A method for optimizing design of a product forming process of upper and lower buckle half-shells

Abstract

The application relates to the field of stamping forming process optimization, and discloses a forming process optimization design method for upper and lower buckle half-shell products, which comprises the following steps: obtaining design data of the upper and lower buckle half-shell products and discretizing the design data into a set of basic manufacturing features, combining process constraint parameters to establish a multi-dimensional correlation table; constructing a process feature conflict matrix, outputting an initial process topology graph, a multi-process execution sequence and an initial mold grid; performing multi-process forming simulation, obtaining an out-of-tolerance node set and mechanical state data, constructing a reverse displacement compensation field and performing compensation pre-evaluation; when it is determined that there is a forming safety risk, the process feature conflict matrix is reversely corrected to decouple and reconstruct the process topology graph and to re-simulate; when there is no risk, multi-process compensation distribution is performed based on the compensation field and iterative calculation is driven until the size deviation and the forming risk index both meet the safe convergence condition, the iterative convergence result is locked, and a process guidance file is generated. The application realizes the collaborative optimization of the process execution sequence and the mold surface displacement compensation.

Description

Technical Field

[0001] This invention relates to the field of stamping process optimization, specifically to a method for optimizing the forming process of a half-shell product with upper and lower snap-fit ​​parts. Background Technology

[0002] In the multi-stage molding process of half-shell products, process route planning and springback compensation are crucial factors determining the final molding quality. In the early stages of molding process design, existing methods often lack quantitative analysis of the geometric and physical states of adjacent manufacturing features and machine tool parameters. This makes it difficult to identify and eliminate theoretically unmergeable feature combinations early on, leading to spatial interference or equipment overload during subsequent verification or actual production, thus prolonging the overall optimization cycle of the molding process.

[0003] When compensating for dimensional deviations caused by multi-pass stamping, existing technologies typically do not take into account the internal stress release characteristics of metal materials during the unloading process of multi-stage forming. Because the springback compensation weight is not allocated according to the proportion of local elastic strain energy actually released in each stage, the accuracy of the global springback amount distribution to the mold surfaces of each forming stage is poor, making it difficult for the final product dimensions to consistently meet the preset process tolerance requirements.

[0004] Furthermore, when reverse displacement compensation is applied to the mold surface, the change in surface morphology can easily induce secondary forming defects such as local cracking or wrinkling. Faced with the forming safety risks caused by displacement compensation, existing technologies are mainly limited to repeated local fine-tuning of the mold surface. They cannot trace the damage contribution rate of each process based on multi-dimensional mechanical state data, and it is difficult to automatically decompose high-risk merged processes by tracing back to the process route level. This limitation of relying solely on mold surface adjustment leads to a disconnect between the process topology and springback compensation design, ultimately resulting in frequent physical mold repairs during the actual trial molding stage, increasing the trial-and-error costs of the mold. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an optimized design method for the molding process of top and bottom snap-on half-shell products. This method solves the problem that in the multi-process molding of top and bottom snap-on half-shell products, the process route planning and springback compensation design are disconnected, which easily leads to molding defects such as local cracking or wrinkling when performing mold surface displacement compensation. Furthermore, it is difficult to automatically adjust the process topology after defects occur, resulting in long mold design cycles and high trial and error costs.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an optimized design method for the molding process of a snap-fit ​​semi-shell product, comprising the following steps:

[0007] The design data of the top and bottom snap-fit ​​half-shell product is obtained and discretized into a basic manufacturing feature set, and a multi-dimensional association table is established in combination with the preset process constraint parameters;

[0008] Construct a process feature conflict matrix based on the basic manufacturing feature set, output the initial process topology diagram and multi-process execution sequence, and generate the initial mold mesh by combining it with a multi-dimensional association table;

[0009] Based on the initial process topology diagram, multi-process execution sequence and initial mold mesh, multi-process forming simulation is performed to obtain the set of out-of-tolerance nodes and mechanical state data;

[0010] Based on the set of out-of-tolerance nodes and mechanical state data, a reverse displacement compensation field is constructed and a compensation pre-evaluation is performed. When a forming safety risk is determined, the process feature conflict matrix is ​​reverse-corrected to decouple and reconstruct the process topology and multi-process forming simulation is re-executed.

[0011] When it is determined that there is no forming safety risk, multi-process compensation allocation is performed based on the reverse displacement compensation field, and iterative calculation is driven until the preset process tolerance is met to obtain the iterative convergence result;

[0012] Lock in the iterative convergence results and generate the process guidance document.

[0013] This invention establishes a physical correlation between forming risk and process topology through multi-process simulation and reverse compensation allocation. In the compensation pre-evaluation stage, strain changes caused by displacement compensation are obtained using small-step pre-calculation. When a forming defect risk is identified, mechanical state data is extracted to calculate the damage contribution rate of each process, and the process characteristic conflict matrix is ​​reverse-corrected. This decouples high-risk merged processes at the process route level, overcoming the limitations of simply adjusting the mold surface. In the absence of forming risk, based on the proportion of elastic strain energy released during the unloading stage of each process, the global springback compensation is allocated to the mold surface of the corresponding process, completing the collaborative optimization of the execution sequence and mold surface dimensions.

[0014] Preferably, the design data of the top and bottom snap-fit ​​semi-shell product is obtained and discretized into a basic manufacturing feature set, and a multi-dimensional association table is established in combination with preset process constraint parameters, including:

[0015] Based on the geometric topology and preset stamping direction vector of the 3D CAD model in the design data, the features of the drawing surface, sidewall cutting edge, hole position, side cutting, and side finishing are identified and extracted, and combined to construct a basic manufacturing feature set. The process constraint parameters of the target production equipment are obtained, and the preset process tolerance is obtained. The basic manufacturing feature set is associated and bound with the mesh binding identifier, candidate process type, and process constraint parameters, and written together into a multidimensional association table covering the basic manufacturing feature set.

[0016] Preferably, a process feature conflict matrix is ​​constructed based on the basic manufacturing feature set, including: for any two adjacent basic manufacturing features, calculating the minimum geodesic distance, the angle between normal vectors, the tonnage required for merging processes, and the mold space requirement, to obtain interference evaluation parameters; and calling the interference evaluation parameters to execute the following four independent constraint judgment formulas:

[0017] If the minimum geodetic distance is less than the set distance safety threshold, the distance constraint formula is valid; if the included angle of the normal vector is greater than the set normal angle tolerance, the angle constraint formula is valid; if the tonnage required for the merged process is greater than the set allowable tonnage of the equipment, the load constraint formula is valid; if the mold space requirement exceeds the set allowable closed height of the target machine tool, the worktable size, the lateral mechanism layout space, or the allowable space interference threshold, the space constraint formula is valid.

[0018] If any one of the four independent constraint judgment formulas is true, it is determined that there is physical interference between two adjacent basic manufacturing features, and the corresponding element in the process feature conflict matrix is ​​assigned a value of 1; if none of the four independent constraint judgment formulas are true, it is determined that two adjacent basic manufacturing features are theoretically allowed to be executed together, and the corresponding element in the process feature conflict matrix is ​​assigned a value of 0. The process feature conflict matrix is ​​constructed by traversing all basic manufacturing feature combinations consisting of any two adjacent basic manufacturing features.

[0019] Preferably, the initial process topology diagram and multi-process execution sequence are output, and an initial mold mesh is generated by combining it with a multi-dimensional association table, including:

[0020] The system reads the preset locking matrix, process feature conflict matrix, and process constraint parameters, outputs the initial process topology diagram and multi-process execution sequence, and generates the initial mold surface based on the multi-dimensional association table according to the multi-process execution sequence and design data, and performs meshing on the initial mold surface to form the initial mold mesh.

[0021] Preferably, the set of out-of-tolerance nodes is obtained, including:

[0022] Using the initial mold mesh as the mold geometric boundary and based on the logical association of the initial process topology diagram, the physical boundary conditions of the simulation process are set. Based on the physical boundary conditions, the internally configured finite element solver is called to start the multi-process forming simulation according to the multi-process execution sequence until the simulation ends, obtaining the mesh model of the final state of the upper and lower half-shell product. The nominal model is extracted based on the design data. After comparing the node coordinates of the mesh model of the final state of the upper and lower half-shell product with the corresponding node coordinates of the nominal model, the node normal deviation is calculated. It is determined whether the absolute value of the node normal deviation of each node is greater than the dimensional tolerance requirement in the preset process tolerance, and the nodes that are greater than the dimensional tolerance requirement are selected. The selected nodes constitute the set of out-of-tolerance nodes.

[0023] Preferably, constructing a reverse displacement compensation field based on the set of out-of-tolerance nodes and mechanical state data, and performing a compensation pre-evaluation, includes:

[0024] Spatial clustering is performed based on the set of out-of-tolerance nodes and mechanical state data to delineate risk areas. Displacement vectors in the reverse displacement compensation field are obtained based on the node normal deviation mapping. The displacement vectors are combined with preset step-size relaxation coefficients and applied to the mold mesh or mold surface of the corresponding process for small-step pre-calculation. Based on the local mesh plastic strain change caused by the pre-calculation, forming risk indicators including local cracking risk indicators and local wrinkling risk indicators are calculated to complete the compensation pre-evaluation. When the local cracking risk indicator or local wrinkling risk indicator indicates that there is forming safety risk in the corresponding risk area, and the forming safety risk cannot be eliminated by adjusting the set forming control parameters, the process contribution rate is triggered to drive the reverse correction of the process feature conflict matrix.

[0025] Preferably, the process contribution rate is triggered by the following:

[0026] For each process in the multi-process execution sequence, the first principal strain increment, the thickness reduction rate increment, and the springback release displacement increment generated in each process stage are directly extracted. The normalized damage increment is obtained by performing weighted processing on the first sensitivity weight factor corresponding to the first principal strain increment, the second sensitivity weight factor corresponding to the thickness reduction rate increment, and the third sensitivity weight factor corresponding to the springback release displacement increment. Based on the normalized damage increment corresponding to each process, the process contribution rate of each process to the risk area is calculated and determined.

