A method and system for designing a weld site for multi-segment target material welding

Abstract

This invention belongs to the field of intelligent welding technology. It discloses a method and system for designing welding parts for multi-segment target welding. The method includes scanning the multi-segment target using laser scanning and visible light imaging equipment to obtain the three-dimensional morphology of the assembled multi-segment target; identifying the weld region based on the three-dimensional morphology and extracting the geometric deviation features of the weld region; generating dynamically adapted bevel design parameters based on the geometric deviation features of the weld region and the thermal expansion characteristics of the welding material, according to preset bevel design rules; collecting thermal infrared image data and strain field data during the welding process to construct a thermo-mechanical coupled welding process simulation model; and obtaining the distribution characteristics of the strain concentration area in the heat-affected zone by analyzing the strain concentration and residual stress distribution in the heat-affected zone during welding, thereby constructing a weld path with a fractal structure. This improves the overall quality and performance of the welding parts of the multi-segment target.

Description

Technical Field

[0001] This invention relates to the field of intelligent welding technology, and more specifically, to a method and system for designing welding parts for multi-segment target welding. Background Technology

[0002] With the rapid development of high-tech industries such as semiconductors, flat panel displays, and solar photovoltaics, multi-segment sputtering targets are typically assembled and welded together from segments of different materials and functions. Their flexible material combinations and high deposition efficiency have led to their widespread application. In the manufacturing process of multi-segment sputtering targets, the welding process is a crucial step in ensuring the performance of the target material. The design of the welding points directly affects the overall structural strength, conductivity, thermal stability, and uniformity of thin film deposition.

[0003] Traditional methods and systems for designing welding areas in multi-segment target welding have the following main problems:

[0004] Traditional welding design methods are often based on experience and simple mechanical analysis, lacking systematicity and precision. This design approach fails to fully consider the complex stress distribution, thermal conductivity characteristics, and physicochemical compatibility between different materials during use, easily leading to problems such as stress concentration, insufficient interfacial bonding strength, and mismatched coefficients of thermal expansion at the welding points. This can result in target cracking and detachment, affecting the quality of thin film deposition and the normal operation of equipment, and increasing production and maintenance costs.

[0005] Furthermore, with the continuous increase in the size and increasing complexity of target materials, traditional design methods can no longer meet the manufacturing requirements of high-precision, high-performance targets. At the same time, existing welding area designs lack effective digital simulation and optimization methods, making it difficult to accurately predict the performance of welding areas during the design phase. This results in long design cycles, high testing costs, and an inability to quickly respond to market demands for new target materials.

[0006] In view of this, the present invention proposes a design method and system for welding parts of multi-segment target welding to solve the above problems. Summary of the Invention

[0007] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a method for designing welding positions for multi-segment target welding, comprising:

[0008] S1. Scan the multi-segment target material using laser scanning and visible light imaging equipment to obtain the three-dimensional morphology of the multi-segment target material assembly; identify the weld area based on the three-dimensional morphology of the assembly and extract the geometric deviation features of the weld area.

[0009] S2. Based on the geometric deviation characteristics of the weld area and the thermal expansion characteristics of the welding material, dynamically adapted groove design parameters are generated according to the preset groove design rules.

[0010] S3. Collect thermal infrared image data and strain field data during the welding process, construct a thermo-mechanical coupling welding process simulation model, and obtain the distribution characteristics of the strain concentration area in the heat-affected zone by analyzing the strain concentration and residual stress distribution in the heat-affected zone during the welding process, and then construct a weld path with a fractal structure.

[0011] S4. Based on scanning electron microscopy and electron backscatter diffraction, and combined with cross-scale image fusion method, the microstructure and element diffusion characteristics of the welded joint interface region are analyzed, and the pore distribution, ripple amplitude spectrum and element diffusion curve of the joint interface region are obtained.

[0012] S5. Based on the pore distribution, ripple amplitude spectrum and element diffusion curve of the joint interface region, design the gradient transition layer structure of the welding interface to achieve a continuous transition of the welding interface performance.

[0013] S6. Based on the dynamically adapted groove design parameters, the weld path of the fractal structure, and the gradient transition layer structure of the welding interface, generate a design scheme for the welding part that can be used for multi-segment target welding.

[0014] Preferably, the method for obtaining the three-dimensional morphology of the multi-segment target assembly includes:

[0015] The principle of line laser triangulation is adopted, a laser scanning device is selected, and a visible light imaging device is equipped. The multi-segment target is installed on a clamping platform with stable support and multi-axis motion control, and the multi-segment target is scanned using the laser scanning device and the visible light imaging device.

[0016] Based on the preset scanning path of the multi-segment target's geometry, the laser scanning device is controlled to emit a laser beam along the preset scanning path to scan the target surface point by point, and the point cloud data of the complete shape of the target is obtained by receiving reflected light; while the laser scanning is being performed, a visible light imaging device is used to simultaneously capture images of the target surface texture, and each frame of the image is accompanied by a timestamp and a spatial coordinate label consistent with the laser scanning device.

[0017] By iterating the nearest point ICP algorithm, the point cloud data in each direction are stitched together and aligned to the camera coordinate system; by using nearest neighbor interpolation, the RGB values ​​of the target surface texture image are mapped to the corresponding point cloud data, thereby generating the assembled three-dimensional shape of the multi-segment target.

[0018] Preferably, the method for extracting the geometric deviation features of the weld area includes:

[0019] By solving the local neighborhood covariance matrix, the normal vectors of all points in the point cloud data of the complete target shape are obtained; the regions of abrupt change in normal vectors are extracted, and the candidate regions of welds are located in combination with the preset point cloud height difference threshold.

[0020] Canny edge detection is performed on the synchronously acquired target surface texture map to extract the contour line composed of two-dimensional edge points with abrupt gray-level gradient changes. The contour line is mapped to point cloud coordinates by the timestamp and spatial coordinate label attached to each frame image, and the two-dimensional edge points are back-projected to the three-dimensional point cloud space to obtain the three-dimensional candidate weld edge point set.

[0021] Region growing and morphological filtering: Merge the points in the candidate weld region and the set of 3D candidate weld edge points to form a seed point set; perform region clustering growth based on the spatial proximity and normal vector angle between point clouds to gradually expand the weld region; at the same time, use morphological opening operation to remove noise points and close small gaps, and finally form a set of weld region point clouds.

