove.

[0028] By adding a safety factor after springback to the target wall thickness of the segment, the depth of the groove is pushed out in combination with the two, so that the mold can smoothly make way when it is pushed into the wing, and can immediately reset and self-lock by its own elasticity when the external force is removed.

[0029] The final width and depth values ​​will be written into the segmentation parameter table for direct use in subsequent 3D modeling and CNC machining.

[0030] S24, Thickened U-shaped contour reconstruction In the section reconstruction layer, the section baseline is coupled with curvature and target wall thickness to complete the outer wall shaping. This contour ensures continuous contact between the masking edge and the skin surface, preventing paint leakage gaps, while maintaining overall stress balance of the mold.

[0031] S25. Segmentation Nesting and Localization Feature Generation Once the entire outline is determined, each segment is nested sequentially from the wing root to the wingtip according to the index, and numbered positioning holes are automatically arranged on the end face of each segment to establish a consistent benchmark from virtual to physical.

[0032] S26, 3D Geometry Export and Virtual Assembly Backtracking The scanned data is converted into an editable 3D solid file through reverse modeling and then virtually assembled in the Catia software. If there are still areas with missing paint in the closed-angle region, fine-tuning is performed based on the assembly stress and occlusion angle feedback until all detected surfaces reach the set lower limit of coating thickness before the final geometry is output.

[0033] Optionally, S3 includes the following steps: S31. Preliminary thermal modeling The 3D solid model is sliced ​​layer by layer along the Z-axis, and the number, local wall thickness and curvature geometry of each layer are recorded in sequence.

[0034] Based on the wall thickness and curvature of the layer, and according to the melting point of polylactic acid material, the heat source temperature is set to 220℃.

[0035] Next, heat conduction links are automatically established between adjacent layers, and all links are merged into a complete thermal resistance network; the link coefficients are uniformly based on the thermal conductivity constant of polylactic acid material.

[0036] After completing the network setup, the finite volume solver library was called to quickly calculate the temperature distribution of the entire component over time using a time-stepping method, resulting in a temperature cloud map.

[0037] The temperature cloud map serves as the thermal boundary for subsequent warpage prediction, providing a foundation for thermo-mechanical coupling analysis.

[0038] S32, Thermal-Mechanical Coupled Warpage Prediction Path planning is based on thermo-mechanical coupling to calculate the instantaneous shrinkage of each grid; the grid is marked as a risk zone and its center coordinates are recorded and synchronized to the calibration list. Initializing the risk zone is the sole input for subsequent velocity and cooling back-calculations, ensuring that the calculations focus on the actual deformation source.

[0039] S33, Inverse calculation of extrusion speed and cooling air volume When thermo-mechanical coupling analysis reveals that certain meshes are at risk of exceeding the allowable shrinkage amount, the allowable shrinkage value is set as the target that must be achieved, and the optimal extrusion speed and cooling airflow are calculated in reverse: Based on the current temperature of the grid, the cumulative heating time, and the geometry of the nozzle and cooling vent, the actual shrinkage that would occur if the current process conditions were maintained is estimated. If this estimate is still higher than the tolerance, then iterative adjustments will be initiated.

[0040] During the iteration process, the controller prioritizes reducing the extrusion speed, allowing the material more time to dissipate heat and gradually solidify; If slowing down is still not enough to bring the shrinkage within the tolerance, then appropriately increase the cooling airflow to increase the convective heat dissipation efficiency.

[0041] Repeatedly compare the predicted shrinkage amount with the target shrinkage amount until the difference between the two disappears or narrows to an acceptable range.

[0042] S34. Add a mesh or mesh with solids locally. If the risk zone is located in a closed-angle area and the alarm is still triggered after the above adjustment, then add a grid-like or grid-plus-solid support around the corresponding layer to further suppress deformation by increasing local stiffness.

[0043] S35, Global Warpage Verification and Cyclic Convergence After updating the speed, cooling, and stiffening parameters, the thermo-mechanical model is re-initiated to predict the global warpage. If the warpage is less than or equal to the target value, the final CNC code (G-Code) is generated directly.

[0044] Optionally, step S4 includes the following steps: S41. Nozzle installation and heated bed releveling First, remove the original nozzle from the equipment and install a 0.40mm brass nozzle with a tolerance of no more than 3μm. Tighten the nozzle to 0.35N·m using a torque wrench to prevent loosening during thermal cycling. Then, use the inductive height gauge inside the machine and the three-point method to relevel the heated bed, ensuring that the minimum gap between the nozzle and the bed surface is completely consistent with the thickness of the first layer of G-Code. The origin of the machine tool coordinate system is locked at the center of the reference hole at the wing root, so as to achieve synchronization of physical zero point and assembly datum.