[0027] Preferably, the reverse correction of the process feature conflict matrix to decouple and reconstruct the process topology graph includes:

[0028] When the contribution rate of a process exceeds the preset threshold for a single process to dominate, and the corresponding process node is a merged process node, and the merged process node corresponds to at least two basic manufacturing features, the decoupling judgment condition is met. When the decoupling judgment condition is met, the basic manufacturing feature combination corresponding to the merged process node is located, the state of the corresponding element in the process feature conflict matrix is ​​overwritten to 1, and the state of the corresponding element in the preset locking matrix is ​​simultaneously assigned to 1, thereby controlling the decoupling and reconstructing of the process topology.

[0029] Preferably, multi-process compensation allocation is performed based on the reverse displacement compensation field, including:

[0030] Extract the local elastic strain energy released during the unloading stage of each process. Based on the ratio of the local elastic strain energy released during the unloading stage of the corresponding process to the sum of the local elastic strain energy released during the unloading stages of all processes, determine the springback compensation weight of the corresponding process. Combine the springback compensation weight with the preset global compensation relaxation coefficient and project it along the local normal of the mold to update the spatial coordinates of the corresponding node in the mold mesh of the corresponding process.

[0031] Preferably, the process involves obtaining and locking the iterative convergence results, and generating a process guidance document, including:

[0032] Based on the ratio of the current maximum node normal deviation to the maximum node normal deviation of the previous iteration step, the global compensation relaxation coefficient in the set forming control parameters is adaptively scaled and adjusted; the mold mesh and forming control parameters after updating the node coordinates are rewritten as the solver input file, and the multi-process forming simulation is re-executed for the next round of iteration calculation until the set of out-of-tolerance nodes is empty and the local cracking risk index in the forming risk index does not exceed the set first critical safety factor and the local wrinkling risk index does not exceed the set second critical safety factor, so as to obtain the iteration convergence result; the final process topology diagram, mold mesh and forming control parameters corresponding to the iteration convergence result are locked, and the generated auxiliary process data and equipment information are integrated and compiled to generate a process guidance document.

[0033] This invention provides an optimized design method for the molding process of a top-and-bottom snap-fit ​​semi-shell product. It offers the following advantages:

[0034] 1. This invention performs small-step pre-calculation during the compensation pre-evaluation stage. When forming safety risks are detected, it calculates the damage contribution rate of each process based on multi-dimensional mechanical state data and reversely corrects the process characteristic conflict matrix to decouple and reconstruct the process topology. When faced with local cracking or wrinkling problems caused by displacement compensation, this method can automatically split high-risk merged processes directly at the process route level, eliminating the limitations of simply relying on mold surface adjustments and reducing the frequency of physical mold repair and trial-and-error costs during the actual mold trial stage.

[0035] 2. This invention extracts the local elastic strain energy released during the unloading stage of multiple processes, determines the springback compensation weight of each process according to the ratio of the elastic strain energy released in each process to the total released amount, and then performs multi-process compensation allocation based on the reverse displacement compensation field. This process conforms to the stress release law of metallic materials in multi-pass forming, rationally distributes the global springback amount to the mold surface of different forming stages, improves the accuracy of mold surface displacement compensation, and ensures that the dimensional tolerance of the final product meets the standards.

[0036] 3. This invention calculates the minimum geodesic distance, normal vector angle, required tonnage, and mold space requirements for adjacent basic manufacturing features. It then performs four independent constraint checks on distance, angle, load, and space to construct a process feature conflict matrix and output an initial process topology diagram. This step quantitatively checks geometric and physical interferences, stress limits, and actual machine tool parameters in the early design stage, eliminating theoretically unmergeable feature combinations. This prevents equipment interference or overload during subsequent execution of the planned process sequence, shortening the overall optimization cycle of the molding process. Attached Figure Description

[0037] Figure 1 This is a framework diagram of the molding process optimization design system for the upper and lower snap-fit ​​semi-shell product of the present invention.

[0038] Figure 2 A flowchart of the optimized design method for the molding process of the upper and lower snap-fit ​​semi-shell product of the present invention;

[0039] Figure 3 This is a flowchart of the initial process topology generation for the present invention;

[0040] Figure 4 This is a flowchart of the topology decoupling evaluation and reconstruction process of the present invention;

[0041] Figure 5 This is a flowchart of the parameter iteration and mesh compensation process of the present invention;

[0042] Figure 6 This is a diagram of the compensation iteration convergence curve according to an embodiment of the present invention;

[0043] Figure 7 This is a risk area process contribution rate distribution diagram according to an embodiment of the present invention;

[0044] Figure 8 This is a comparison diagram of the forming limit state according to an embodiment of the present invention. Detailed Implementation

[0045] The technical solutions in 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.

[0046] Reference Figure 1 This invention provides a system for optimizing the molding process of a top and bottom snap-fit ​​semi-shell product. The system includes a model analysis module, a topology generation module, a simulation calculation module, a decoupling evaluation module, a parameter control module, and a file output module.

[0047] The model parsing module is used to acquire and parse the 3D CAD model of the upper and lower half-shell product, forming basic manufacturing features and their process constraint data; the topology generation module is used to generate a process topology diagram based on the basic manufacturing features; the simulation calculation module is used to perform multi-process forming simulation and output deviation and forming state data; the decoupling evaluation module is used to determine whether process topology needs to be decoupled based on simulation results; the parameter control module is used to perform mold compensation and forming control parameter iteration when decoupling is not triggered; and the file output module is used to output process guidance documents and manufacturing data after iteration convergence.

[0048] Each computing module runs under processor scheduling, and through process topology generation, forming simulation, risk assessment, topology feedback and mold compensation iteration, a process optimization design flow for the top and bottom half-shell products is formed.

[0049] Reference Figure 2 This invention provides a method for optimizing the molding process of a snap-fit ​​half-shell product. This method is executed by the aforementioned system for optimizing the molding process of snap-fit ​​half-shell products and mainly includes six stages: design data parsing, process topology generation, multi-process forming simulation, topology decoupling evaluation, mold compensation iteration, and process document output. The method comprises the following steps:

[0050] S10. Feature Extraction and Association Construction. The model parsing module acquires the design data of the top and bottom half-shell product. The design data includes the 3D CAD (Computer-Aided Design) model and the nominal model data. The model parsing module analyzes the geometric topology of the 3D CAD model and discretizes the continuous surfaces contained in the 3D CAD model into a basic manufacturing feature set. The basic manufacturing features in the basic manufacturing feature set include drawn surface features, sidewall cutting edge features, side cutting features, side finishing features, and hole position features.

[0051] Subsequently, the model parsing module synchronously acquires the process constraint parameters and preset process tolerances of the target production equipment used to produce the semi-shell product, establishes a multi-dimensional association table covering the basic manufacturing feature set, the mesh binding identifier corresponding to the CAE (Computer-Aided Engineering) mesh to be generated, the candidate process type corresponding to the basic manufacturing features, and the process constraint parameters, and transmits it to the topology generation module. The mesh binding identifier can be any one of the feature mapping region, node index rule, and cell set identifier.

[0052] S20. Initial Process Topology Generation. The topology generation module receives a multidimensional association table and constructs a process feature conflict matrix based on the basic manufacturing feature set. Then, based on the candidate process type, it calculates the interference evaluation parameters of adjacent basic manufacturing features. The interference evaluation parameters include the minimum geodesic distance, the angle between normal vectors, the tonnage required for merging processes, and the mold space requirement.

[0053] The topology generation module determines the element states of the process feature conflict matrix based on interference evaluation parameters. Combined with a preset locking matrix, the topology generation module outputs an initial process topology diagram that conforms to the process constraint parameters, generates a multi-process execution sequence, and, based on the basic manufacturing features and corresponding candidate process types in the multi-dimensional association table, generates initial die surfaces corresponding to different stamping stages, and meshes the initial die surfaces to form the initial die mesh. Finally, the topology generation module imports the initial process topology diagram, multi-process execution sequence, and initial die mesh into the simulation calculation module.

[0054] S30. Multi-process Simulation and Deviation Extraction. The simulation calculation module receives the initial process topology diagram, the multi-process execution sequence, and the initial mold mesh. Using the initial mold mesh as the mold geometric boundary and based on the logical connections of the initial process topology diagram, it performs multi-process forming simulation according to the multi-process execution sequence. After the simulation is completed, the node coordinates of the mesh model in the final state of the semi-shell product are extracted and compared with the corresponding node coordinates of the nominal model in the design data. Based on the comparison results, the simulation calculation module calculates the node normal deviation, filters out the set of nodes with out-of-tolerance, and simultaneously extracts the mechanical state data during the forming process. The mechanical state data includes the first principal strain, the second principal strain, the thickness reduction rate, out-of-plane displacement data, the springback release displacement data generated during the unloading stage of each process in the multi-process execution sequence, and the residual stress distribution after each process. This data is then submitted to the decoupling evaluation module.

[0055] S40. Topology Decoupling Assessment and Reconstruction. The decoupling assessment module receives node normal deviation and mechanical state data. It takes the out-of-tolerance nodes in the out-of-tolerance node set and the forming risk nodes in the mechanical state data whose thickness reduction rate exceeds the preset safety threshold (e.g., 15% to 25%) as the node set to be clustered, performs spatial clustering to form risk regions, and constructs a reverse displacement compensation field.

[0056] Subsequently, the decoupling assessment module combines the mechanical state data to perform a compensation pre-evaluation, calculates forming risk indicators, including local cracking risk indicators and local wrinkling risk indicators, and determines whether there is forming safety risk in the risk area based on the forming risk indicators.

[0057] When there is a forming safety risk in the risk area and the forming safety risk cannot be eliminated by adjusting the forming control parameters, the decoupling evaluation module calculates the process contribution rate of each process in the multi-process execution sequence, and judges whether the decoupling judgment condition is met based on the process contribution rate. When the decoupling judgment condition is met, the element states of the process feature conflict matrix and the element states of the locking matrix are reversed, the current compensation pre-evaluation process is terminated, the control topology generation module decouples and reconstructs the process topology diagram, and feeds the decoupled process topology diagram back to the simulation calculation module to re-execute the multi-process forming simulation. If decoupling is not triggered, the parameter control module is activated.