[0022] The main direction of the weld in the point cloud set of the weld area is analyzed. Based on the standard process specification, the ideal welding path is preset. The deviation between the main direction of the weld in the point cloud set of the weld area and the preset ideal welding path is calculated point by point, so as to obtain the geometric deviation characteristics of the weld area.

[0023] Preferably, the method for obtaining the bevel design parameters includes:

[0024] Based on the material of the target welding material, the corresponding coefficient of thermal expansion and thermal conductivity are obtained by querying. Combined with the welding heat input and process parameters, a thermal-structural response analysis model is performed to predict the shrinkage deformation, residual stress and structural warping trend after welding. The preset groove design rules combine the shrinkage deformation, residual stress and structural warping trend after welding with the geometric deviation characteristics of the weld area to form a joint feature vector, which is matched in the preset groove design rules to dynamically generate suitable groove design parameters.

[0025] Preferably, the method for constructing the thermo-mechanical coupling welding process simulation model includes: constructing heat conduction equations and mechanical equations based on heat conduction theory and elastoplastic mechanics theory, thereby obtaining a thermo-mechanical coupling model; and using the thermo-mechanical coupling model to simulate the temperature field and strain field during the welding process.

[0026] Preferably, the method for constructing a weld path with a fractal structure includes:

[0027] The temperature distribution sequence of the heat-affected zone during the welding process is acquired in real time by a thermal infrared camera to obtain thermal infrared image data; a distributed fiber optic strain sensing system is used to deploy monitoring points in the weld area of ​​the weld area point cloud set to dynamically capture the three-dimensional strain distribution during the welding process, form a strain tensor field, and obtain strain field data.

[0028] The three-dimensional shape of the assembly is geometrically modeled and meshed. The finite element method is used to simulate the thermo-mechanical coupling process of welding based on the thermo-mechanical coupling model. The temperature field is obtained by solving the heat conduction equation. The temperature field is used as a load input to the mechanical equation to solve the stress field. The distribution characteristics of strain concentration area in the heat-affected zone are obtained by analyzing the strain concentration and residual stress distribution under the temperature field.

[0029] The maximum principal strain distribution in the strain concentration region distribution characteristics of the heat-affected zone is extracted, and the strain concentration region is identified by the threshold segmentation method to generate a continuous high-strain region. The SIMP method is used for topology optimization modeling, with the minimum strain energy density per unit volume of the structure as the objective function, to construct the fractal structure of the weld path. The fractal structure of the weld path is discretized into welding torch movement path points, thereby obtaining the weld path with fractal structure.

[0030] Preferably, the method for obtaining the pore distribution, ripple amplitude spectrum, and elemental diffusion curves of the joint interface region includes:

[0031] Samples were taken and surface-prepared for the interface region of the multi-segment target welding joint. The surface preparation process included grinding, polishing, and ion beam removal of the surface deformation layer. Secondary electron images and backscattered electron images of the interface region of the welding joint at different scales were obtained using a scanning electron microscope. An EBSD probe was equipped on the scanning electron microscope platform to perform crystallographic orientation analysis on the interface region of the welding joint and obtain the grain orientation distribution map of the region.

[0032] Image stitching and multi-scale enhancement algorithms are used to construct a panoramic image of the welded joint interface containing structural information at the micron and nanoscale. The pore regions of the panoramic image of the welded joint interface are identified to obtain the pore distribution of the joint interface region.

[0033] Boundary tracking and extraction of the ripple profile in the panoramic image of the welded joint interface are performed, and the ripple amplitude spectrum is obtained by Fourier transform analysis. Using spectral analysis technology, cross-sectional scanning paths and surface scanning areas are set on the multi-segment target weld interface and its two sides to obtain spatial distribution data of different metal elements in the welded joint interface area. Based on the spatial distribution data, element diffusion curves of element concentration as a function of spatial location are generated.

[0034] Preferably, the method for designing the gradient transition layer structure of the welding interface includes:

[0035] Based on the pore distribution data identified in the joint interface region, the pore density and size are identified, and a buffer weld bead is designed. By remelting the material and interlayer staggering, the filling density of the pore distribution region is improved, resulting in a pore density field. Micro-concave / gentle slope corrugated structures are arranged in the fluctuating frequency abrupt region of the corrugation amplitude spectrum to release thermal stress along the structural morphology direction, resulting in a corrugation morphology field. Based on the slope of the element diffusion curve, an M-layer alloy combination is designed, with the material composition of each layer gradually changing, resulting in a composition concentration field. A parameter field-driven model based on a neural network is used to fuse the pore density field, corrugation morphology field, and composition concentration field to generate a three-dimensional gradient transition structure map, resulting in a gradient transition layer structure of the welding interface.

[0036] Preferably, the method for generating a welding part design scheme that can be used for multi-segment target welding includes:

[0037] Based on the dynamically adapted groove design parameters, the weld path of the fractal structure, and the gradient transition layer structure of the welding interface, a design scheme for the welding part that can be used for multi-segment target welding is generated.

[0038] A structured welding part model is established using 3D CAD software based on dynamically adapted bevel design parameters, fractal weld paths, and gradient transition layers at the welding interface. The structured welding part model is then converted into a set of instructions that the welding robot can recognize using a CNC programming interface, thus forming a digital design scheme for welding parts that can be used for multi-segment target welding.

[0039] A welding area design system for multi-segment target welding includes:

[0040] The 3D topography acquisition module scans the multi-segment target material using laser scanning and visible light imaging equipment to acquire the assembled 3D topography of the multi-segment target material; based on the assembled 3D topography, it identifies the weld area and extracts the geometric deviation features of the weld area.

[0041] The bevel parameter generation module generates dynamically adapted bevel design parameters based on the geometric deviation characteristics of the weld area and the thermal expansion characteristics of the welding material, according to the preset bevel design rules.

[0042] The weld path generation module collects thermal infrared image data and strain field data during the welding process, constructs a thermo-mechanical coupled welding process simulation model, and obtains the distribution characteristics of the strain concentration area in the heat-affected zone by analyzing the strain concentration and residual stress distribution in the heat-affected zone during the welding process, thereby constructing a weld path with a fractal structure.

