Method for optimizing die structure based on powder forming simulation three-dimensional modeling

Abstract

The application belongs to the technical field of mold optimization design, and relates to a mold structure optimization method based on powder forming simulation three-dimensional modeling. Multiple candidate mold configurations are introduced into a powder forming simulation environment, a pressing process numerical simulation is performed, and the candidate configurations are preliminarily evaluated and screened; the total number of simulation grid units is determined, a unit coordinate reference system is established, and an optimization reference unit is selected; step-by-step approximation adjustment is performed on other units and is summarized to a global integrated module; the coordinates at the common boundaries of adjacent regions are weighted and cooperatively fused, and the coordinate reference values of all grid units are updated; the simulation model is reconstructed and the pressing simulation is performed again, comprehensive sorting is performed on all candidate configurations, and the optimal design scheme is output. The application effectively unifies the multi-region grid coordinate reference, overcomes the defects of grid fixation and boundary mismatch in traditional simulation, significantly improves the adaptability and reliability of the pressing process simulation, and can provide a systematic and high-precision optimization design approach for powder metallurgy molds.

Description

Technical Field

[0001] This invention belongs to the field of mold optimization design technology, and relates to a mold structure optimization method based on three-dimensional modeling simulation of powder forming. Background Technology

[0002] The structural design of powder metallurgy molds directly affects the density uniformity, surface accuracy, and mechanical properties of the pressed parts. Traditional mold design relies heavily on engineering experience and trial-and-error experiments, adjusting punch radii, demolding tapers, and cavity dimensions through repeated mold modifications and pressing tests. This process is time-consuming, costly, and makes it difficult to obtain the globally optimal configuration. With the development of numerical simulation technology, the finite element method has been introduced into the analysis of the powder pressing process, capable of simulating the stress-strain field and density distribution during powder densification. However, existing simulation-aided design typically performs unidirectional calculations only within a single fixed mesh system. The mesh coarseness, density distribution, and node coordinate settings rely entirely on manual experience, lacking the dynamic adaptability to the local deformation characteristics of the mold surface. This results in simulation results that are highly sensitive to mesh quality, making it difficult to accurately capture the subtle deformations and density gradient distribution of the mold surface at the end of pressing.

[0003] Current powder forming simulation optimization schemes often involve simple meshing and assembly of multi-region mold structures such as the upper punch area, lower punch area, die wall, and toothed cavity, with a lack of unified calibration and collaborative correction mechanisms for the coordinate references of mesh nodes in each region. When adjusting or refining the mesh for each local region separately, node misalignment, element distortion, and topological conflicts easily occur at the common boundaries of the regions, significantly reducing the convergence and reliability of subsequent pressing simulations. Furthermore, existing methods lack a complete optimization chain from candidate configuration evaluation and mesh node deviation identification to iterative updating of coordinate references. They also lack adaptive reconstruction methods based on simulation feedback for comparing and selecting different configurations, making it difficult to guarantee that the finally selected mold structure achieves optimal performance in the pressing densification process and surface retention. Summary of the Invention

[0004] To address the aforementioned shortcomings, this invention proposes a powder metallurgy mold optimization design method based on three-dimensional modeling of powder forming simulation. This method evaluates and filters candidate mold configurations during the pressing process in a simulation environment. By establishing a unified unit coordinate reference system, coordinate reference values ​​are distributed to each local mold region. Within each region, the positional deviation between the actual coordinates of the mesh units and the reference values ​​is identified, and the unit with the smallest deviation is selected as the optimization reference unit. The remaining units are then adjusted step-by-step. Subsequently, the boundary nodes of adjacent regions are collaboratively fused in the global integration module, updating the coordinate reference values ​​of all mesh units. Pressing simulation is then performed again based on the updated mesh system. Finally, the optimal configuration is determined by comprehensively ranking the surface deformation and density uniformity indices. This method effectively overcomes the shortcomings of traditional simulations, such as fixed mesh systems, inconsistent regional coordinates, and lack of iterative optimization, providing a systematic and high-precision simulation design approach for the structural optimization of powder metallurgy molds.

[0005] To achieve the above objectives, this application provides the following technical solution:

[0006] A powder metallurgy mold optimization design method based on three-dimensional modeling simulation of powder forming includes:

[0007] S1. Import the initial geometric configuration of the powder metallurgy mold to be optimized into the powder forming simulation environment, and select several candidate configurations from the candidate mold configuration set whose pressing and densification distribution field data meet the preset uniformity criterion.