[0058] S50, Parameter Iteration and Mesh Compensation. The parameter control module performs multi-process compensation allocation for the reverse displacement compensation field that has passed the forming risk verification. Based on the proportion of elastic strain energy released during the unloading stage of each process, the springback compensation weight of each process in the multi-process execution sequence for the node normal deviation is calculated, and the node coordinates of the mold mesh are updated according to the compensation mapping relationship. The process contribution rate is evaluated in the previous steps and is only used to determine whether topology decoupling is triggered. At the same time, the forming control parameters are adjusted synchronously. The forming control parameters include blank holder force, drawbead resistance, local mold fillet, step size relaxation coefficient, and global compensation relaxation coefficient. The mold mesh with updated node coordinates and forming control parameters are rewritten as the solver input file, driving the simulation calculation module to perform the next round of iteration until the set of out-of-tolerance nodes is empty and the local cracking risk index in the forming risk index does not exceed the set first critical safety factor and the local wrinkling risk index does not exceed the set second critical safety factor. Then the file output module is triggered.

[0059] S60. Process document output and mapping to the physical manufacturing system. The document output module obtains the iterative convergence results and locks the final process topology diagram, mold mesh, and forming control parameters. Finally, a process guidance document is generated, which guides the physical manufacturing process to perform the molding and manufacturing of the semi-shell product.

[0060] The following section, with reference to the accompanying drawings and specific steps, explains each stage of the molding process optimization design method.

[0061] For the feature extraction and association construction in step S10, the model parsing module converts the 3D geometric data into process feature parameters, specifically including the following sub-steps:

[0062] S11. Design Data Acquisition and Geometric Topology Analysis. The model analysis module acquires the design data of the top and bottom half-shell product. The design data includes the 3D CAD model and the nominal model data. The design data can be in any of the following formats: STEP format, IGES format, or native 3D modeling software format. After reading the design data, the module analyzes the boundary representation data structure of the 3D CAD model to extract the vertex coordinates, edge curve equations, and freeform surface equations that constitute the nominal model, thereby obtaining the geometric topology.

[0063] S12. Surface Discretization and Basic Feature Recognition. Based on the geometric topology, the model parsing module discretizes the continuous surfaces contained in the 3D CAD model into a basic manufacturing feature set. The basic manufacturing features within this set represent specific processing areas. These features include drawing surface features, sidewall cutting edge features, side-cutting features, side-finishing features, and hole location features. Specifically, during feature recognition, the boundary loops, surface patches, and surface normal data of the nominal model are traversed.

[0064] For drawn surface features, the model analysis module marks the surface region that undertakes the main material flow and initial forming as drawn surface features based on the main forming surface of the product, the normal angle with the preset stamping direction vector, the surface continuity, the drawing depth or the boundary of the drawing region.

[0065] For sidewall trimming line features, the model parsing module identifies linear features located in the sidewall area that are used to remove excess material or form product boundaries based on the product sidewall boundary curve, trimming outline, connection relationship between sidewall surface and open boundary, and candidate trimming process type.

[0066] For borehole location features, the model analysis module determines the borehole location features based on the closed internal boundary ring, the perimeter of the boundary ring, the area of ​​the boundary ring, and the center position of the borehole.

[0067] For side-cutting and side-adjusting features, the local surface normal vector is extracted, and the angle between the local surface normal vector and the preset stamping direction vector is calculated. When the angle between the local surface normal vector and the preset stamping direction vector exceeds the side-machining judgment angle threshold, and a side cam mechanism needs to be introduced to perform machining, the corresponding region is marked as a side-machining candidate region. If the side-machining candidate region contains an open boundary cut-off line or a trimming contour line, it is marked as a side-cutting feature; if the side-machining candidate region corresponds to a local shape correction surface or size shaping surface of an already formed sidewall, it is marked as a side-adjusting feature.

[0068] In addition, the threshold for the lateral machining judgment angle is determined by referring to a table based on the material and thickness of the sheet metal and the allowable working angle of the lateral cam mechanism, and the value ranges from 15° to 30°.

[0069] S13. Acquisition of Process Constraint Parameters and Tolerances. Simultaneously, the model analysis module acquires the process constraint parameters and preset process tolerances for the target production equipment used to produce the semi-shell product. The target production equipment is represented by the machine tools configured in the physical stamping workshop, including a 315T hydraulic press, a 200T punch press, and a 160T punch press. The process constraint parameters include the allowable tonnage of the equipment, closing height, slide stroke, worktable size, mold assembly height, lateral mechanism layout space, and blank holder force adjustment range. The preset process tolerances include the dimensional tolerance requirements specified in the product design drawings. In addition to dimensional tolerance requirements, the preset process tolerances also include safety criteria corresponding to forming risk indicators, including the first critical safety factor for local cracking risk indicators and the second critical safety factor for local wrinkling risk indicators. Dimensional tolerance requirements are set according to industry processing specifications, with a value range between ±0.1mm and ±0.5mm. The process constraint parameters and preset process tolerances are used to limit the boundary conditions for the merging and arrangement of subsequent processes.

[0070] S14. Multidimensional Association Table Establishment and Data Transmission. The model parsing module integrates the discretized basic manufacturing feature set and process constraint parameters to establish a multidimensional association table. This table provides a data index for subsequent conflict determination and mesh compensation. The multidimensional association table contains multiple sets of mapping records. These records record the basic manufacturing features within the basic manufacturing feature set, the mesh binding identifiers corresponding to those features, the candidate process types corresponding to those features, and the process constraint parameters. The mesh binding identifier can be any one of the following: feature mapping region, node index rule, or element set identifier. Candidate process types include drawing, trimming, side-cutting, side-finishing, and punching. After the topology generation module completes the actual mesh generation, it binds the basic manufacturing features to the actual CAE mesh nodes or elements based on the mesh binding identifiers.

[0071] By constructing a multidimensional association table, the transformation from a pure geometric model to a digital model incorporating manufacturing process attributes is completed. Subsequently, the model parsing module transmits the multidimensional association table to the topology generation module.

[0072] Reference Figure 3 For the initial process topology generation of step S20 triggered after step S10, the topology generation module performs physical interference determination and processing sequence arrangement based on the data structure issued by the model parsing module, specifically including the following sub-steps:

[0073] S21. Calculate interference evaluation parameters. The topology generation module receives a multidimensional correlation table and, based on the set of basic manufacturing features recorded in the table, performs quantitative calculations of interference evaluation parameters for any two adjacent basic manufacturing features. Interference evaluation parameters include minimum geodesic distance, normal vector angle, tonnage required for merging processes, and mold space requirements. The geodesic distance and normal vector angle evaluation parameters are configured as physical indicators to avoid material flow competition and draft angle conflicts.

[0074] For two adjacent basic manufacturing features, the topology generation module performs the following parallel computation operations: calculates the shortest path length between the boundary points of the two adjacent basic manufacturing features along the product surface mesh as the minimum geodesic distance; extracts the local unit normal vectors at each node of the surface mesh of the two adjacent basic manufacturing features respectively, calculates the spatial angle between the local normal vectors pairwise and extracts the maximum value as the angle between the normal vectors;

[0075] Extract process constraint parameters from the multidimensional association table. Calculate the required tonnage for the merged process based on the stress sequence of candidate processes corresponding to two adjacent basic manufacturing features within the same stamping stroke. When two candidate processes are simultaneously stressed within the same stroke, the corresponding processing forces are superimposed and multiplied by the equipment load margin coefficient to obtain the required tonnage for the merged process. When two candidate processes are stressed at different times within the same mold, the larger of the peak processing forces of each candidate process is taken and multiplied by the equipment load margin coefficient to obtain the required tonnage for the merged process. The equipment load margin coefficient is a load amplification parameter used for equipment tonnage verification, determined based on the allowable off-center load value in the target production equipment manual, the measured or simulated peak load fluctuations of similar processes, and the enterprise's stamping process specifications. When measured load statistics are lacking, a value of 1.05 to 1.15 is used for ordinary synchronous blanking or punching merging, and 1.15 to 1.30 is used for merged processes containing lateral mechanisms or significant off-center loads. The required tonnage for the merged process after amplification by the equipment load margin coefficient must still be less than or equal to the allowable tonnage of the equipment in the multidimensional association table.

[0076] The three-dimensional bounding box algorithm is used to calculate the mold closing height requirement, horizontal envelope dimension requirement, lateral mechanism travel requirement, and spatial interference evaluation value when assembling two molds simultaneously within a single punch press, so as to jointly characterize the mold spatial requirements. Calculating geodesic distance based on a three-dimensional surface mesh and evaluating spatial interference using bounding boxes are well-known techniques in this field and will not be elaborated upon here.

[0077] S22. Constructing the Process Feature Conflict Matrix. Based on various interference evaluation parameters, the topology generation module constructs a two-dimensional square matrix of process feature conflicts. The process feature conflict matrix is ​​used to quantitatively determine whether any two adjacent basic manufacturing features corresponding to corresponding processes have the physical feasibility of being merged for execution. The topology generation module determines the element state of the process feature conflict matrix using the following four independent constraint judgment formulas:

[0078] Distance constraint formula: ;

[0079] Angle constraint formula: ;

[0080] Load constraint formula: ;

[0081] Spatial constraint formula: ,or ,or ,or ;

[0082] in, Represents the minimum geodesic distance; This represents the set distance safety threshold; Represents the angle between the normal vectors; This represents the set tolerance for the included normal angle; This represents the tonnage required for the merging process. This represents the allowable tonnage of equipment in a multidimensional association table. This represents the required closed height of the mold corresponding to the merging process. This represents the allowable closed height of the target machine tool; This represents the horizontal envelope dimension requirement of the merging mold. Represents the workbench size in a multidimensional association table; Represents the travel needs of lateral mechanisms. This represents the allowable space for the lateral mechanism arrangement of the target machine tool; This represents the spatial interference evaluation value calculated from the bounding box. This represents the permissible spatial interference threshold, which is usually set to 0 or a small positive number as the tolerance for safe assembly clearance of the mold. When a small positive number is used, the specific value range is set from 0.5mm to 2.0mm, and its value is determined according to the assembly positioning accuracy specifications of the target machine tool.