[0043] The microstructure analysis module, based on scanning electron microscopy and electron backscatter diffraction, and combined with cross-scale image fusion methods, analyzes the microstructure and element diffusion characteristics of the weld joint interface region, and obtains the pore distribution, ripple amplitude spectrum and element diffusion curve of the joint interface region.

[0044] The interface structure design module designs the gradient transition layer structure of the welding interface based on the pore distribution, ripple amplitude spectrum and element diffusion curve of the joint interface area, so as to achieve a continuous transition of the welding interface performance; the welding scheme integration module generates a welding part design scheme that can be used for multi-segment target welding based on the dynamically adapted groove design parameters, the weld path of the fractal structure and the gradient transition layer structure of the welding interface.

[0045] The technical effects and advantages of the welding part design method and system for multi-segment target welding of the present invention are as follows:

[0046] By employing two methods—initial screening of weld candidate regions and edge enhancement of texture images—the weld is located from two dimensions: point cloud data and texture images. These two methods complement each other, improving the accuracy and reliability of weld location.

[0047] A thermo-mechanical coupling model is constructed based on the theories of heat conduction and elastoplastic mechanics. By solving the heat conduction equation and the mechanical equation, the temperature field and stress field distribution in the heat-affected zone during the welding process can be accurately simulated. This helps to predict the magnitude and distribution of residual stress and deformation of the structure after welding, and provides a theoretical basis for the optimization of welding process.

[0048] By simulating the welding process under different parameters using a thermo-mechanical coupling model, the changes in temperature and stress fields are analyzed to find the optimal welding process parameters, reduce the probability of welding defects, and improve process stability and product qualification rate. The model has good simulation efficiency and engineering adaptability, and can support the simulation analysis of multi-pass welding and multi-layer welding of complex components. It provides online prediction and feedback control basis for intelligent welding systems, enhances the automation and intelligence capabilities of the system, and improves welding quality and efficiency.

[0049] The weld path design method based on actual welding process data and simulation analysis can optimize the weld path, reduce stress concentration during welding, and improve welding quality. By rationally designing the gradient transition layer structure of the welding interface, stress concentration at the weld joint interface can be effectively alleviated, interface defects can be reduced, and the strength, toughness, and corrosion resistance of the weld joint can be improved, thus extending the service life of the welded structure.

[0050] The gradient transition layer structure enables tighter bonding between different materials, reduces uneven element diffusion at the interface, thereby improving welding quality and ensuring the reliability of the welded structure under complex working conditions. It can quickly and accurately generate digital design schemes for welding parts that meet actual needs, reducing the time and errors of manual design and improving design efficiency and quality. Attached Figure Description

[0051] Figure 1This is a schematic flowchart of a welding part design method for multi-segment target welding according to the present invention;

[0052] Figure 2 This is a schematic diagram of a welding part design system for multi-segment target welding according to the present invention;

[0053] Figure 3 This is a schematic diagram of the method for obtaining bevel design parameters provided by the present invention. Detailed Implementation

[0054] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] Example 1

[0056] Please see Figure 1 and Figure 3 As shown in the figure, this embodiment further illustrates the welding part design method for multi-segment sputtering proposed in this invention. This design can be applied to transform large / heavy metal sputtering materials that are difficult to transport into a modular design that can be transported in segments: Due to limitations in size, weight, or special shape, large or heavy metal sputtering materials are often difficult to transport as a whole, and therefore need to be disassembled into multiple metal parts (multi-segment sputtering materials), which are then transported to the destination and welded together. The welding part design method for multi-segment sputtering materials proposed in this invention can effectively solve the welding problems in such scenarios, ensuring that the spliced ​​sputtering material can meet the standards of an integral sputtering material.

[0057] Against the backdrop of semiconductor chip manufacturing processes reaching more advanced nodes, flat panel displays evolving towards ultra-high-definition resolution, and solar photovoltaics pursuing higher photoelectric conversion efficiency, multi-segment sputtering targets, with their flexible material combinations and efficient thin-film deposition performance, have become a key basic material for high-tech industries. This special structure, composed of spliced ​​and welded segments of different materials and functions, greatly improves industrial production efficiency and reduces production costs.

[0058] However, in the manufacturing process of multi-segment sputtering targets, the welding process has become a core bottleneck restricting their performance. The design quality of the welded parts directly determines the overall structural strength, electrical and thermal conductivity, and thin film deposition uniformity of the target, playing a decisive role in the yield of downstream products and the stability of equipment operation. Traditional welded part design relies on accumulated experience and simple mechanical calculations, lacking systematic theoretical support and precise quantitative analysis models. This extensive design mode is unable to cope with the complex operating conditions of the target during service, such as the high-frequency alternating electric field in semiconductor sputtering and the high-temperature and high-vacuum environment in flat panel display coating, resulting in the inability to fully consider key factors such as stress distribution, thermal conductivity, and physicochemical compatibility between materials.

[0059] This has led to a series of serious problems: stress concentration at the welding points causes microcracks in the target material under cyclic thermal stress, eventually resulting in structural cracking; insufficient interfacial bonding strength causes the target material to detach under high-speed ion bombardment, contaminating the coating equipment and reducing film quality; mismatch in thermal expansion coefficients causes permanent deformation at the welding points, damaging the flatness of the target material. These problems not only significantly increase production and maintenance costs, but also severely restrict capacity expansion due to frequent downtime for maintenance.

[0060] Meanwhile, industrial development is driving the evolution of sputtering targets towards larger sizes and more complex structures, highlighting the limitations of traditional design methods in terms of accuracy and efficiency. Due to the lack of effective support from digital methods such as finite element analysis and molecular dynamics simulations, it is difficult to accurately predict the mechanical, thermal, and chemical properties of the welded areas during the design phase. This leads to repeated testing and corrections, lengthy design cycles, and high R&D costs, hindering a rapid response to the market's urgent demand for new high-performance sputtering targets. Therefore, constructing a scientific and systematic multi-segment sputtering target welding area design method and supporting system has become a key task in overcoming the technological bottlenecks in industrial development and ensuring the reliability of sputtering target performance.