[0008] S2, perform simulation meshing operation on the three-dimensional solid model corresponding to each candidate configuration selected, and distribute the coordinate reference values ​​of each calibrated simulation mesh unit to the calculation nodes corresponding to each local mold region divided by the candidate configuration.

[0009] S3, for each local mold region in each candidate configuration, determine the optimized reference element for that local mold region;

[0010] S4. For each local mold region of the optimized baseline unit in each candidate configuration, perform local positioning parameter densification and refinement processing, and summarize and send the adjusted positioning coordinate data of all simulation mesh units in each local mold region and the mesh unit affiliation information of the candidate configuration to which the local mold region belongs to to the global integration module.

[0011] S5, the global integration module receives the positioning coordinate data and grid cell affiliation information of all local mold regions from each candidate configuration, and obtains the updated coordinate reference values ​​of each simulation grid cell.

[0012] S6 reconstructs the simulation mesh system of the three-dimensional solid model of each candidate configuration based on the updated coordinate reference values ​​of each simulation mesh unit, and outputs the optimized design results as powder metallurgy molds.

[0013] Furthermore, S1 also includes:

[0014] A three-dimensional solid model corresponding to the initial geometric configuration is established in the simulation environment. Powder material property parameters and pressing boundary conditions are applied to the three-dimensional solid model. The powder forming pressing process simulation is run to obtain the pressing densification distribution field data corresponding to the initial geometric configuration. The initial geometric configuration is evaluated based on the pressing densification distribution field data. Several candidate configurations that satisfy the preset uniformity criterion are selected from the candidate mold configuration set.

[0015] The powder material properties include the initial relative density of the powder particles, the flow stress curve data of the powder material, the friction coefficient between the powder particles and the mold surface, and the volumetric compressibility modulus of the powder body during the pressing process.

[0016] Furthermore, in S1, the pressing boundary conditions include the pressing speed curve of the upper die punch, the support reaction force setting value of the lower die punch, the lateral constraint stiffness value of the female die, and the termination position parameter of the pressing stroke.

[0017] Furthermore, S2 also includes:

[0018] Determine the total number of simulation mesh elements required for each candidate configuration. Using the geometric centroid of each candidate configuration as the reference point, establish a unified element coordinate reference system for the simulation mesh space corresponding to the three-dimensional solid model of each candidate configuration. In this element coordinate reference system, calibrate the coordinate reference value of each simulation mesh element and distribute the calibrated coordinate reference value of each simulation mesh element to the calculation nodes corresponding to each local mold region divided by the candidate configuration.

[0019] The specific implementation method for establishing a unified element coordinate reference system in the simulation mesh space corresponding to the 3D solid model of each candidate configuration is as follows:

[0020] Using the projection point of the geometric centroid of each candidate configuration in three-dimensional space as the origin, and the direction of the principal inertial axis of the three-dimensional solid model of the candidate configuration as the coordinate axis direction, an orthogonal right-handed coordinate system is established as the unit coordinate reference system of the candidate configuration. In this unit coordinate reference system, each simulation mesh unit is assigned a unique coordinate reference value triplet according to the topological connection relationship of the simulation mesh unit.

[0021] Furthermore, in S2, the specific method for distributing the coordinate reference values ​​of each calibrated simulation mesh element to the computation nodes corresponding to each local mold region divided by the candidate configuration is as follows:

[0022] Based on the spatial occupancy of each local mold region in the three-dimensional solid model, the coordinate reference values ​​of the simulation mesh elements falling within the spatial occupancy range are packaged into a data packet. This data packet is then sent to the computing node corresponding to the local mold region through a parallel computing communication interface. Each computing node stores the data packet upon receiving it and returns an acknowledgment signal.

[0023] Furthermore, S3 also includes:

[0024] Obtain the actual spatial coordinate values ​​of each simulation mesh unit within the local mold area, calculate the spatial position deviation between the actual spatial coordinate value of each simulation mesh unit and the coordinate reference value in the unit coordinate reference system of the candidate configuration to which the local mold area belongs, select the simulation mesh unit corresponding to the coordinate reference value with the smallest spatial position deviation, and set the simulation mesh unit as the optimized reference unit of the local mold area.

[0025] The method for calculating the spatial position deviation between the actual spatial coordinate value of each simulation mesh element and the coordinate reference value of each element in the candidate configuration to which the local mold region belongs is as follows:

[0026] Calculate the Euclidean distance between the actual spatial coordinates of each simulation grid cell and each coordinate reference value in the cell coordinate reference system, and use the calculated Euclidean distance as the spatial position deviation between the simulation grid cell and the coordinate reference value.