[0083] If any one of the distance constraint formula, angle constraint formula, load constraint formula, and spatial constraint formula is true, it is determined that there is physical interference between two adjacent basic manufacturing features, and the state of the corresponding element in the process feature conflict matrix is ​​assigned to 1; if none of the distance constraint formula, angle constraint formula, load constraint formula, and spatial constraint formula is true, it is determined that the two adjacent basic manufacturing features are theoretically allowed to be executed together, and the state of the corresponding element in the process feature conflict matrix is ​​assigned to 0.

[0084] The allowable tonnage of the equipment, the allowable closed height of the target machine tool, the size of the worktable, and the space for the lateral mechanism arrangement are determined according to the target machine tool equipment manual of the physical stamping workshop. The allowable spatial interference threshold is determined according to the mold safety assembly clearance tolerance; when no specific assembly clearance tolerance is obtained, the allowable spatial interference threshold is 0, to indicate that the two mold envelopes must not overlap. Regarding the distance safety threshold and normal angle tolerance, the topology generation module directly determines specific parameter values ​​through preset material mapping criteria: The topology generation module reads the sheet thickness and yield strength of the semi-shell product, normalizes the sheet thickness and yield strength according to the minimum and maximum values ​​in the material database, and interpolates within the range of 30mm to 50mm based on the normalized comprehensive result; the larger the sheet thickness or the higher the yield strength, the closer the distance safety threshold is to 50mm; combining the presence of a negative angle demolding area in the physical mold and the maximum allowable working tilt angle of the lateral cam mechanism, the maximum allowable angle is determined as the normal angle tolerance within the range of 15° to 20°; when there is a negative angle demolding area or the working angle of the lateral cam mechanism is close to the maximum allowable working tilt angle, the normal angle tolerance is taken as a value close to 15°; when there are no such restrictions, the normal angle tolerance is taken as a value close to 20°.

[0085] S23. Read and apply the locking matrix. The topology generation module synchronously reads the preset locking matrix. The two-dimensional square matrix locking matrix is ​​used to introduce a forming risk feedback control mechanism. A value of 0 for an element in the locking matrix corresponding to a combination of feature processes indicates that the element state of the corresponding process feature conflict matrix is ​​determined by interference evaluation parameters; a value of 1 for an element in the locking matrix corresponding to a combination of feature processes indicates that the corresponding feature process combination is determined by the subsequent risk feedback mechanism to be in a prohibited merging state. During the initial layout phase, the state of each element in the locking matrix is ​​initialized to 0. When performing the merging determination, the topology generation module forcibly rewrites the state of the corresponding element in the process feature conflict matrix to 1 for feature combinations where the element state in the locking matrix is ​​1.

[0086] S24. Generate the initial process topology diagram and export the data. After completing the matrix assignment, the topology generation module jointly reads the process feature conflict matrix, locking matrix, and process constraint parameters, and uses a directed graph generation algorithm to output an initial process topology diagram that conforms to the current process constraint parameters. The nodes of the initial process topology diagram represent specific stamping processes, and the directed edges represent the order in which the physical processing is executed. To avoid dependency deadlocks in node arrangement, the directed graph generation algorithm uses topological sorting logic to ensure that all process nodes follow the macroscopic unidirectional constraint of first drawing, then edge trimming and side trimming, then side straightening, and finally punching.

[0087] For feature combinations where the element state in the process feature conflict matrix is ​​0, the topology generation module first generates candidate merging groups and verifies that the elements in the process feature conflict matrix corresponding to any two basic manufacturing features within the candidate merging group are both 0, and the elements in the corresponding logical locking matrix are also both 0. Only when all pairwise constraints within the candidate merging group satisfy the allowed merging conditions is the candidate merging group mapped to the same process node. For feature combinations where the element state in the process feature conflict matrix is ​​1, they are split and mapped to different process nodes, and directed edges are assigned sequential constraints.

[0088] Based on the generated initial process topology diagram, the topology generation module generates a linear multi-process execution sequence. Simultaneously, it reads the basic manufacturing features from the multi-dimensional association table, the feature mapping regions in the mesh binding identifiers, and the corresponding candidate process types. Following the multi-process execution sequence, and based on the nominal product surface, sheet thickness, blanking clearance, die clearance, stamping direction, and the process attributes of each process, it generates initial die surfaces corresponding to different stamping stages. These initial die surfaces are then meshed to form the initial die mesh. Finally, the initial process topology diagram, the multi-process execution sequence, and the initial die mesh are exported and imported into the simulation calculation module, triggering the multi-process forming simulation verification process.

[0089] For the multi-process simulation and deviation extraction in step S30, the simulation calculation module performs physical boundary loading, deformation calculation, and state parameter separation, specifically including the following sub-steps:

[0090] S31. Simulation Boundary Setting. The simulation calculation module receives the initial process topology diagram, multi-process execution sequence, and initial mold mesh from the topology generation module. Using the initial mold mesh as the mold's geometric boundary and based on the logical connections in the initial process topology diagram, it sets the physical boundary conditions for the simulation process. The physical boundary conditions include the material's constitutive model parameters, friction coefficient, and the time-varying blank holder force loading curve. The material's constitutive model parameters are obtained by fitting uniaxial tensile test data of the half-shell product material and are typically defined using an anisotropic yield criterion. The friction coefficient is determined based on the actual lubrication conditions of the stamping workshop and Coulomb's friction theory, with a value ranged between 0.10 and 0.15.

[0091] Based on the multi-process execution sequence, the simulation calculation module calls the internally configured finite element solver to perform multi-process forming simulation. For the explicit dynamic stamping solution and implicit static springback solution in the finite element simulation calculation, those skilled in the art can use a nonlinear solution algorithm based on the updated Lagrange formula for matrix iteration. The mesh large deformation solution mechanism inside the nonlinear solution algorithm is a well-known technology in this field and will not be described in detail here.

[0092] S32. Deviation Calculation and Comparison. After the simulation is completed and the elastic rebound is triggered by unloading the mold constraints, the simulation calculation module extracts the node coordinates of the mesh model in the final state of the semi-shell product. For each extracted mesh node, a spatial mapping algorithm is used to compare the position information of the mesh model node coordinates with the corresponding node coordinates of the nominal model in the design data to measure the degree of local surface deformation. For the execution process of using the spatial mapping algorithm for spatial comparison of node positions, those skilled in the art can use the nearest point search algorithm. The matching mechanism of the nearest point search algorithm is a well-known technology in this field and will not be described in detail here.

[0093] In actual stamping, sheet metal undergoes significant in-plane tangential slip. Directly calculating the spatial Euclidean distance introduces spurious deviations. Therefore, the calculation of nodal normal deviation follows the logic of normal projection, projecting the nodal displacement vector onto the normal of the target surface. The specific calculation relationship is as follows:

[0094] ;

[0095] in, The sign represents the normal deviation of the node, and its positive or negative sign indicates whether the node bulges outward or is concave inward relative to the nominal surface. The coordinates of the mesh model nodes representing the final state of the semi-shell product; This represents the coordinates of the nodes corresponding to the nominal model in the design data; This represents the unit normal vector of the nominal model at the corresponding node. Spatial deviation is calculated using vector dot product to separate the springback deformation perpendicular to the product surface normal and filter out tangential slip data parallel to the surface direction.

[0096] S33. State Parameter Separation and Data Output. Based on the comparison results, the simulation calculation module calculates the node normal deviation and determines whether the absolute value of the node normal deviation of each node exceeds the dimensional tolerance requirement in the preset process tolerance. The dimensional tolerance requirement is set according to the product assembly accuracy specification and the 3D dimension annotation of the 3D CAD model, and the conventional value is set to 0.1mm to 0.5mm. After the determination is completed, the simulation calculation module filters out the set of nodes with out-of-tolerance and simultaneously extracts the mechanical state data during the forming process.

[0097] Specifically, the simulation calculation module reads the strain tensor at the integration point of the mesh element, performs principal value decomposition on the strain tensor, and then uses the shape function extrapolation interpolation algorithm to calculate the mechanical state data at the node position. For the execution process of calculating the mechanical state data using the shape function extrapolation interpolation algorithm, those skilled in the art can call conventional finite element post-processing interpolation functions. The shape function extrapolation interpolation algorithm is a well-known technology in this field and will not be described in detail here.

[0098] The mechanical state data includes the first principal strain, the second principal strain, the thickness reduction rate, out-of-plane displacement data, springback release displacement data generated during the unloading phase of each process in the multi-process execution sequence, and the residual stress distribution after each process. The out-of-plane displacement data reflects the degree of normal bulging of the sheet metal in the unconstrained region without the die; the springback release displacement data characterizes the local elastic recovery of the material during die unloading; and the residual stress distribution after each process characterizes the cumulative springback effect of continuous stamping in multiple processes. Finally, the simulation calculation module outputs a complete data package containing the set of out-of-tolerance nodes, node normal deviations, and mechanical state data, and submits it to the decoupled evaluation module.

[0099] Reference Figure 4 For the topology decoupling assessment and reconstruction in step S40, the decoupling assessment module receives node normal deviation and mechanical state data, performs forming risk determination and topology feedback based on physical interference correlation, specifically including the following sub-steps:

[0100] S41. Spatial Clustering and Reverse Displacement Compensation Field Construction of Out-of-Tolerance Nodes and Forming Risk Nodes. The decoupled evaluation module receives node normal deviation and mechanical state data. Out-of-tolerance nodes within the out-of-tolerance node set, and forming risk nodes whose thickness reduction rate exceeds a preset safety threshold (the preset safety threshold is set according to the material's allowable thinning rate specification, typically between 15% and 25%), are collectively treated as a node set to be clustered. Spatial clustering is performed on this node set. Based on the set node neighborhood search radius and minimum number of nodes, nodes that are spatially continuous and have a cluster density meeting preset requirements are grouped into independent risk regions. The node neighborhood search radius is configured to be 2 to 3 times the average grid side length, and the minimum number of nodes is configured to be 5 to 10. Initially, the node neighborhood search radius is set to 2 times the average grid side length, and the minimum number of nodes is set to 5. When the CAE mesh is locally refined and risk nodes are scattered, the node neighborhood search radius is increased to 3 times the average grid side length, and the minimum number of nodes is increased to 10.

[0101] For the computer processing logic of identifying geometrically continuous nodes using density-based spatial clustering algorithms, those skilled in the art can use the existing DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering algorithm. Its spatial clustering process is a well-known technology in this field and will not be described in detail here.