[0045] S42, G-code import and sequence verification Import the G-Code into the motion control system, use the built-in previewer to quickly scroll through the layers, and carefully check the nozzle path for each print. After confirming that the zero point, Z-axis positive direction, and wing assembly coordinate system are consistent, perform read-only writing.

[0046] S43, Process Temperature Preheating and Flow Calibration Heat the nozzle to 220°C and the heated bed to 60°C, and maintain this temperature for 120 seconds until the temperature fluctuation is below ±1°C.

[0047] Continuously extrude 100mm of sample material and measure the line width online; when the line width is between 0.38mm and 0.42mm, the flow rate is deemed qualified to ensure that the first layer of material is continuous and there are no areas with insufficient material.

[0048] S44. First-layer laying and real-time adhesion monitoring Start printing and complete the first layer at 80% of the speed set in the G-Code. Monitor the nozzle tracking error and trajectory width in real time. If a gap of more than 0.10mm is found at the edge of the first layer, immediately press the pause button and eliminate the gap by fine-tuning the heated bed temperature by ±5℃ or reducing the nozzle-bed gap by 0.02mm before resuming printing to avoid the root cause of subsequent warping.

[0049] S45, Dynamic airflow-speed coupled execution During the printing process, the controller strictly follows the cooling airflow and speed instructions in the G-Code to switch in real time, and reads the stepper motor drive current at 5ms intervals to calculate the nozzle back pressure.

[0050] S46. Gradual Cooling Removal and Internal Stress Relief After the final layer is completed, the heated bed temperature is reduced to 40°C at a gradient of -5°C / min, and held for 60 seconds to achieve uniform cooling. The bed is then slowly removed from the heated bed using a mold, and the gradual cooling process releases internal stress, preventing new warping from forming at closed corners or the roots of the ribs due to concentrated temperature differences.

[0051] S47. Dimensional Re-inspection and Knowledge Base Backtracking Immediately after unloading, use a 0.01mm precision vernier caliper to re-inspect the preset test points, focusing on measuring the flange width, rib height, and center distance of the positioning holes. If any of these deviations exceeds 0.15mm, bind the deviation value, its coordinates, and the corresponding G-code line number and upload it to the knowledge base; if all indicators are qualified, mark the G-Code as a verified template, allowing mass production replication.

[0052] Optionally, S5 includes the following steps: S51. Precise positioning and initial seating Insert the cylindrical positioning hole of mold one into the wing root reference pin, and then slowly place it vertically to control the pin-hole fit clearance within 5μm. Then use a 0.03mm feeler gauge to check the coplanarity of the top of the rib and the bottom surface of the mold; if it exceeds the tolerance, gradually approach it by replacing the bottom metal shims to ensure that the initial posture does not introduce additional torsion.

[0053] S52, 3D scanning and point cloud acquisition After locating the S71's reference pins and maintaining a stationary position, a handheld structured light scanner is slowly moved along the wingspan and chord direction to perform a full-coverage scan of the mold's exterior and exposed wing skin. Before scanning, several coded targets are attached to the wing root, wingtip, and typical rib nodes for rapid synchronization between the scanning coordinate system and the assembly coordinate system. The system monitors the point cloud density in real time; if insufficient local data is detected, a supplementary scan prompt is immediately given until all areas meet the preset sampling interval requirements. After scanning, noise point removal and isolated point clustering are automatically performed to generate a clean point cloud, which is then precisely aligned with the CAD coordinate system.

[0054] S53, Best Fit and Fit Determination The software calls the global best-fit algorithm to rigidly register the newly obtained point cloud with the mold design surface and generate a color mark deviation map in real time.

[0055] The judgment criteria are as follows: The maximum positive deviation and the maximum negative deviation shall not exceed ±0.15 mm; The area of ​​the continuous out-of-tolerance region must be less than the threshold specified by the process. All critical locations, such as closed angles and the base of reinforcing bars, must meet the required standards.

[0056] If any condition is not met, a revision list is automatically generated, and the process is traced back to S3 or S5 for geometric or process compensation; if all indicators are qualified, the bonding is deemed successful, the deviation report is archived along with the point cloud data, and the process proceeds to the subsequent painting process.