[0027] Furthermore, S4 also includes:

[0028] The positioning coordinates of the simulation mesh units other than the optimized reference unit in each local mold area are adjusted step by step to approach the coordinate reference value of the optimized reference unit. After the step-by-step approximation adjustment is completed, the adjusted positioning coordinate data of all simulation mesh units in each local mold area and the 3D solid model mesh unit affiliation information of the candidate configuration to which the local mold area belongs are summarized and sent to the global integration module.

[0029] The specific implementation method for adjusting the positioning coordinates of the simulation mesh elements other than the optimized reference element within each local mold area to gradually approximate the coordinate reference value of its optimized reference element is as follows:

[0030] Determine the deviation vector between the current positioning coordinates of each simulation mesh cell (excluding the optimized reference cell) within the local mold area and the coordinate reference value of the optimized reference cell. Decompose the deviation vector into several levels of approximation sub-vectors according to a preset approximation step size ratio coefficient. In each level of approximation operation, add an approximation sub-vector to the positioning coordinates of the simulation mesh cell until the remaining deviation between the positioning coordinates of the simulation mesh cell and the coordinate reference value of the optimized reference cell is less than a preset deviation threshold.

[0031] Furthermore, S5 also includes:

[0032] Perform a collaborative fusion operation on the positioning coordinate data of all local mold regions of each candidate configuration. Based on the collaborative fusion result, recalculate the update amount of the coordinate reference value of each simulation mesh unit in the unit coordinate reference system in the three-dimensional solid model of each candidate configuration. Superimpose the update amount onto the current coordinate reference value of each simulation mesh unit to obtain the updated coordinate reference value of each simulation mesh unit.

[0033] The specific implementation method for performing collaborative fusion operation on the positioning coordinate data of all local mold regions of each candidate configuration is as follows:

[0034] Using the common boundary simulation mesh element between each local mold area as the fusion reference point, the positioning coordinate data of adjacent local mold areas at the common boundary are weighted and averaged. The original positioning coordinate data of each adjacent local mold area are replaced with the weighted averaged coordinate value. The global consistency verification of all fused positioning coordinate data is performed according to the topology consistency constraint. The positioning coordinate data that passes the global consistency verification is taken as the collaborative fusion result.

[0035] Furthermore, S6 also includes:

[0036] The powder forming and pressing process simulation was run again on the reconstructed simulation grid system to obtain updated pressing and densification distribution field data. The mold surface deformation index and density distribution uniformity index corresponding to each candidate configuration were extracted from the updated pressing and densification distribution field data. All candidate configurations were comprehensively ranked according to the mold surface deformation index and density distribution uniformity index. The candidate configuration with the highest comprehensive ranking was output as the optimized design result of the powder metallurgy mold.

[0037] The specific method for extracting the mold surface deformation index corresponding to each candidate configuration from the updated compacted distribution field data is as follows:

[0038] On the reconstructed simulation mesh system, all surface simulation mesh elements corresponding to the mold surface in the three-dimensional solid model are identified. The spatial displacement of each surface simulation mesh element at the moment of pressing is extracted. The difference between the maximum and minimum values ​​of the spatial displacement of all surface simulation mesh elements is calculated, and this difference is used as the mold surface deformation index corresponding to the candidate configuration.

[0039] Furthermore, in step S6, the specific method for extracting the density distribution uniformity index corresponding to each candidate configuration from the updated compressed densification distribution field data is as follows:

[0040] The local relative density value of each simulation mesh cell at the moment of compression completion is extracted on the reconstructed simulation mesh system. The arithmetic mean and standard deviation of the local relative density values ​​of all simulation mesh cells are calculated. The ratio of the standard deviation to the arithmetic mean is used as the density distribution uniformity index corresponding to the candidate configuration.