[0061] To effectively solve the above problems, this invention proposes a design method for welding parts in multi-segment target welding, including:

[0062] S1. Scan the multi-segment target material using laser scanning and visible light imaging equipment to obtain the three-dimensional morphology of the multi-segment target material assembly; identify the weld area based on the three-dimensional morphology of the assembly and extract the geometric deviation features of the weld area.

[0063] S2. Based on the geometric deviation characteristics of the weld area and the thermal expansion characteristics of the welding material, dynamically adapted groove design parameters are generated according to the preset groove design rules.

[0064] S3. Collect thermal infrared image data and strain field data during the welding process, construct a thermo-mechanical coupling welding process simulation model, and obtain the distribution characteristics of the strain concentration area in the heat-affected zone by analyzing the strain concentration and residual stress distribution in the heat-affected zone during the welding process, and then construct a weld path with a fractal structure.

[0065] S4. Based on scanning electron microscopy and electron backscatter diffraction, and combined with cross-scale image fusion method, the microstructure and element diffusion characteristics of the welded joint interface region are analyzed, and the pore distribution, ripple amplitude spectrum and element diffusion curve of the joint interface region are obtained.

[0066] S5. Based on the pore distribution, ripple amplitude spectrum and element diffusion curve of the joint interface region, design the gradient transition layer structure of the welding interface to achieve a continuous transition of the welding interface performance.

[0067] S6. Based on the dynamically adapted groove design parameters, the weld path of the fractal structure, and the gradient transition layer structure of the welding interface, generate a design scheme for the welding part that can be used for multi-segment target welding.

[0068] Methods for obtaining the three-dimensional morphology of multi-segment target assembly include:

[0069] The principle of line laser triangulation is adopted, and laser scanning equipment (such as laser triangulation or structured light) is selected and equipped with visible light imaging equipment (such as industrial camera). The multi-segment target is mounted on a clamping platform with stable support and multi-axis motion control to ensure the stability and repeatability of the target during the scanning process. The multi-segment target is scanned using laser scanning equipment and visible light imaging equipment.

[0070] Based on the geometric shape (cylindrical / irregular) of the multi-segment target material, a spiral or grid-shaped scanning path is preset to ensure that the overlap rate of adjacent scanning areas is ≥30%. The laser scanning device is controlled to emit a linear laser beam along the preset scanning path to scan the target material surface point by point. The point cloud data of the complete shape of the target material is obtained by receiving reflected light. At the same time as the laser scanning, a visible light imaging device is used to simultaneously capture the surface texture image of the target material. Each frame of the image is accompanied by a timestamp and a spatial coordinate label consistent with the laser scanning device to help identify surface details such as weld edges, contour lines, positioning holes, and assembly errors.

[0071] By iterating the nearest point ICP algorithm, the point cloud data in each direction are stitched together and aligned to the camera coordinate system; by using nearest neighbor interpolation, the RGB values ​​of the target surface texture image are mapped to the corresponding point cloud data, thereby generating the assembled three-dimensional shape of the multi-segment target.

[0072] Methods for extracting geometric deviation features of the weld area include:

[0073] Initial screening of weld candidate regions: By solving the local neighborhood covariance matrix, the normal vectors of all points in the point cloud data of the complete target shape are obtained; a covariance matrix is ​​constructed for the k-neighbor point cloud of each point, and the eigenvector corresponding to its smallest eigenvalue is the normal vector; regions with abrupt changes in normal vectors (there are usually drastic changes in normal vectors at the edge of the weld) are extracted, and the weld candidate regions are located by combining the preset point cloud height difference threshold (such as the height difference between adjacent point clouds > 0.5 mm);

[0074] Since the weld edge is the junction of target segments, there are usually abrupt changes in surface morphology (such as protrusions, depressions, or misalignments), resulting in drastic changes in the normal vector direction (normal vector angle > 30°). Simultaneously, the height difference between the weld area and the targets on both sides is significant (height difference between adjacent point clouds > 0.5 mm, height difference direction perpendicular to the target reference plane). Combining these two characteristics, the candidate weld area can be preliminarily located, namely, a continuous region in the point cloud where the normal vector changes abruptly and the height difference exceeds the threshold, manifesting as an "abnormal zone" in the point cloud.

[0075] Texture image edge enhancement: Canny edge detection is performed on the synchronously acquired target surface texture map to extract the contour line with a sudden change in gray level composed of two-dimensional edge points. The weld edge appears as a clear step edge in the visual image, such as the light-dark boundary or oxide color boundary of the metal target. Through the timestamp and spatial coordinate label attached to each frame image, the contour line is mapped to the point cloud coordinates. The two-dimensional edge points are back-projected to the three-dimensional point cloud space to obtain the three-dimensional candidate weld edge point set, which accurately corresponds to the physical edge of the weld in three-dimensional space, such as the boundary line between the weld and the base material.

[0076] Region growing and morphological filtering: The candidate weld region and the points in the 3D candidate weld edge point set are merged to form a seed point set. These seed points cover both the macroscopic anomaly region at the point cloud level and the microscopic edge features at the image level, ensuring the comprehensiveness and accuracy of the seed points. Based on the spatial proximity and normal vector angle between point clouds, region clustering and growing are performed to gradually expand the weld region. At the same time, morphological opening operations (such as sphere structuring elements with a radius of 2mm) are used to remove noise points and close small gaps, ultimately forming a closed and coherent weld region point cloud set.

[0077] The main direction of the weld in the point cloud set of the weld area is analyzed. Based on the standard process specification, the ideal welding path is preset. The deviation between the main direction of the weld in the point cloud set of the weld area and the preset ideal welding path is calculated point by point, so as to obtain the geometric deviation characteristics of the weld area.

[0078] The deviation includes normal deviation (the distance perpendicular to the preset ideal welding path) and tangential deviation (the offset along the direction of the weld). After the deviation is calculated, the extreme values ​​of the deviation, the average deviation and standard deviation of all deviation points are statistically analyzed and output. These are then integrated with the deviation of each point to obtain the final geometric deviation characteristics.