[0057] Figures 2-6 The images shown are, in sequence, an isometric side view, a front view, a left view, a top view, and a three-dimensional view of a U-shaped paint masking mold prepared by the preparation method provided in the embodiments of the present invention.

[0058] This invention utilizes core strategies such as curvature difference-driven segmented indexing, rib fitting baseline reconstruction, and elastic back-thrust of the expansion groove to achieve a fully integrated process of close contact, clearance, and springback self-locking between the mold outer wall and the wing ribs. This allows the mold to automatically avoid rib and fastener protrusions during insertion and automatically lock itself after insertion, effectively avoiding the pain points of traditional hard molds being difficult to insert and paper patterns easily becoming suspended. It ensures that there are no visible gaps or hard-impact interference between the skin surface and the shielding wall.

[0059] Based on a dual-loop closed-loop system of coupled thermal, mechanical, and fluid field path correction and risk grid wall thickness fine-tuning, this invention can predict potential warping sources before additive manufacturing and simultaneously adjust extrusion speed, cooling airflow, and local filling stiffness during the manufacturing process. After manufacturing, targeted compensation is made through wall thickness stiffening or weight reduction. This multi-layered protection system of source prevention, process correction, and result feedback significantly suppresses the warping risk in easily deformable areas such as thin walls and closed angles, ensuring that the self-locking groove of the mold maintains a stable locking force after multiple insertions and removals.

[0060] The process iteration scheme, which combines hierarchical locking of local thermal field re-solution with initial compensation vector of knowledge base, enables thermal solution to be recalculated only for the modified area, while existing results are directly reused for other areas, which greatly shortens the time for secondary or even multiple solutions. At the same time, the system automatically records and reuses the compensation experience of similar mold tasks, which enables new tasks to converge quickly and reduces repeated trial and error and manual parameter tuning.

Claims

1. A rapid preparation method for a U-shaped paint masking mold based on the curved surface features of an airfoil, characterized in that, Includes the following steps: S1. Mark the unpainted feature areas on the wing surface and make their origin coaxial with the wing root reference hole. Use a laser scanner to scan the wing surface area at a constant speed in a strip manner, calculate the distance measurement index in real time and collect point cloud data, and scan the missing areas to complete noise removal and continuous patching of hole curvature to obtain the corresponding model of the wing surface. S2. Establish a segmented index for the corresponding model of the wing curved surface, design the expansion groove parameters of the mold for each segment, reconstruct the thickened U-shaped contour, and output the three-dimensional solid model of the mold. S3. Perform Z-axis slicing on the three-dimensional solid model of the mold, quickly solve the thermal field and perform thermo-mechanical coupling warpage prediction. If a risk area is detected, it is filled by implantable mesh or mesh plus solid support. The correction is repeated until the residual warpage meets the paint adhesion tolerance and output G code is generated. S4. Mold printing is completed on a fused deposition modeling (FDM) machine; S5. Perform deviation calibration on the printed mold.

2. The rapid preparation method of a U-shaped paint masking mold based on the curved surface features of an airfoil according to claim 1, characterized in that, S1 includes the following steps: S11, Calibration and Coordinate System Affix calibration points to unpainted feature areas on the curved surface of the wing. The calibration points should be affixed randomly, and the origin of the calibration points should be fixed coaxially with the reference hole at the wing root. A handheld laser scanner is used to scan the curved area of ​​the wing in a constant speed strip manner. The homogeneous transformation matrix from the scanner coordinate system to the wing assembly coordinate system is solved in real time, and the scanner coordinate system is transformed to the wing assembly coordinate system in real time throughout the scanning process. S12, Strip Division and Constant Speed ​​Scanning Based on the spacing between adjacent ribs, the curved surface area of ​​the wing is divided into several longitudinal strips along the wingspan. A constant linear velocity is maintained to cover the rib flanges within each strip, and then the rib grooves of the same strip are swept back to ensure that the point spacing inside the strip is evenly distributed. S13. Data Acquisition, Ranging, Monitoring, and Real-time Rescanning The scanning software calculates the distance measurement index and collects point cloud data for each sliding window in real time. When the scanner indicates that the distance is too close or too far, it returns to S12 to perform a rescan on the corresponding area. S14. Noise Removal and Isolated Node Clustering The original point cloud data is filtered twice based on the normal change rate threshold and the reflected light intensity threshold to remove unstable noise points; a clustering method based on local point density is used to merge or delete floating isolated points. S15, Hole Patch and Curvature Continuity Constraint In the data processing software, the curvature of the hole edge corresponding to the calibration point is detected, and a patch is generated by least squares fitting of the neighborhood of the hole edge. The curvature of the patch is continuous with the curvature of the perimeter, and the shape of the wing surface region is completely restored. The output is a fine point cloud that retains only the geometric features of the wing surface region. S16, Mesh-Spline Hybrid Reconstruction Using the refined point cloud as input, the surface reconstruction algorithm is called to generate closed triangular mesh data in one go. Normal unification and adaptive resampling are performed on the mesh, and automatic subdivision is performed at the ribs and fastener protrusions to preserve local curvature. After completing the triangular mesh, based on moving least squares smoothing, only the high-frequency noise components are softened to ensure that the geometric features of the ribs and closed angles are not excessively weakened, thus obtaining the corresponding model of the wing curved surface.