[0041] This invention belongs to the field of mold optimization design technology, and relates to a mold structure optimization method based on three-dimensional modeling of powder forming simulation. Multiple candidate mold configurations are imported into a powder forming simulation environment, and numerical simulation of the pressing process is performed to conduct preliminary evaluation and screening of the candidate configurations. The total number of simulation mesh elements is determined, and a unit coordinate reference system is established, selecting the optimal reference element. Other elements are adjusted stepwise and aggregated into a global integration module. The coordinates at the common boundary of adjacent regions are weighted and collaboratively fused to update the coordinate reference values ​​of all mesh elements. The simulation model is reconstructed, and the pressing simulation is performed again. All candidate configurations are comprehensively ranked, and the optimal design scheme is output. This invention effectively unifies the coordinate reference of multi-region meshes, overcomes the defects of fixed meshes and mismatched boundaries in traditional simulations, significantly improves the adaptability and reliability of the pressing process simulation, and can provide a systematic and high-precision optimization design approach for powder metallurgy molds. Attached Figure Description

[0042] Figure 1 The flowchart illustrates a powder metallurgy mold optimization design method based on three-dimensional modeling of powder forming simulation, as claimed in this embodiment of the invention. Detailed Implementation

[0043] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0044] The terms "first," "second," and "third" in this application are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of those features. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. All directional indications in the embodiments of this application, such as up, down, left, right, front, back, etc., are only used to explain the relative positional relationships and movements between components in a specific orientation as shown in the accompanying drawings. If the specific orientation changes, the directional indications will change accordingly. Furthermore, the terms "including" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0045] References to embodiments herein mean that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0046] According to the first embodiment of the present invention, referring to Figure 1 This invention claims protection for a powder metallurgy mold optimization design method based on three-dimensional modeling of powder forming simulation, comprising:

[0047] S1. Import the initial geometric configuration of the powder metallurgy mold to be optimized into the powder forming simulation environment, and select several candidate configurations from the candidate mold configuration set whose pressing and densification distribution field data meet the preset uniformity criterion.

[0048] S2, perform simulation meshing operation on the three-dimensional solid model corresponding to each candidate configuration selected, and distribute the coordinate reference values ​​of each calibrated simulation mesh unit to the calculation nodes corresponding to each local mold region divided by the candidate configuration.

[0049] S3, for each local mold region in each candidate configuration, determine the optimized reference element for that local mold region;

[0050] S4. For each local mold region of the optimized baseline unit in each candidate configuration, perform local positioning parameter densification and refinement processing, and summarize and send the adjusted positioning coordinate data of all simulation mesh units in each local mold region and the mesh unit affiliation information of the candidate configuration to which the local mold region belongs to to the global integration module.

[0051] S5, the global integration module receives the positioning coordinate data and grid cell affiliation information of all local mold regions from each candidate configuration, and obtains the updated coordinate reference values ​​of each simulation grid cell.

[0052] S6 reconstructs the simulation mesh system of the three-dimensional solid model of each candidate configuration based on the updated coordinate reference values ​​of each simulation mesh unit, and outputs the optimized design results as powder metallurgy molds.

[0053] In this embodiment, a certain type of iron-based powder metallurgy gear mold is used as the optimization object. This mold has an upper punch, a lower punch, and a female mold, with an initial geometric configuration of a cylinder with toothed cavities. During implementation, the initial 3D CAD model of the mold is first imported into commercially available powder forming simulation software such as DEFORM or MSC.Marc. A finite element solid geometry representation of the model is then established within the simulation environment. Powder material parameters are assigned to this solid model: the initial relative density of the powder particles is set to 0.48, and the flow stress of the powder is given by a user-defined hardening curve, specifically 320 MPa at an equivalent plastic strain of 0.1, 400 MPa at 0.3, and 460 MPa at 0.5; the coefficient of friction between the powder and the mold wall is set to 0.25, and the bulk compressive modulus of the powder is set to 1200 MPa. Subsequently, pressing boundary conditions were applied: the upper die punch moved downwards at a constant speed of 3 mm / s, with the stroke ending 20 mm from the initial position; the lower die punch maintained a fixed support reaction force of 5 kN; the lateral constraint stiffness of the female die was set to 150 kN / mm; the first pressing process simulation was performed using an explicit dynamic solver with a time step of 0.001 seconds and a total of 2000 steps. After the simulation, the relative density values ​​of each element inside the pressed part were extracted, and the variance of this relative density value in spatial distribution was used as an evaluation index. If the variance is not greater than 0.002, it is considered to meet the homogeneity criterion. In this embodiment, the variance of the initial configuration is 0.005, which does not meet the criterion. Therefore, the same simulation evaluation is performed one by one from the pre-constructed candidate configuration set, which includes nine configurations with punch fillet radii of 1 mm, 2 mm, and 3 mm, and female die taper of 0.5 degrees, 1 degree, and 1.5 degrees. Configurations with variance less than or equal to 0.002 are selected, and a total of three candidate configurations are selected.