[0079] Region clustering growth: First, create an empty weld region point cloud set to store the final weld region point cloud; select an unvisited seed point from the seed point set, mark the selected seed point as visited, and add it to the weld region point cloud set; find neighboring points with a radius less than 1 mm (to ensure the compactness of the weld region) in the point cloud data of the entire target shape, and calculate the normal vector angle between each neighboring point and the marked seed point; retain neighboring points with a normal vector angle less than 15 degrees, mark them as visited, and add them to the seed point set and the weld region point cloud set. Repeat the above steps until all seed points in the seed point set have been visited. At this point, the expanded weld region point cloud set is obtained, which includes the weld body and its adjacent transition region.

[0080] Example explanation:

[0081] Assuming the seed point set S contains 100 seed points, in the initial stage of region growing, the first unvisited seed point p1 is selected, marked as visited, and added to the weld region point cloud set W. Next, based on p1, its neighboring points are searched in the point cloud data of the entire target's complete shape. If any neighboring points meet the constraints of spatial proximity and the angle between their normal vectors, these neighboring points are also added to sets S and W; otherwise, they are not included. In subsequent iterations, unvisited seed points are selected from set S, and the above operation is repeated until all points in set S have been visited. The final set W is the weld region point cloud set determined after filtering and expansion.

[0082] Methods for obtaining bevel design parameters include:

[0083] Based on the material of the target welding material, the corresponding coefficient of thermal expansion and thermal conductivity are obtained by querying. Combined with the welding heat input and process parameters, a thermal-structural response analysis model is performed to predict the shrinkage deformation, residual stress and structural warping trend after welding. The preset groove design rules combine the shrinkage deformation, residual stress and structural warping trend after welding with the geometric deviation characteristics of the weld area to form a joint feature vector, which is matched in the preset groove design rules to dynamically generate suitable groove design parameters.

[0084] A generator network is used to analyze the joint feature vector and output bevel design parameter samples. A discriminator network is then used to match and discriminate the bevel design parameter samples according to the preset bevel design rules. The bevel design parameter samples are continuously updated through iterative generation and adversarial training until suitable bevel design parameters are obtained. The bevel design parameters include bevel angle, blunt edge width, gap compensation value, symmetry factor, and melt depth prediction parameters.

[0085] The coefficient of thermal expansion reflects the degree of expansion or contraction of a material when the temperature changes, while thermal conductivity affects heat transfer during welding. Different welding materials have different thermal expansion characteristics, which has a significant impact on groove design. Based on welding process requirements, target material properties, welding quality standards, and past welding experience, pre-defined groove design rules are formulated. These rules typically specify the range and adjustment methods for parameters such as groove angle, groove depth, blunt edge dimensions, and root clearance under different conditions.

[0086] The method for constructing a thermo-mechanical coupling welding process simulation model includes: constructing heat conduction equations and mechanical equations based on heat conduction theory and elastoplastic mechanics theory, and then obtaining a thermo-mechanical coupling model; using the thermo-mechanical coupling model to simulate the temperature field and strain field during the welding process.

[0087] Heat conduction equation: Where ρ is the density of the multi-segment target material; c is the specific heat capacity of the multi-segment target material. Let T be the partial derivative of temperature with respect to time, and t be the rate of change of temperature over time; T is temperature; t is time. Here, is the Nabla operator, representing the vector differential operator; k is the thermal conductivity of the multi-segment target material. For temperature gradient, the rate of temperature change in space; Q is the welding heat input;

[0088] This term represents the net heat outflow rate per unit volume due to heat conduction. It describes the process of heat diffusion in space through conduction and is the core term of the heat conduction equation, determining how the temperature field evolves over time. This indicates that heat is flowing out from that point, causing a decrease in local temperature; This indicates that heat is flowing into that point, causing a local temperature increase;

[0089] The mechanical equation is: in, σ is the divergence of the stress tensor, describing the non-uniformity of stress distribution in space; F is the stress tensor; D is the volume force, an external force acting on a unit volume (such as gravity or inertial force); ep ε is the elastoplastic constitutive matrix, describing the stress-strain relationship of the material during the elastoplastic deformation stage; ε is the strain tensor, a measure of the degree of material deformation; Dep ε is the product of the elastoplastic constitutive matrix and the strain tensor.

[0090] Thermo-mechanical coupling models can accurately simulate the temperature and stress field distribution in the heat-affected zone during welding, helping to predict the magnitude and distribution of residual stress and deformation in the welded structure in advance, facilitating early compensation. Simultaneously, thermo-mechanical coupling models can be used to optimize welding process parameters, reducing the probability of welding defects and improving process stability and product yield. Furthermore, thermo-mechanical coupling models possess good simulation efficiency and engineering adaptability, supporting the simulation analysis of multi-pass and multi-layer welding of complex components, providing online prediction and feedback control basis for intelligent welding systems, and enhancing the system's automation and intelligence capabilities.

[0091] Methods for constructing weld paths with fractal structures include:

[0092] The temperature distribution sequence of the heat-affected zone during the welding process is acquired in real time by a thermal infrared camera to obtain thermal infrared image data; a distributed fiber optic strain sensing system is used to deploy monitoring points in the weld area of ​​the weld area point cloud set to dynamically capture the three-dimensional strain distribution during the welding process, form a strain tensor field, and obtain strain field data.

[0093] The three-dimensional shape of the assembly is geometrically modeled and meshed. The finite element method is used to simulate the thermo-mechanical coupling process of welding based on the thermo-mechanical coupling model. The temperature field is obtained by solving the heat conduction equation. The temperature field is used as a load input to the mechanical equation to solve the stress field. The distribution characteristics of strain concentration area in the heat-affected zone are obtained by analyzing the strain concentration and residual stress distribution under the temperature field.

[0094] It should be noted that the heat conduction equation and the mechanical equation are discretized in both space and time. Spatial discretization uses the finite element method, which discretizes the three-dimensional shape of the assembly into a finite number of elements and nodes. The temperature and displacement of the nodes are expressed as the temperature and displacement of any point within the element through the element interpolation function. Temporal discretization uses the finite difference method, which divides the time domain into several time steps and solves the temperature and stress fields of each time step step by step. The solution result of the heat conduction equation (temperature field) serves as the input to the mechanical model and affects the mechanical properties of the material. The solution result of the mechanical equation (stress field) affects the heat generation and heat transfer during the heat conduction process.