3. The rapid preparation method of a U-shaped paint masking mold based on the curved surface features of an airfoil according to claim 1, characterized in that, S2 includes the following steps: S21. Creation of Curvature-Driven Fragment Index The curvature values ​​of each strip are read sequentially along the wingspan direction to form a curvature sequence ordered in order; Compare the sequence before and after: subtract the curvature value of the current band from the curvature value of the previous band and observe the change range; if the change range at a certain point reaches or exceeds the preset threshold, then the position is regarded as a curvature inflection point, and at the same time, it is determined that this should be the boundary between two segments; record all such inflection points in chronological order to form a segment index table. S22, Rib-fitting baseline reconstruction For each segment of the mold, the corresponding rib centerline is obtained by first searching the primary label, and the hard collision area is removed by using the curvature expansion strategy, and finally the cross-sectional baseline is reconstructed. The cross-sectional baseline is collinear with the rib axis in the longitudinal direction and avoids the hard collision zone in the transverse direction, ensuring that the mold wall and the rib fit together without gaps and without weakening the structural stress. S23, Solving for expansion joint parameters Import all raised labels and hard collision area labels, and calculate the assembly gap required for insertion of each piece; Find the highest protrusion that the segment will encounter during the advancement process, add a small margin for safe entry and exit, and obtain the total dimension that the expansion groove must give way in the lateral direction. The total dimension is used to determine the width of the expansion groove. By adding a safety factor after springback to the target wall thickness of the segmented mold, the depth of the telescopic groove is deduced, so that the mold can smoothly make way when it is pushed into the wing, and can immediately reset and self-lock by its own elasticity when the external force is released. The final generated width and depth values ​​will be written into the piecewise parameter table for direct use in subsequent 3D modeling and CNC machining; S24, Thickened U-shaped contour reconstruction The cross-sectional baseline is coupled with the mold curvature and the target mold wall thickness to complete the mold outer wall shaping; the mold outer wall ensures that the mold's shielding edge and the wing surface are continuously in contact, without producing paint leakage gaps, while maintaining the overall stress balance of the mold; S25. Segmentation Nesting and Localization Feature Generation Once the entire thickened U-shaped contour is determined, each segment of the mold is nested sequentially from the wing root to the wing tip according to the index, and numbered positioning holes are automatically arranged on the end face of each segment to establish a consistent benchmark from virtual to physical, thus obtaining the model corresponding to the mold. S26, 3D Geometry Export and Virtual Assembly Backtracking Closed triangular mesh data is converted into editable 3D solid files through reverse modeling, and the wing surface corresponding model and the mold corresponding model are virtually assembled in Catia software. If there are still paint leakage areas in the closed angle area of ​​the wing surface, fine adjustments are made based on assembly stress and shading angle feedback until all inspection surfaces reach the set coating thickness lower limit before the final 3D model of the mold is output.