[0102] After delineating independent risk zones, the decoupling assessment module constructs a reverse displacement compensation field for each node within the risk zone. The calculation relationship of the reverse displacement compensation field is expressed by the following formula:

[0103] ;

[0104] in, This represents the displacement vector in the reverse displacement compensation field. Represents the normal deviation of the node; This represents the unit normal vector of the nominal model at the corresponding node in the design data. To construct a theoretical compensation spatial displacement opposite to the direction of stamping springback, and to provide an input basis for subsequent finite element compensation verification.

[0105] S42. Forming Risk Index Calculation and Compensation Pre-evaluation. Combining mechanical state data, the decoupled evaluation module performs a compensation pre-evaluation, multiplying the displacement vector in the reverse displacement compensation field by a preset step relaxation coefficient, and determining the mapping relationship between the product CAE mesh nodes and the corresponding process mold mesh nodes within the risk area through mesh binding identifiers. Small step compensation amounts are applied to the mold mesh or mold surface of the corresponding process, and small step pre-calculation is performed to obtain the local mesh plastic strain change caused by reverse compensation.

[0106] The step relaxation coefficient is used to prevent mesh distortion caused by applying the full compensation amount at once, and its value is set between 0.1 and 0.3. During the initial compensation pre-evaluation, the step relaxation coefficient is set to 0.2; if mesh distortion warnings or sudden increases in local plastic strain appear in the small step pre-calculation, the step relaxation coefficient is reduced to 0.1; if no mesh distortion appears in two consecutive pre-evaluations and the forming risk index changes steadily, the step relaxation coefficient can be increased but not exceeding 0.3. Based on the extracted local mesh plastic strain changes, principal value decomposition is performed on the estimated changed plastic strain tensor to update the mechanical state data at each node location. The updated first principal strain and the updated second principal strain are extracted, and then the forming risk index, including local cracking risk index and local wrinkling risk index, is calculated. The compensation pre-evaluation does not directly modify the final product mesh, but rather performs small-step compensation on the corresponding process mold surface through the mapping relationship between the product CAE mesh and the mold mesh before performing the pre-calculation.

[0107] The local cracking risk index is constructed based on the forming limit diagram theory. It assesses the likelihood of material thinning and cracking by comparing the actual strain state with the boundary conditions at which the material enters necking fracture. The calculation relationship of the local cracking risk index is expressed by the following formula:

[0108] ;

[0109] in, It represents the risk index of local cracking, characterizing the degree to which local material approaches the extreme value of drawing cracking; This represents the first primary strain after the update; This represents the updated second principal strain; The first principal strain represents the ultimate limit strain at the updated second principal strain on the forming limit diagram. The forming limit diagram can be obtained from material testing, material databases, supplier material cards, or historical simulation models. After obtaining the forming limit diagram, the corresponding ultimate first principal strain is obtained by looking up a table or interpolation based on the updated second principal strain.

[0110] The calculation relationship of the local wrinkling risk index is expressed by the following formula:

[0111] ;

[0112] in, This represents an indicator of the risk of localized wrinkling. The compressive strain represents the strain extracted from the mechanical state data. The value of the compressive strain is taken from the negative part of the updated second principal strain. This represents the set compressive strain instability threshold. This represents the out-of-plane amplitude of the nodal neighborhood mesh extracted from out-of-plane displacement data in mechanical state data; This represents the set tolerance threshold for wrinkle amplitude. Represents the first weighting coefficient; This represents the second weighting coefficient; It represents the ratio of the absolute value of compressive strain to the compressive strain instability threshold, characterizing the degree of in-plane compressive instability; It represents the ratio of out-of-plane amplitude to the wrinkling amplitude tolerance threshold, characterizing the degree of out-of-plane geometric buckling.

[0113] By establishing a dual instability criterion, the tendency of local material to enter the plastic instability limit is quantitatively measured. The compressive strain instability threshold and the wrinkling amplitude tolerance threshold are set according to the standard stamping design manual based on the sheet thickness specifications. The compressive strain instability threshold is defined as the absolute value threshold of compressive strain, which is usually set between 0.05 and 0.10. The wrinkling amplitude tolerance threshold is usually set between 10% and 20% of the sheet thickness. The sum of the first weighting coefficient and the second weighting coefficient is 1, and they are usually evenly distributed at 0.5.

[0114] S43. Forming Safety Risk Assessment and Process Contribution Rate Evaluation. The decoupled assessment module determines whether there is a forming safety risk in the risk area based on the calculated forming risk index. When the maximum local cracking risk index in the corresponding risk area is greater than the set first critical safety factor, or the maximum local wrinkling risk index in the corresponding risk area is greater than the set second critical safety factor, it is determined that there is a forming safety risk in the corresponding risk area.

[0115] The first critical safety factor is used only for determining the exceedance of local cracking risk indicators, and the second critical safety factor is used only for determining the exceedance of local wrinkling risk indicators. The first critical safety factor is determined based on the material forming limit diagram, the allowable material thinning rate, and the qualified trial production records of similar semi-shell products; the second critical safety factor is determined based on the plate thickness, compressive strain instability threshold, wrinkling amplitude tolerance threshold, and wrinkling defect records of similar semi-shell products. For high-strength steel, deep-drawing areas, or areas with frequent historical defects, the first and second critical safety factors are taken as 0.85 to 0.90; for ordinary low-carbon steel or low-risk areas with stable mass production, they are taken as 0.90 to 0.95.

[0116] When there is a forming safety risk, the decoupling assessment module adjusts the forming control parameters according to preset parameter adjustment rules to reduce the probability of local material instability. Specifically, when the local cracking risk index increases or the thickness reduction rate exceeds the threshold, at least one of the following operations is performed according to the preset single adjustment step size: reducing the blank holder force in the corresponding area, reducing the drawbead resistance, increasing the local die radius, and reducing the step relaxation coefficient. When the local wrinkling risk index increases, the blank holder force in the corresponding area is increased or the drawbead resistance is increased according to the preset single adjustment step size to reduce the local material inflow. The single adjustment step size is set to 5% to 10% of the initial value of the corresponding parameter.

[0117] When a forming safety risk exists in a risk area, and the adjusted forming control parameters exceed the adjustment range limited by the process constraint parameters in the multi-dimensional correlation table, or the number of local adjustment iterations of the forming control parameters reaches the set local optimization upper limit, it is determined that the forming safety risk cannot be eliminated by adjusting the forming control parameters. The process contribution rate of each process in the multi-process execution sequence for the corresponding risk area is calculated. The local optimization upper limit is determined based on historical mold adjustment records and simulation convergence records of similar semi-shell products, with a value range of 3 to 5 times; if historical records are lacking, the local optimization upper limit is set to 3 times. The process contribution rate is used to measure the proportion of the impact of a single process on the cumulative material damage in the corresponding risk area. The decoupling evaluation module first calculates the normalized damage increment of each process on the corresponding risk area, and then calculates the process contribution rate. The calculation relationship is as follows:

[0118] ;

[0119] ;

[0120] in, Representing the The first process is related to the second... The normalized damage increment of each risk area represents the comprehensive physical risk level caused by a single process. To avoid the default positive stable term with a denominator of zero, its value range is set to 10. -6 Up to 10-4 The specific value is set according to the floating-point operation precision of the computing system; Representing the The first principal strain increment generated in each process stage; Representing the The increase in thickness reduction rate generated in each process stage; Representing the The rebound release displacement increment generated during the unloading stage of each process; , and These are the normalized reference values ​​for the first principal strain increment, the thickness reduction rate increment, and the springback release displacement increment, respectively. The first sensitivity weighting factor represents the first principal strain increment; The second sensitivity weighting factor represents the increase in thickness reduction rate. The third sensitivity weighting factor represents the rebound release displacement increment;

[0121] Representing the The first process is related to the second... The process contribution rate of each risk area represents the dominant proportion of the current merged process in the global cumulative damage. Represents the total number of processes in a multi-process execution sequence; This represents the sum of the normalized damage increments of all processes for the corresponding risk area.

[0122] The first, second, and third sensitivity weighting factors are set using the analytic hierarchy process (AHP) to assess the influence of each physical quantity on the forming limit, ensuring uniformity across multiple physical quantities and satisfying the constraint that the sum of the first, second, and third sensitivity weighting factors is 1. In practical applications, the first sensitivity weighting factor typically ranges from 0.4 to 0.6, the second from 0.2 to 0.3, and the third from 0.1 to 0.3. Since the dimensions of the first principal strain increment, thickness reduction rate increment, and springback displacement increment are different, a normalized reference value is used for dimensionless processing. The normalized reference value can be determined by the allowable strain, allowable thinning rate, and dimensional tolerance threshold of the material.

[0123] S44. Decoupling Judgment and Topology Reconstruction Flow. The decoupling evaluation module determines whether the decoupling judgment conditions are met based on the calculated contribution rates of each process. The decoupling judgment condition is configured such that the contribution rate of a certain merged process exceeds a preset single-process dominance judgment threshold, which is set to a range of 0.65 to 0.80. The single-process dominance judgment threshold is calibrated based on defect attribution records of similar products and records of changes in risk indicators before and after simulation decoupling; if historical records are lacking, the single-process dominance judgment threshold is set to 0.70. When the decoupling judgment conditions are met, it indicates that there is a dominant correlation between the current risk area and the corresponding merged process combination, and the forming safety risk is difficult to eliminate through adjustments to conventional forming control parameters.

[0124] If the contribution rate of all processes is lower than the threshold for determining the dominant process, then the forced decoupling process will not be executed. Instead, the risk area, the contribution rate of each process, the adjustment record of the formed control parameters that have been executed, and the current process topology diagram will be output as the abnormal assessment result. The automatic forced decoupling process of the risk area will be terminated for subsequent manual review.

[0125] A process node is considered a merged process node only if its process contribution rate exceeds the threshold for determining the dominance of a single process. The merged process node corresponds to a combination of basic manufacturing features consisting of at least two basic manufacturing features. In this case, the decoupling evaluation module locates the corresponding combination of basic manufacturing features through a multidimensional association table, overwrites the element state of the corresponding process feature conflict matrix with 1, and assigns the element state of the corresponding feature combination in the locking matrix to 1.