[0095] The maximum principal strain distribution in the strain concentration region distribution characteristics of the heat-affected zone is extracted, and the strain concentration region is identified by the threshold segmentation method to generate a continuous high-strain region. The SIMP method is used for topology optimization modeling, with the minimum strain energy density per unit volume of the structure as the objective function, to construct the fractal structure of the weld path. The fractal structure of the weld path is discretized into welding torch movement path points, thereby obtaining the weld path with fractal structure.

[0096] Methods for obtaining the porosity distribution, ripple amplitude spectrum, and elemental diffusion curves of the joint interface region include:

[0097] Samples were taken and surface-prepared for the interface region of the multi-segment target welding joint. The surface preparation process included grinding, polishing, and ion beam removal of the surface deformation layer. Secondary electron images and backscattered electron images of the interface region of the welding joint at different scales were obtained using a scanning electron microscope. An EBSD probe was equipped on the scanning electron microscope platform to perform crystallographic orientation analysis on the interface region of the welding joint and obtain the grain orientation distribution map of the region.

[0098] Image stitching and multi-scale enhancement algorithms are used to construct a panoramic image of the welded joint interface containing structural information at the micron and nanoscale. The pore regions of the panoramic image of the welded joint interface are identified to obtain the pore distribution of the joint interface region.

[0099] Boundary tracking and extraction of the ripple profile in the panoramic image of the welded joint interface are performed, and the ripple amplitude spectrum is obtained by Fourier transform analysis. Using spectral analysis technology, cross-sectional scanning paths and surface scanning areas are set on the multi-segment target weld interface and its two sides to obtain spatial distribution data of different metal elements in the welded joint interface area. Based on the spatial distribution data, element diffusion curves of element concentration as a function of spatial location are generated.

[0100] Methods for designing gradient transition layer structures at welding interfaces include:

[0101] Based on the pore distribution data identified in the joint interface region, the pore density and size are identified, and a buffer weld bead is designed. By remelting the material and interlayer staggering, the filling density of the pore distribution region is improved, resulting in a pore density field. Micro-concave / gentle slope corrugated structures are arranged in the fluctuating frequency abrupt region of the corrugation amplitude spectrum to release thermal stress along the structural morphology direction, resulting in a corrugation morphology field. Based on the slope of the element diffusion curve, an M-layer alloy combination is designed, with the material composition of each layer gradually changing, resulting in a composition concentration field. A parameter field-driven model based on a neural network is used to fuse the pore density field, corrugation morphology field, and composition concentration field to generate a three-dimensional gradient transition structure map, resulting in a gradient transition layer structure of the welding interface.

[0102] Methods for generating welding part design schemes applicable to multi-segment target welding include:

[0103] Based on the dynamically adapted bevel design parameters, the weld path of the fractal structure, and the gradient transition layer structure of the welding interface, a design scheme for the welding part that can be used for multi-segment target welding is generated.

[0104] By using 3D CAD software, a structured welding part model is established based on dynamically adapted bevel design parameters, fractal weld paths, and gradient transition layers at the welding interface, achieving a synergistic expression of geometric structure and thermo-mechanical properties. The structured welding part model is then converted into a set of instructions recognizable by the welding robot using a CNC programming interface, including welding torch movement trajectory, energy input parameters, and multi-layer welding sequence, thus forming a digital design scheme for welding parts used in multi-segment target welding, ensuring comprehensive coordination of welding accuracy, structural strength, and interface performance.

[0105] The preset point cloud height difference threshold is set by the staff. By collecting different point cloud height differences, the average value of multiple point cloud height differences is taken as the preset point cloud height difference threshold; similarly, a preset error threshold is set.

[0106] In this embodiment, weld seams are located from two dimensions: point cloud data and texture image, through two methods: initial screening of weld seam candidate areas and edge enhancement of texture images. These two methods complement each other, improving the accuracy and reliability of weld seam location.

[0107] A thermo-mechanical coupling model is constructed based on the theories of heat conduction and elastoplastic mechanics. By solving the heat conduction equation and the mechanical equation, the temperature field and stress field distribution in the heat-affected zone during the welding process can be accurately simulated. This helps to predict the magnitude and distribution of residual stress and deformation of the structure after welding, and provides a theoretical basis for the optimization of welding process.

[0108] By simulating the welding process under different parameters using a thermo-mechanical coupling model, the changes in temperature and stress fields are analyzed to find the optimal welding process parameters, reduce the probability of welding defects, and improve process stability and product qualification rate. The model has good simulation efficiency and engineering adaptability, and can support the simulation analysis of multi-pass welding and multi-layer welding of complex components. It provides online prediction and feedback control basis for intelligent welding systems, enhances the automation and intelligence capabilities of the system, and improves welding quality and efficiency.

[0109] The weld path design method based on actual welding process data and simulation analysis can optimize the weld path, reduce stress concentration during welding, and improve welding quality. By rationally designing the gradient transition layer structure of the welding interface, stress concentration at the weld joint interface can be effectively alleviated, interface defects can be reduced, and the strength, toughness, and corrosion resistance of the weld joint can be improved, thus extending the service life of the welded structure.

[0110] The gradient transition layer structure enables tighter bonding between different materials, reduces uneven element diffusion at the interface, thereby improving welding quality and ensuring the reliability of the welded structure under complex working conditions. It can quickly and accurately generate digital design schemes for welding parts that meet actual needs, reducing the time and errors of manual design and improving design efficiency and quality.

[0111] Example 2

[0112] Please see Figure 2 As shown in the figure, this embodiment of a welding part design system for multi-segment target welding includes:

[0113] The 3D topography acquisition module scans the multi-segment target material using laser scanning and visible light imaging equipment to acquire the assembled 3D topography of the multi-segment target material; based on the assembled 3D topography, it identifies the weld area and extracts the geometric deviation features of the weld area.

[0114] The bevel parameter generation module generates dynamically adapted bevel design parameters based on the geometric deviation characteristics of the weld area and the thermal expansion characteristics of the welding material, according to the preset bevel design rules.