[0054] For each selected candidate configuration, its 3D solid model was meshed. A hexahedral-dominated mapping meshing algorithm was used, with a global mesh size of 0.5 mm. Local feature regions such as the tooth root and tooth tip were refined to a size of 0.15 mm. After meshing, the total number of mesh elements for the three configurations were 182,345, 195,678, and 171,234, respectively. For each configuration, a local Cartesian coordinate system was established as the element coordinate reference system, using the geometric centroid (i.e., the centroid coordinate of the 3D solid volume) as the origin and the principal inertial axis direction of that configuration obtained by calculating the inertial matrix of the solid model as the X, Y, and Z axes. Within this reference system, the mesh elements were numbered according to the topological connection order, starting from the bottom-level nodes. The process involves incrementally assigning centroid coordinates to each grid cell as a coordinate reference value. For example, the centroid coordinates of cell number 1 are (0.523, 1.204, -0.876), in millimeters. All the assigned coordinate reference values ​​are then divided into spatial regions: the entire mold is divided into four local mold regions: the upper punch region, the lower punch region, the female mold wall region, and the toothed cavity region. For each local region, the grid cell numbers and their corresponding coordinate reference values ​​are packaged into a data message and sent to the corresponding computing node via the MPI parallel communication interface. Each region is allocated an independent CPU core, and after transmission, the system waits for each node to return a receipt confirmation flag.

[0055] For each local mold region in each candidate configuration, this embodiment specifically uses the tooth-shaped cavity region of the first candidate configuration as an example. This region contains 42,356 simulated mesh elements. The actual centroid coordinates of these elements in the current simulation coordinate system, the global Cartesian coordinate system, are obtained. For example, the actual coordinates of a certain element are (5.231, 8.452, 3.124), while the reference value (5.200, 8.400, 3.100) corresponding to that element is stored in the element coordinate reference system of this region. The straight-line distance between them is calculated: the square root of the sum of the squares of the differences in the three coordinate axes is calculated respectively. In this example, the differences are (0.031, 0.052, 0.024), and the square root of the sum of squares is approximately 0.065 mm. The reference values ​​of all elements in this region are then calculated one by one.

[0056] Next, local positioning parameter refinement processing is performed on the remaining units in each local mold area except for the optimized reference unit. Taking a non-reference unit numbered 12345 in the tooth cavity area with actual coordinates of (6.100, 9.200, 4.500) as an example, its corresponding optimized reference unit number is 12346, and the reference value of the reference unit's coordinates is (6.050, 9.100, 4.400). The deviation vector from the actual coordinates to the reference value is calculated as (0.050, 0.100, 0.100). The approximation step size ratio is set to 0.2, meaning that the deviation vector is adjusted by 20% each time. In the first level of approximation, the unit coordinates are adjusted to (6.100-0.010, 9.200-0.020, 4.500-0.020) = (6.090, 9.180, 4.480). In the second level, another 20% of the remaining amount is adjusted, and so on. In this embodiment, a deviation threshold of 0.005 mm is set; approximation stops when the remaining deviation is less than 0.005 mm. After 8 levels of approximation, the unit coordinates become (6.0501, 9.1002, 4.4001), satisfying the threshold. The above step-by-step approximation adjustment is performed on all non-reference units within the region. After completion, the adjusted positioning coordinates of each unit, along with the candidate configuration number and local region number to which the unit belongs, are summarized into a dataset and sent to a data collection program running on the main control node of the global integration module.

[0057] The global integration module receives data from all local regions of all candidate configurations. Since each region is adjusted independently, mesh mismatches or coordinate inconsistencies may occur at the common boundary of adjacent regions. In this embodiment, for the shared node at the common boundary between the upper punch region and the toothed cavity region, such as node number 5678, two adjusted coordinate values ​​were received from the two regions: (8.123, 5.234, 2.000) and (8.120, 5.230, 2.005). The global integration module takes a weighted average of these two values, with the weights proportional to the number of elements contained in each region: the upper punch region has more elements (weight 0.6), and the toothed cavity region has a weight of 0.4, resulting in the fused coordinates (8.1218, 5.2324, 2.002). A similar weighted average is performed on all common boundary nodes. Then, a topology consistency check is performed on all fused coordinates: the volume of each element is checked for positive values; if negative volumes exist, the weights are readjusted. In this embodiment, all element volumes are verified to be positive, confirming the fusion result is usable. Based on the fusion results, the update amount of the coordinate reference value of each grid cell is calculated—that is, the difference between the fused coordinates and the original reference value. For example, the original reference value of the cell corresponding to the above node is (8.100, 5.200, 2.000), and the update amount is (0.0218, 0.0324, 0.002). The update amount is superimposed on the current coordinate reference values ​​of all cells to obtain the updated coordinate reference values.