[0126] After performing matrix overwriting, the compensation pre-evaluation process for the current round is terminated, and the topology generation module is controlled to decouple and reconstruct the process topology diagram based on the updated underlying matrix. The decoupled process topology diagram is then fed back to the simulation calculation module to re-execute the multi-process forming simulation. If the evaluation results show that all forming risk indicators are within limits, or if adjusting the forming control parameters successfully brings the risk area indicators back to within the safe threshold, i.e., the decoupling operation is not triggered, the parameter control module is directly activated to proceed to the next step.

[0127] Reference Figure 5 For the parameter iteration and mesh compensation that triggers step S50 after step S40, the parameter control module performs multi-process compensation allocation for the reverse displacement compensation field that has passed the forming risk verification, specifically including the following sub-steps:

[0128] S51. Calculation of Springback Compensation Weights for Multiple Processes. The parameter control module calculates the springback compensation weight for each process in the multi-process execution sequence based on the node normal deviation. The influence of each process on the final springback amount is positively correlated with the elastic strain energy released during the unloading of the corresponding process at the end of forming. Therefore, the springback compensation weight is determined based on the strain energy release ratio during the unloading stage of each process. Specifically, for each process in the multi-process execution sequence, the parameter control module extracts the node... The local elastic strain energy released by the risk area or its neighboring units during the unloading phase of each process is used as the ratio of the local elastic strain energy released during the unloading phase of the corresponding process to the sum of the local elastic strain energy released during the unloading phases of all processes. This ratio is taken as the value of the local elastic strain energy released during the unloading phase of the corresponding process for the node. rebound compensation weight .

[0129] The execution logic for extracting elastic strain energy can be obtained by those skilled in the art through volume integral calculation of the stress and elastic strain tensor of each element in the finite element model of the corresponding process. The execution logic for extracting elastic strain energy is a well-known technology in this field and will not be described in detail here.

[0130] S52. Mold Mesh Node Coordinate Compensation Mapping Update. The parameter control module distributes the displacement vectors in the reverse displacement compensation field generated by the decoupled evaluation module to the mold mesh of the corresponding process according to the calculated springback compensation weight, so as to update the node coordinates of the mold mesh.

[0131] To avoid geometric distortion of the mold surface caused by compensation, the node coordinates are updated by projection along the local normal of the mold. For nodes in the CAE mesh... The parameter control module determines the first [parameter] based on the mesh binding identifier, nearest point projection relationship, or parameter coordinate mapping relationship. The corresponding node in the mold grid for each process The calculation relationship is expressed by the following formula:

[0132] ;

[0133] in, Represents the updated number The corresponding node in the mold mesh of each process Spatial coordinates; Represents the current iteration step. The corresponding node in the mold mesh of each process Spatial coordinates; Representative product: CAE mesh node In the The corresponding node index in the mold mesh for each process; For the first The compensation direction coefficient of the mold surface for each process; This represents the global compensation relaxation coefficient, and its initial value is usually set between 0.5 and 0.8 to control the amount of mold surface variation in a single iteration. Representing the The first process is related to the second... The rebound compensation weight of each node; This represents the displacement vector in the reverse displacement compensation field. Represents the current iteration step. Each process mold grid is at the corresponding node The local unit normal vector at that location, i.e., the local normal of the mold; This represents the projected displacement component of the displacement vector in the reverse displacement compensation field onto the local normal direction of the mold.

[0134] Compensation Direction Coefficient The value is either 1 or -1, used to unify the definition of the normal vector on different mold working surfaces. When the local unit normal vector of the mold points to the contact side of the sheet metal, The value is 1. When the local unit normal vector of the mold deviates from the contact side of the sheet metal, The value is set to -1 to ensure that the compensation displacement of the mold surface in each process always moves in the opposite direction to the physical springback.

[0135] Global compensation relaxation coefficient The value is determined based on the current maximum node normal deviation and the allowable single surface change of the mold. When the current maximum node normal deviation is large or the local curvature change of the mold is large, the value is taken as 0.5 to 0.6. When the current maximum node normal deviation is small and the mold surface has good continuity, the value is taken as 0.7 to 0.8. When historical iteration data is lacking, the initial value is taken as 0.6.

[0136] S53. Synchronous Adjustment and File Rewriting of Forming Control Parameters. After the node coordinates are updated, the parameter control module extracts the spatial gradient distribution of forming risk indicators and corrects the process parameters: for areas where the local cracking risk indicator reaches 90% to 95% of the first critical safety factor, the parameter control module reduces the blank holder force or drawbead resistance in the corresponding area; for areas where the local wrinkling risk indicator reaches 90% to 95% of the second critical safety factor, the parameter control module increases the blank holder force or drawbead resistance in the corresponding area to limit local material inflow.

[0137] Meanwhile, the global compensation relaxation coefficient is adaptively scaled based on the ratio of the current maximum node normal deviation to the maximum node normal deviation of the previous iteration step. The specific adjustment logic is as follows: if the current iteration is in its first iteration step, the maximum node normal deviation of the previous iteration step is recorded as the maximum node normal deviation calculated in the initial simulation; when the ratio of the current maximum node normal deviation to the maximum node normal deviation of the previous iteration step is greater than or equal to 0.8, it indicates that the deviation convergence speed has slowed down, and the global compensation relaxation coefficient is multiplied by a reduction factor of 0.5; when the ratio of the current maximum node normal deviation to the maximum node normal deviation of the previous iteration step is less than 0.8, the global compensation relaxation coefficient remains unchanged. The convergence judgment ratio of 0.8 and the reduction factor of 0.5 are both set based on historical convergence test experience of stamping forming simulation. Subsequently, the parameter control module rewrites the updated mold mesh and forming control parameters into the solver input file.

[0138] S54. Compensation Iteration Drive and Convergence Judgment. The parameter control module sends out the solver input file, driving the simulation calculation module to perform the next round of iterative calculation based on the new mold surface. This iterative cycle continues until the set of out-of-tolerance nodes output by the simulation calculation module is empty, and the local cracking risk index among the various forming risk indicators does not exceed the set first critical safety factor, and the local wrinkling risk index does not exceed the set second critical safety factor. When the set convergence conditions are met, the parameter control module terminates the iteration and triggers the file output module; when the set maximum iteration threshold is reached but the set of out-of-tolerance nodes is still not empty, or any forming risk indicator still exceeds the corresponding set critical safety factor, the parameter control module terminates the current automatic iteration process, but does not trigger the file output module to enter the physical manufacturing file generation process, but instead outputs the non-converged risk area, the current process topology diagram, parameter adjustment records, and suggested manual review information. The maximum iteration threshold is set to 15 to 20 times to prevent infinite loops caused by non-physical convergence.

[0139] For the process document output and physical manufacturing system mapping of step S60 triggered after the iteration convergence of step S50, the document output module performs mold surface reconstruction, process parameter solidification, and manufacturing system data interaction, specifically including the following sub-steps:

[0140] S61. Data Locking and Surface Reconstruction. The file output module obtains the iterative convergence results and performs locking and storage of the iterative convergence data, namely locking the final process topology diagram, mold mesh, and forming control parameters. The locked mold mesh specifically covers the three-dimensional discrete data of the entire set of multi-process mold surfaces, including the optimized compensation surface data of each process mold, namely the compensation surface data of the drawing die, trimming die, side cutting die, side finishing die, and punching die. For the locked mold mesh, since stamping forming simulation relies on discrete meshes, while physical CNC machining relies on continuous geometric surfaces, the file output module performs reverse engineering on the mold mesh, reconstructing the discrete node coordinates into a continuous three-dimensional CAD surface model.

[0141] In the process of reconstructing a continuous 3D CAD surface model, the file output module divides the feature region based on the curvature change trend of the mesh nodes, and generates non-uniform rational B-spline surface patches within the feature region. Adjacent surface patches are smoothly transitioned at the boundary through tangent continuity conditions to reduce the risk of machining marks caused by sudden changes in local curvature. The surface approximation tolerance during the fitting process is usually controlled within the range of 0.01mm to 0.05mm to ensure reconstruction accuracy. The specific tolerance value is set according to the surface quality grade requirements of the final stamped part.

[0142] For the execution logic that uses the non-uniform rational B-spline algorithm to fit a discrete mesh into a smooth continuous surface, those skilled in the art can call conventional inverse modeling surface fitting operators. Surface fitting and continuous smoothing techniques are well-known in the field and will not be elaborated here. The reconstructed 3D CAD surface model is saved in a standard 3D data exchange format, namely STEP or IGES format.

[0143] S62. Parameter Solidification and Process File Compilation. The file output module extracts the final solidified forming control parameters after multiple iterations, as well as auxiliary process data generated during simulation and analysis. The forming control parameters include the final blank holder force value, drawbead resistance parameters, local die fillet radius, step relaxation coefficient, and global compensation relaxation coefficient. The auxiliary process data includes the actual forming tonnage consumed at each process node, the cross-sectional geometry of the drawbead calculated based on the drawbead resistance parameters using preset drawbead parameter conversion rules, and basic configuration parameters derived from a multi-dimensional correlation table. The basic configuration parameters include blanking clearance, side cutting angle, and die closing height.

[0144] The file output module integrates the final solidified forming control parameters, auxiliary process data, decoupled process topology diagram, and equipment information from the multidimensional association table, and compiles them into a process guidance document. In specific application scenarios, the process guidance document is packaged and distributed using an extensible markup language format to facilitate parsing and extraction by the manufacturing execution system. The process guidance document details the process sequencing and flow rules, machine tool selection mapping relationships, and parameter setting boundaries for on-site debugging when performing semi-shell product processing in the physical stamping workshop. The process sequencing rules include sheet preparation, drawing, edge trimming, side cutting, side finishing, punching, welding, and product inspection.

[0145] S63. CNC Code Generation and Equipment Mapping. The file output module imports the reconstructed 3D CAD surface model into the computer-aided manufacturing software. Based on the mold material properties and preset machining specifications, it generates roughing and finishing tool path data. Specifically, it sets different machining allowance step distributions, such as reserving 0.5mm to 1.0mm allowance in the roughing stage and performing body-fit cutting in the finishing stage. The specific machining allowance values ​​are determined based on the tool radius and machine tool cutting rigidity parameters. For the post-processing operation of generating CNC machining tool paths based on 3D continuous surfaces, those skilled in the art can set the cutting depth, feed rate, and spindle speed parameters based on the machine tool kinematic model. CNC programming and post-processing technologies are well-known in the field and will not be elaborated here.