[0115] The weld path generation module collects thermal infrared image data and strain field data during the welding process, constructs a thermo-mechanical coupled welding process simulation model, and obtains the distribution characteristics of the strain concentration area in the heat-affected zone by analyzing the strain concentration and residual stress distribution in the heat-affected zone during the welding process, thereby constructing a weld path with a fractal structure.

[0116] The microstructure analysis module, based on scanning electron microscopy and electron backscatter diffraction, and combined with cross-scale image fusion methods, analyzes the microstructure and element diffusion characteristics of the weld joint interface region, and obtains the pore distribution, ripple amplitude spectrum and element diffusion curve of the joint interface region.

[0117] The interface structure design module designs a gradient transition layer structure for the welding interface based on the pore distribution, ripple amplitude spectrum, and element diffusion curves of the joint interface region, enabling a continuous transition of the welding interface performance.

[0118] The welding scheme integration module generates welding part design schemes that can be used for multi-segment target welding based on dynamically adapted groove design parameters, fractal weld paths, and gradient transition layer structures at the welding interface.

[0119] Since the electronic device described in this embodiment is the electronic device used to implement the welding part design method and system for multi-segment target welding in the embodiments of this application, those skilled in the art can understand the specific implementation and various variations of the electronic device in this embodiment based on the welding part design method and system for multi-segment target welding described in the embodiments of this application. Therefore, how the electronic device implements the method in the embodiments of this application will not be described in detail here. As long as those skilled in the art implement the electronic device used in the welding part design method and system for multi-segment target welding in the embodiments of this application, it falls within the scope of protection of this application.

[0120] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0121] The above description is merely a preferred embodiment of the present invention, and the scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for users of ordinary technical skills, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for designing welding positions for multi-segment target welding, characterized in that, include: S1. The multi-segment target material is scanned using laser scanning and visible light imaging equipment to obtain the three-dimensional morphology of the multi-segment target material assembly. Based on the three-dimensional shape of the assembly, the weld area is identified, and the geometric deviation features of the weld area are extracted. S2. Based on the geometric deviation characteristics of the weld area and the thermal expansion characteristics of the welding material, dynamically adapted groove design parameters are generated according to the preset groove design rules. S3. Collect thermal infrared image data and strain field data during the welding process, construct a thermo-mechanical coupling welding process simulation model, and obtain the distribution characteristics of the strain concentration area in the heat-affected zone by analyzing the strain concentration and residual stress distribution in the heat-affected zone during the welding process, and then construct a weld path with a fractal structure. S4. Based on scanning electron microscopy and electron backscatter diffraction, and combined with cross-scale image fusion method, the microstructure and element diffusion characteristics of the welded joint interface region are analyzed, and the pore distribution, ripple amplitude spectrum and element diffusion curve of the joint interface region are obtained. S5. Based on the pore distribution, ripple amplitude spectrum and element diffusion curve of the joint interface region, design the gradient transition layer structure of the welding interface to achieve a continuous transition of the welding interface performance. S6. Based on the dynamically adapted groove design parameters, the weld path of the fractal structure, and the gradient transition layer structure of the welding interface, generate a design scheme for the welding part that can be used for multi-segment target welding.

2. The method for designing welding positions for multi-segment target welding according to claim 1, characterized in that, The method for obtaining the three-dimensional morphology of the multi-segment target assembly includes: The principle of line laser triangulation is adopted, a laser scanning device is selected, and a visible light imaging device is equipped. The multi-segment target is installed on a clamping platform with stable support and multi-axis motion control, and the multi-segment target is scanned using the laser scanning device and the visible light imaging device. Based on the preset scanning path of the multi-segment target's geometry, the laser scanning device is controlled to emit a laser beam along the preset scanning path to scan the target surface point by point, and the point cloud data of the complete shape of the target is obtained by receiving reflected light; while the laser scanning is being performed, a visible light imaging device is used to simultaneously capture images of the target surface texture, and each frame of the image is accompanied by a timestamp and a spatial coordinate label consistent with the laser scanning device. By iterating the nearest point ICP algorithm, the point cloud data in each direction are stitched together and aligned to the camera coordinate system; by using nearest neighbor interpolation, the RGB values ​​of the target surface texture image are mapped to the corresponding point cloud data, thereby generating the assembled three-dimensional shape of the multi-segment target.

3. The method for designing welding positions for multi-segment target welding according to claim 2, characterized in that, The method for extracting the geometric deviation features of the weld area includes: By solving the local neighborhood covariance matrix, the normal vectors of all points in the point cloud data of the complete target shape are obtained; the regions of abrupt change in normal vectors are extracted, and the candidate regions of welds are located in combination with the preset point cloud height difference threshold. Canny edge detection is performed on the synchronously acquired target surface texture map to extract the contour line composed of two-dimensional edge points with abrupt gray-level gradient changes. The contour line is mapped to point cloud coordinates by the timestamp and spatial coordinate label attached to each frame image, and the two-dimensional edge points are back-projected to the three-dimensional point cloud space to obtain the three-dimensional candidate weld edge point set. Region growing and morphological filtering: Merge the points in the candidate weld region and the set of 3D candidate weld edge points to form a seed point set; perform region clustering growth based on the spatial proximity and normal vector angle between point clouds to gradually expand the weld region; at the same time, use morphological opening operation to remove noise points and close small gaps, and finally form a set of weld region point clouds. The main direction of the weld in the point cloud set of the weld area is analyzed. Based on the standard process specification, the ideal welding path is preset. The deviation between the main direction of the weld in the point cloud set of the weld area and the preset ideal welding path is calculated point by point, so as to obtain the geometric deviation characteristics of the weld area.

4. The method for designing welding positions for multi-segment target welding according to claim 3, characterized in that, The methods for obtaining the bevel design parameters include: Based on the material of the target welding material, the corresponding coefficient of thermal expansion and thermal conductivity are obtained by querying. Combined with the welding heat input and process parameters, a thermal-structural response analysis model is performed to predict the shrinkage deformation, residual stress and structural warping trend after welding. The preset groove design rules combine the shrinkage deformation, residual stress and structural warping trend after welding with the geometric deviation characteristics of the weld area to form a joint feature vector, which is matched in the preset groove design rules to dynamically generate suitable groove design parameters.