4. The rapid preparation method of a U-shaped paint masking mold based on the curved surface features of an airfoil according to claim 1, characterized in that, Step S3 includes the following steps: S31. Preliminary thermal modeling The three-dimensional solid model is sliced ​​layer by layer along the Z-axis, and the number, local wall thickness and curvature geometry of each layer are recorded in sequence. The heat source temperature is set according to the wall thickness and curvature of each layer; Automatically establish heat conduction links between adjacent layers and merge all links into a complete thermal resistance network; After completing the thermal resistance network construction, the finite volume solver library was called to quickly calculate the temperature distribution of the entire mold over time using a time-stepping method, resulting in a temperature cloud map. The temperature cloud map serves as the thermal boundary for subsequent warpage prediction, providing a foundation for thermo-mechanical coupling analysis. S32, Thermal-Mechanical Coupled Warpage Prediction Path planning is based on the thermo-mechanical coupling to calculate the instantaneous shrinkage of each mesh in the thermal resistance network; meshes with thermal stress exceeding a set threshold are marked as risk zones and their center coordinates are recorded and synchronized to the correction list; initializing the risk zone is the only input for subsequent velocity and cooling back calculations, ensuring that the calculation focuses on the source of substantial deformation; S33. Fill the risk area with an implantable mesh or mesh plus solid support, and cyclically correct until the residual warpage meets the paint adhesion tolerance. S35, Global Warpage Verification and Cyclic Convergence After updating the speed, cooling, and support, the thermo-mechanical model is called back to predict the global warpage and generate the final CNC code G-code.

5. The rapid preparation method of a U-shaped paint masking mold based on the curved surface features of an airfoil according to claim 4, characterized in that, S33 specifically refers to: Set the allowable shrinkage value for each grid in the risk zone as the target that must be achieved, and then reverse-engineer the most suitable extrusion speed and cooling airflow for the die material: Based on the current temperature of the grid, the cumulative heating time, and the geometry of the nozzle and cooling vent, the actual shrinkage that would occur if the current process conditions were maintained is estimated. If this estimate is still higher than the tolerance, then iterative adjustment will be initiated; During the iteration process, the controller prioritizes reducing the extrusion speed, allowing the material more time to dissipate heat and gradually solidify; If slowing down is still not enough to compress the contraction into the tolerance, then appropriately increase the cooling airflow to increase the convective heat dissipation efficiency. Repeatedly compare the predicted shrinkage amount with the target shrinkage amount until the difference between the two disappears or narrows to an acceptable range.

6. The rapid preparation method of a U-shaped paint masking mold based on the curved surface features of an airfoil according to claim 4, characterized in that, S33 also includes: If the risk zone is located in a closed-angle region and the predicted shrinkage is still greater than the target shrinkage after the above adjustments, then add a grid-like or grid-plus-solid support around the corresponding layer to further suppress deformation by increasing local stiffness.

7. The rapid preparation method of a U-shaped paint masking mold based on the curved surface features of an airfoil according to claim 1, characterized in that, S5 includes the following steps: Obtain the clean point cloud data of the mold; The software calls the global best-fit algorithm to rigidly register the newly obtained fine point cloud of the mold with the design surface of the mold's 3D solid model, and generates a color mark deviation map in real time.

8. A rapid preparation method for a U-shaped paint masking mold based on the curved surface features of an airfoil, as described in claim 7, is characterized in that... The criteria for judging color mark deviation diagrams are as follows: The maximum positive deviation and the maximum negative deviation shall not exceed ±0.15 mm; The area of ​​the continuous out-of-tolerance region must be less than the threshold specified by the process. All critical locations, including closed angles and the base of reinforcing bars, must meet the required standards. If any clause is not met, the process is back to S2 or S4 for geometric or process compensation; if all indicators are qualified, the bonding is deemed successful.

9. A rapid preparation method for a U-shaped paint masking mold based on the curved surface features of an airfoil, as described in claim 7, is characterized in that... Obtaining the clean point cloud data of the mold includes: S51. Precise positioning and initial seating Insert the cylindrical positioning hole of the mold into the reference pin of the wing root, and then slowly place it vertically to control the pin-hole fit clearance within 5μm. Use a feeler gauge to check the coplanarity of the top of the rib and the bottom surface of the mold. If it exceeds the tolerance, gradually approach it by replacing the bottom metal shim to ensure that the initial posture does not introduce additional torsion. S52, 3D scanning and point cloud acquisition After completing the positioning of the S51 reference pin, the device remains stationary. A handheld structured light scanner is slowly moved along the wingspan and chord direction to perform a full-coverage scan of the mold surface and exposed wing skin. Before scanning, several coded targets are attached to the wing root, wingtip, and typical rib nodes to facilitate rapid synchronization between the scanning coordinate system and the assembly coordinate system. After scanning, noise point removal and isolated point clustering are automatically performed to generate a clean point cloud of the mold and precisely align it with the CAD coordinate system.

10. A rapid preparation method for a U-shaped paint masking mold based on the curved surface features of an airfoil, as described in claim 7, is characterized in that... The fine point cloud data of the mold can also be obtained by direct scanning.

Images