[0146] After generating roughing and finishing toolpath data, the toolpath data is converted into CNC machining code files suitable for the target CNC machine tool controller. Finally, the file output module distributes the CNC machining code files and process guidance documents to the control terminals of the physical mold CNC machining workshop and the stamping production line through the manufacturing execution system network interface. This allows the process guidance documents to guide or assist operators, the manufacturing execution system, or the equipment control system in setting the corresponding machining parameters for the target production equipment.

[0147] Subsequently, the physical manufacturing process sequentially executes product blanking, continuous stamping, and welding robot assembly. Specifically, this is reflected in determining the target production equipment corresponding to each process based on the final process topology diagram and equipment constraint parameters. In one embodiment, the drawing process is guided to be executed by a 315T hydraulic press, the trimming and side-cutting processes are both executed by a 200T punch press, the side-finishing process is executed by any one of the following: a 315T hydraulic press or a stamping equipment that meets the tonnage and closing height requirements, and the punching process is executed by a 160T punch press.

[0148] Finally, assembly is completed using welding fixtures and welding robots, thus completing the data transfer and process mapping from process topology design and mold compensation data generation to physical manufacturing document output.

[0149] Reference Figures 6 to 8The following example, using a high-strength steel lower shell product for a new energy vehicle power battery, illustrates the execution process of the aforementioned molding process optimization design system in a specific product. This embodiment is used to illustrate the implementation of the present invention and does not constitute a limitation on product type, material type, or equipment model.

[0150] The model parsing module acquires the STEP format 3D CAD design drawings and nominal model data of the target lower shell product. After the design data is input, the boundary representation data structure is parsed, vertex coordinates and edge curve equations are extracted, and the continuous surface is discretized into a basic manufacturing feature set. The basic manufacturing feature set corresponds to a set of 3D mesh regions defined by a node index array or a cell set identifier in the data structure. The mesh binding identifier specifically adopts node index rules and cell set identifiers.

[0151] The extracted basic manufacturing feature set covers drawing surface features, side-finishing features, sidewall trimming line features, and three hole position features. Simultaneously, production workshop configuration parameters are read to obtain process constraint parameters, including the allowable tonnage of the target machine tool (315T hydraulic press) at 3000kN, a closed height of 800mm, and a preset dimensional tolerance requirement of ±0.2mm. A multidimensional association table is established, forming a mapping index dictionary that integrates the basic manufacturing feature set, mesh binding identifiers, candidate process types, and process constraint parameters.

[0152] The topology generation module calculates interference evaluation parameters for adjacent foundation manufacturing features based on a multidimensional association table. For sidewall tangent features and hole location features, geodesic distances are calculated along the product surface grid, yielding a minimum geodesic distance of 12.5 mm. Based on a sheet thickness of 1.2 mm, a distance safety threshold of 35 mm is calculated using interpolation. This is then applied according to the distance constraint formula. (in Represents the actual calculated minimum geodetic distance. (Representing the set distance safety threshold), if spatial interference competition occurs, the element state of the corresponding sidewall cutting edge feature and hole position feature in the process feature conflict matrix is ​​assigned a value of 1.

[0153] Since all interference evaluation parameters for the drawing surface feature and the side-forming feature did not exceed the limits, their corresponding element states were assigned a value of 0. The topology generation module uses a directed graph generation algorithm to output an initial process topology diagram that conforms to the current equipment parameter constraints, generating a multi-process execution sequence that sequentially arranges the drawing and side-forming composite process, the trimming process, and the punching process. Based on the multi-process execution sequence, an initial mold surface is generated, and the initial mold surface is meshed to form an initial mold mesh. The initial process topology diagram, the multi-process execution sequence, and the initial mold mesh are then imported into the simulation calculation module.

[0154] The simulation module uses the initial mold mesh as the geometric boundary, loads the anisotropic constitutive model parameters of high-strength steel, sets the friction coefficient to 0.12, and executes multi-process forming simulation. It extracts the node coordinates of the mesh model in the final state of the semi-shell product. A spatial mapping algorithm is used to compare the positional information of the mesh model node coordinates with the corresponding node coordinates of the nominal model in the design data. For the execution logic of the geometric point set matching operation using the spatial mapping algorithm, those skilled in the art can call the nearest point iteration function in existing 3D analysis libraries; its point cloud matching and projection calculation are well-known technologies in the field and will not be elaborated here.

[0155] The nodal normal deviation of each node was calculated, and a set of out-of-tolerance nodes with an absolute value of nodal normal deviation greater than 0.2 mm was selected. The integral point strain was extracted and interpolated using shape functions to obtain mechanical state data. The mechanical state data includes the first principal strain, the second principal strain, the thickness reduction rate, out-of-plane displacement data, the springback release displacement data generated during the unloading stage of each process in the multi-process execution sequence, and the residual stress distribution after each process.

[0156] The decoupling evaluation module considers the out-of-tolerance nodes within the out-of-tolerance node set, along with forming risk nodes with a thickness reduction rate exceeding 20%, as the node set to be clustered. The DBSCAN clustering algorithm is run, with a node neighborhood search radius of 5mm and a minimum number of nodes of 6, to aggregate geometrically connected discrete nodes. In this analysis, the system successfully isolated two independent risk regions: Region 1 (flange corner) and Region 2 (sidewall). A reverse displacement compensation field is constructed for each node in the risk region. The displacement vector in the reverse displacement compensation field is calculated using the formula... Theoretical displacement construction is performed (wherein) This represents the displacement vector in the reverse displacement compensation field. Represents the normal deviation of the node. (Represents the unit normal vector of the nominal model at the corresponding node in the design data). Compensation pre-calculation was performed with a relaxation coefficient of 0.2 to update the mechanical state data at each node and calculate the wrinkling index. The evaluation results showed that the maximum wrinkling index in region 1 (flange corner) reached 1.12. After attempting to increase the local blank holder force, the wrinkling index did not decrease, indicating that the forming safety risk could not be eliminated by adjusting the forming control parameters.

[0157] The absolute damage increment of each process in the multi-process execution sequence for the corresponding risk area is extracted, and the process contribution rate is calculated. It is found that the contribution rate of the drawing and side-forming composite process is as high as 0.82, which meets the decoupling judgment condition (i.e., the node is a merged process node and corresponds to the two basic manufacturing features of drawing surface and side-forming). The decoupling evaluation module locates the corresponding basic manufacturing feature combination through the multi-dimensional association table, and overwrites the element state of the corresponding process feature conflict matrix with 1. At the same time, it assigns the element state of the corresponding feature combination in the lock matrix to 1, terminates the compensation pre-evaluation, and controls the topology generation module to split the original drawing and side-forming composite process into two independent process nodes, the drawing process and the side-forming process, decouples and reconstructs the process topology diagram and feeds it back to the simulation calculation module.

[0158] The parameter control module receives the reverse displacement compensation field that has passed the forming risk verification. For the drawing, side-forming, and trimming processes, stress and elastic strain tensors are extracted using finite element analysis. Local elastic strain energy is obtained by calculating the volume integral of the element, determining the springback compensation weight for each process on the flange corner node. Based on the formula... Update the coordinates of the mold mesh nodes for the corresponding process (where Represents the updated number The first process mold grid The spatial coordinates of each node; Represents the current iteration step. The first process mold grid The spatial coordinates of each node; For the first The compensation direction coefficient of the mold surface for each process; Represents the global compensation relaxation coefficient. Representing the The first process is related to the second... The rebound compensation weight of each node; This represents the displacement vector in the reverse displacement compensation field. Represents the current iteration step. The mold grid for the first process is in the... (Local unit normal vector at each node).

[0159] Global compensation relaxation coefficient The initial value is assigned to 0.7. The mold mesh with updated node coordinates is rewritten as the solver input file to drive the next iteration. As the iteration progresses, by the 8th iteration, the set of out-of-tolerance nodes is empty, and all forming risk indicators are within the preset process tolerance, triggering the file output module.

[0160] The file output module locks the finally converged mold mesh, divides feature regions based on the curvature variation trend of the mesh nodes, and fits and generates a non-uniform rational B-spline surface patch. The surface approximation tolerance is controlled within 0.02mm, and a 3D CAD continuous surface model in STEP format is exported. The final solidified blank holder force value, drawbead resistance parameters, and blanking clearance are extracted, machine tool equipment information is integrated, and a process guidance document in Extensible Markup Language format is compiled. Based on the mold material properties, a machining allowance step distribution is set, CNC machining tool trajectory data is generated, and converted into CNC machining code files. The data is sent to the physical mold CNC machining workshop through the manufacturing execution system network interface. The manufacturing system maps the drawing process to a 315T hydraulic press for execution, and the trimming and punching processes are assigned to a 200T punch press, guiding the physical manufacturing system to sequentially complete the production of the half-shell product.

[0161] This invention effectively overcomes the physical bottleneck problem caused by blind trial and error in traditional stamping die compensation. Based on the execution process and figures of this embodiment, the following technical conclusions can be drawn:

[0162] First, refer to Figure 7 The distribution of process contribution rates in the risk areas shown indicates that the system quantitatively calculates the dominant risk sources corresponding to different risk areas. In area 1, i.e., the flange corner, the process contribution rate of the drawing and side-forming composite process is 0.82, exceeding the decoupling threshold of 0.75, indicating that the local wrinkling risk in this area is mainly related to the combined processing of the drawing surface features and the side-forming features. In area 2, i.e., the sidewall area, the process contribution rates of both the drawing and side-forming composite process and the edge-cutting process do not exceed the decoupling threshold. Therefore, the system does not perform automatic topology decoupling in this area, but instead outputs the corresponding risk area and process contribution rate results for subsequent parameter adjustment or manual review.

[0163] Secondly, after the system triggers topology decoupling and executes mold compensation iteration, refer to Figure 6 The compensation iteration convergence curve shown indicates that the maximum nodal normal deviation decreased from the initial 1.65 mm to 0.08 mm at the 8th iteration, which is below the 0.2 mm dimensional tolerance threshold; the number of out-of-tolerance nodes decreased from the initial 450 to 0. These results demonstrate that updating the mold mesh through the compensation mapping relationship using the reverse displacement compensation field enables the dimensional deviations in this embodiment to gradually converge.