5. The method for designing welding positions for multi-segment target welding according to claim 4, characterized in that, The method for constructing the thermo-mechanical coupling welding process simulation model includes: constructing heat conduction equations and mechanical equations based on heat conduction theory and elastoplastic mechanics theory, and then obtaining the thermo-mechanical coupling model; using the thermo-mechanical coupling model to simulate the temperature field and strain field during the welding process.

6. The method for designing welding positions for multi-segment target welding according to claim 5, characterized in that, The method for constructing a weld path with a fractal structure includes: The temperature distribution sequence of the heat-affected zone during the welding process is acquired in real time by a thermal infrared camera to obtain thermal infrared image data; a distributed fiber optic strain sensing system is used to deploy monitoring points in the weld area of ​​the weld area point cloud set to dynamically capture the three-dimensional strain distribution during the welding process, form a strain tensor field, and obtain strain field data. The three-dimensional shape of the assembly is geometrically modeled and meshed. The finite element method is used to simulate the thermo-mechanical coupling process of welding based on the thermo-mechanical coupling model. The temperature field is obtained by solving the heat conduction equation. The temperature field is used as a load input to the mechanical equation to solve the stress field. The distribution characteristics of strain concentration area in the heat-affected zone are obtained by analyzing the strain concentration and residual stress distribution under the temperature field. The maximum principal strain distribution in the strain concentration region distribution characteristics of the heat-affected zone is extracted, and the strain concentration region is identified by the threshold segmentation method to generate a continuous high-strain region. The SIMP method is used for topology optimization modeling, with the minimum strain energy density per unit volume of the structure as the objective function, to construct the fractal structure of the weld path. The fractal structure of the weld path is discretized into welding torch movement path points, thereby obtaining the weld path with fractal structure.

7. The method for designing welding positions for multi-segment target welding according to claim 6, characterized in that, The method for obtaining the porosity distribution, ripple amplitude spectrum, and element diffusion curves of the joint interface region includes: Samples were taken and surface-prepared for the interface region of the multi-segment target welding joint. The surface preparation process included grinding, polishing, and ion beam removal of the surface deformation layer. Secondary electron images and backscattered electron images of the interface region of the welding joint at different scales were obtained using a scanning electron microscope. An EBSD probe was equipped on the scanning electron microscope platform to perform crystallographic orientation analysis on the interface region of the welding joint and obtain the grain orientation distribution map of the region. Image stitching and multi-scale enhancement algorithms are used to construct a panoramic image of the welded joint interface containing structural information at the micron and nanoscale. The pore regions of the panoramic image of the welded joint interface are identified to obtain the pore distribution of the joint interface region. Boundary tracking and extraction of the ripple profile in the panoramic image of the welded joint interface are performed, and the ripple amplitude spectrum is obtained by Fourier transform analysis. Using spectral analysis technology, cross-sectional scanning paths and surface scanning areas are set on the multi-segment target weld interface and its two sides to obtain spatial distribution data of different metal elements in the welded joint interface area. Based on the spatial distribution data, element diffusion curves of element concentration as a function of spatial location are generated.

8. The method for designing welding positions for multi-segment target welding according to claim 7, characterized in that, The method for designing the gradient transition layer structure of the welding interface includes: Based on the pore distribution data identified in the joint interface region, the pore density and size are identified, and a buffer weld bead is designed. By remelting the material and interlayer staggering, the filling density of the pore distribution region is improved, resulting in a pore density field. Micro-concave / gentle slope corrugated structures are arranged in the fluctuating frequency abrupt region of the corrugation amplitude spectrum to release thermal stress along the structural morphology direction, resulting in a corrugation morphology field. Based on the slope of the element diffusion curve, an M-layer alloy combination is designed, with the material composition of each layer gradually changing, resulting in a composition concentration field. A parameter field-driven model based on a neural network is used to fuse the pore density field, corrugation morphology field, and composition concentration field to generate a three-dimensional gradient transition structure map, resulting in a gradient transition layer structure of the welding interface.

9. The method for designing welding positions for multi-segment target welding according to claim 8, characterized in that, The method for generating welding part design schemes that can be used for multi-segment target welding includes: Based on the dynamically adapted groove design parameters, the weld path of the fractal structure, and the gradient transition layer structure of the welding interface, a design scheme for the welding part that can be used for multi-segment target welding is generated. A structured welding part model is established using 3D CAD software based on dynamically adapted bevel design parameters, fractal weld paths, and gradient transition layers at the welding interface. The structured welding part model is then converted into a set of instructions that the welding robot can recognize using a CNC programming interface, thus forming a digital design scheme for welding parts that can be used for multi-segment target welding.

10. A welding part design system for multi-segment target welding, used to implement the welding part design method for multi-segment target welding as described in any one of claims 1 to 9, characterized in that, include: The three-dimensional morphology acquisition module scans the multi-segment target material using laser scanning and visible light imaging equipment to acquire the assembled three-dimensional morphology of the multi-segment target material. Based on the three-dimensional shape of the assembly, the weld area is identified, and the geometric deviation features of the weld area are extracted. The bevel parameter generation module generates dynamically adapted bevel design parameters based on the geometric deviation characteristics of the weld area and the thermal expansion characteristics of the welding material, according to the preset bevel design rules. The weld path generation module collects thermal infrared image data and strain field data during the welding process, constructs a thermo-mechanical coupled welding process simulation model, and obtains the distribution characteristics of the strain concentration area in the heat-affected zone by analyzing the strain concentration and residual stress distribution in the heat-affected zone during the welding process, thereby constructing a weld path with a fractal structure. The microstructure analysis module, based on scanning electron microscopy and electron backscatter diffraction, and combined with cross-scale image fusion methods, analyzes the microstructure and element diffusion characteristics of the weld joint interface region, and obtains the pore distribution, ripple amplitude spectrum and element diffusion curve of the joint interface region. The interface structure design module designs a gradient transition layer structure for the welding interface based on the pore distribution, ripple amplitude spectrum, and element diffusion curves of the joint interface region, enabling a continuous transition of the welding interface performance. The welding scheme integration module generates welding part design schemes that can be used for multi-segment target welding based on dynamically adapted groove design parameters, fractal weld paths, and gradient transition layer structures at the welding interface.

Images