[0164] Finally, further reference Figure 8The provided forming limit diagram state comparison, with the first principal strain as the vertical axis and the second principal strain as the horizontal axis, shows that before optimization, a large number of state points representing dangerous areas were distributed above the forming limit curve, exceeding the cracking zone, and crossing the wrinkling boundary to the upper left, entering an extremely unsafe wrinkling zone. After topology reconstruction and parameter iteration, the optimized state points, representing the final stress state of the material, shifted significantly downwards and converged towards the center, falling entirely back into the safe zone. These results demonstrate that the system can correct the process topology based on forming risk feedback and reduce the risks of local cracking and local wrinkling through mold compensation and iterative forming control parameters.

[0165] While specific embodiments of the present invention have been described above, those skilled in the art should understand that these specific embodiments are merely illustrative. Those skilled in the art can omit, substitute, and modify the details of the above methods and systems in various ways without departing from the principles and essence of the present invention. For example, combining the above method steps to perform substantially the same function and achieve substantially the same result using substantially the same method falls within the scope of the present invention.

Claims

1. A method for optimizing the molding process of a semi-shell product with upper and lower snap-fit ​​joints, characterized in that, Includes the following steps: The design data of the top and bottom snap-fit ​​half-shell product is obtained and discretized into a basic manufacturing feature set, and a multi-dimensional association table is established in combination with the preset process constraint parameters; Based on the basic manufacturing feature set, a process feature conflict matrix is ​​constructed, an initial process topology diagram and a multi-process execution sequence are output, and an initial mold mesh is generated by combining the multi-dimensional association table. Based on the initial process topology diagram, the multi-process execution sequence, and the initial mold mesh, perform multi-process forming simulation to obtain the set of out-of-tolerance nodes and mechanical state data; Based on the set of out-of-tolerance nodes and the mechanical state data, a reverse displacement compensation field is constructed and a compensation pre-evaluation is performed. When it is determined that there is a forming safety risk, the process feature conflict matrix is ​​reversed to decouple and reconstruct the process topology and the multi-process forming simulation is re-executed. When it is determined that there is no forming safety risk, multi-process compensation allocation is performed based on the reverse displacement compensation field, and iterative calculation is driven until the preset process tolerance is met to obtain the iterative convergence result; Lock the iterative convergence result and generate a process guidance document.

2. The method for optimizing the molding process of the upper and lower snap-fit ​​semi-shell product according to claim 1, characterized in that, The design data of the top and bottom snap-fit ​​half-shell product is obtained and discretized into the basic manufacturing feature set. A multidimensional association table is then established based on the preset process constraint parameters, including: Based on the geometric topology of the 3D CAD model in the design data and the preset stamping direction vector, the features of the drawing surface, sidewall cutting edge, hole position, side cutting, and side finishing are identified and extracted, and combined to construct the basic manufacturing feature set. Obtain the process constraint parameters of the target production equipment and the preset process tolerance, and associate and bind the basic manufacturing feature set with the grid binding identifier, candidate process type and the process constraint parameters, and write them together into the multidimensional association table covering the basic manufacturing feature set.

3. The method for optimizing the molding process of the upper and lower snap-fit ​​semi-shell product according to claim 1, characterized in that, Constructing the process feature conflict matrix based on the basic manufacturing feature set includes: For any two adjacent basic manufacturing features, calculate the minimum geodesic distance, the angle between normal vectors, the tonnage required for the merged process, and the mold space requirement to obtain interference evaluation parameters. The following four independent constraint judgment formulas are executed by calling the interference evaluation parameters: If the minimum geodesic distance is less than the set distance safety threshold, then the distance constraint formula holds true; If the included angle of the normal vectors is greater than the set tolerance for the included angle of the normal vectors, then the angle constraint formula holds. If the tonnage required for the merging process is greater than the set allowable tonnage of the equipment, then the load constraint formula holds true; If the mold space requirement exceeds the set target machine tool's allowable closed height, worktable size, lateral mechanism arrangement space, or allowable space interference threshold, then the space constraint formula holds true. If any of the four independent constraint judgment formulas is true, it is determined that there is physical interference between two adjacent basic manufacturing features, and the corresponding element in the process feature conflict matrix is ​​assigned a state value of 1. If none of the four independent constraint judgment formulas are true, it is determined that two adjacent basic manufacturing features are theoretically allowed to be executed together, and the corresponding element state in the process feature conflict matrix is ​​assigned a value of 0. The process feature conflict matrix is ​​constructed by traversing all basic manufacturing feature combinations composed of any two adjacent basic manufacturing features.

4. The method for optimizing the molding process of the upper and lower snap-fit ​​semi-shell product according to claim 1, characterized in that, Output the initial process topology diagram and the multi-process execution sequence, and generate the initial mold mesh by combining the multi-dimensional association table, including: The system reads the preset locking matrix, the process feature conflict matrix, and the process constraint parameters together, and outputs the initial process topology diagram and the multi-process execution sequence. Based on the multidimensional association table, an initial mold surface is generated according to the multi-process execution sequence and the design data, and the initial mold surface is meshed to form the initial mold mesh.

5. The method for optimizing the molding process of the upper and lower snap-fit ​​semi-shell product according to claim 1, characterized in that, Obtaining the set of out-of-tolerance nodes includes: Using the initial mold mesh as the mold geometric boundary and based on the logical association of the initial process topology diagram, the physical boundary conditions of the simulation process are set; Based on the physical boundary conditions, the internally configured finite element solver is invoked to start the multi-process forming simulation according to the multi-process execution sequence until the simulation ends, and the mesh model of the final state of the upper and lower half-shell product is obtained. Extract the nominal model based on the design data; After comparing the node coordinates of the mesh model of the final state of the upper and lower half-shell product with the node coordinates of the corresponding nominal model, the node normal deviation is calculated. Determine whether the absolute value of the node normal deviation of each node is greater than the dimensional tolerance requirement in the preset process tolerance, filter out the nodes that are greater than the dimensional tolerance requirement, and form the set of out-of-tolerance nodes by the filtered nodes.

6. The method for optimizing the molding process of the upper and lower snap-fit ​​semi-shell product according to claim 5, characterized in that, Constructing the reverse displacement compensation field based on the set of out-of-tolerance nodes and the mechanical state data, and performing the compensation pre-evaluation, includes: Based on the set of out-of-tolerance nodes and the mechanical state data, spatial clustering is performed to delineate the risk area, and the displacement vector in the reverse displacement compensation field is obtained according to the node normal deviation mapping. The displacement vector, combined with a preset step relaxation coefficient, is applied to the mold mesh or mold surface of the corresponding process for small step pre-calculation. Based on the change in local mesh plastic strain caused by the small step pre-calculation, the forming risk index, including local cracking risk index and local wrinkling risk index, is calculated to complete the execution of the compensation pre-evaluation. When the local cracking risk index or the local wrinkling risk index indicates that the corresponding risk area has the forming safety risk, and the forming safety risk cannot be eliminated by adjusting the set forming control parameters, the process contribution rate is calculated to drive the reverse correction of the process feature conflict matrix.

7. The method for optimizing the molding process of the upper and lower snap-fit ​​semi-shell product according to claim 6, characterized in that, Triggering the calculation of the process contribution rate includes: For each process in the multi-process execution sequence, the first principal strain increment, thickness reduction rate increment, and springback release displacement increment generated in each process stage are directly extracted. The normalized damage increment is obtained by performing weighted processing calculations using the first sensitivity weighting factor corresponding to the first principal strain increment, the second sensitivity weighting factor corresponding to the thickness reduction rate increment, and the third sensitivity weighting factor corresponding to the rebound release displacement increment. Based on the normalized damage increment corresponding to each of the aforementioned processes, the process contribution rate of each of the aforementioned processes to the risk area is calculated and determined.

8. The method for optimizing the molding process of the upper and lower snap-fit ​​semi-shell product according to claim 7, characterized in that, The reverse modification of the process feature conflict matrix to decouple and reconstruct the process topology includes: When the contribution rate of the process exceeds the preset threshold for single process dominance, and the corresponding process node is a merged process node, and the merged process node corresponds to at least two of the basic manufacturing features, it is determined that the decoupling determination condition is met. When the decoupling determination condition is met, the basic manufacturing feature combination corresponding to the merged process node is located, the corresponding element state in the process feature conflict matrix is ​​overwritten to 1, and the corresponding element state in the preset locking matrix is ​​simultaneously assigned to 1, thereby controlling the decoupling and reconstructing of the process topology diagram.

9. The method for optimizing the molding process of the upper and lower snap-fit ​​semi-shell product according to claim 1, characterized in that, Performing the multi-process compensation allocation based on the reverse displacement compensation field includes: Extract the local elastic strain energy released during the unloading stage of each process, and determine the springback compensation weight of the corresponding process based on the ratio of the local elastic strain energy released during the unloading stage of the corresponding process to the sum of the local elastic strain energy released during the unloading stages of all processes. Combining the springback compensation weight with the preset global compensation relaxation coefficient, projection is performed along the local normal of the mold to update the spatial coordinates of the corresponding node in the mold mesh of the corresponding process.

10. The method for optimizing the molding process of the upper and lower snap-fit ​​semi-shell product according to claim 9, characterized in that, Obtaining the iterative convergence result, locking the iterative convergence result, and generating the process guidance document include: Based on the ratio of the current maximum node normal deviation to the maximum node normal deviation of the previous iteration step, the global compensation relaxation coefficient in the set forming control parameters is adaptively scaled and adjusted. The updated mold mesh and the forming control parameters are rewritten as the solver input file, and the multi-process forming simulation is re-executed for the next round of iterative calculation until the set of out-of-tolerance nodes is empty and the local cracking risk index in the forming risk index does not exceed the set first critical safety factor and the local wrinkling risk index does not exceed the set second critical safety factor. The iterative convergence result is then obtained. The final process topology, mold mesh, and forming control parameters corresponding to the iterative convergence result are locked, and the generated auxiliary process data and equipment information are integrated and compiled to generate the process guidance document.

Images