[0058] Finally, based on the updated coordinate reference values ​​of each simulation mesh element, the simulation mesh of the 3D solid model of each candidate configuration is reconstructed, that is, the element topology is regenerated according to the new node coordinates. The powder pressing simulation is run again on the reconstructed mesh, with the same material parameters and boundary conditions set, and the pressing process is calculated to obtain the updated relative density distribution field. From the simulation results, the surface elements corresponding to the mold surface are extracted and the element surfaces in contact with the mold surface are identified. The displacement of these surface elements along the mold axis Z at the end of pressing is calculated, and the maximum and minimum values ​​are recorded. In this embodiment, the maximum displacement of the first candidate configuration is 0.45 mm, the minimum displacement is 0.32 mm, and the difference of 0.13 mm is used as the surface deformation index. At the same time, the relative density values ​​of all internal elements are extracted, and their arithmetic mean is calculated to be 0.965, the standard deviation is 0.018, and the density uniformity index is 0.018 / 0.965≈0.0187. The same calculations were performed on other candidate configurations, and the deformation amounts of the three configurations were 0.13, 0.09, and 0.15 mm, respectively, with uniformity indices of 0.0187, 0.0142, and 0.0210, respectively. The smaller the deformation amount, the better. If the deformation amounts are the same, the smaller the uniformity index, the better. After sorting, the second configuration is the best overall, and the three-dimensional model of this configuration is output as the final optimization design result.

[0059] Furthermore, S1 also includes:

[0060] A three-dimensional solid model corresponding to the initial geometric configuration is established in the simulation environment. Powder material property parameters and pressing boundary conditions are applied to the three-dimensional solid model. The powder forming pressing process simulation is run to obtain the pressing densification distribution field data corresponding to the initial geometric configuration. The initial geometric configuration is evaluated based on the pressing densification distribution field data. Several candidate configurations that satisfy the preset uniformity criterion are selected from the candidate mold configuration set.

[0061] The powder material properties include the initial relative density of the powder particles, the flow stress curve data of the powder material, the friction coefficient between the powder particles and the mold surface, and the volumetric compressibility modulus of the powder body during the pressing process.

[0062] In this embodiment, the specific values ​​and settings of the powder material property parameters are as follows: For iron-based powder containing 2% copper and 0.8% graphite by mass, its initial relative density is obtained by dividing the loose density by the theoretical density. In this embodiment, the loose density is 3.2 g / cm³ and the theoretical density is 7.2 g / cm³, so the initial relative density is set to 0.444. The flow stress curve data uses stress-strain discrete points obtained from compression tests. In this embodiment, five points are used: stress of 280 MPa when strain is 0.05, 330 MPa when strain is 0.1, 390 MPa when strain is 0.2, 440 MPa when strain is 0.3, and 490 MPa when strain is 0.5. The intermediate values ​​are obtained by linear interpolation during simulation. The coefficient of friction between the powder and the mold surface was measured by an annular shear test. In this embodiment, it is 0.22 under lubricated conditions and 0.35 under dry friction conditions. The lubrication state is selected according to the pressing process. This embodiment uses lubrication, so it is set to 0.22. The volumetric compressive modulus of the powder was obtained by hydrostatic pressure test. When the hydrostatic pressure increases from 0 to 100 MPa, the volumetric strain changes by 0.08, and the modulus is 100 / 0.08 = 1250 MPa. This value is taken in this embodiment. All parameters are entered into the material library before simulation and assigned to the three-dimensional solid model as the basis data for all subsequent simulations.

[0063] Furthermore, in S1, the pressing boundary conditions include the pressing speed curve of the upper die punch, the support reaction force setting value of the lower die punch, the lateral constraint stiffness value of the female die, and the termination position parameter of the pressing stroke.

[0064] In this embodiment, the pressing boundary conditions are set according to the actual press parameters. The downward pressing speed curve of the upper punch adopts a three-segment structure: in the initial acceleration segment, the speed linearly increases from 0 to 4 mm / s within a stroke of 0 to 2 mm; in the intermediate constant speed segment, the speed remains at 4 mm / s within a stroke of 2 to 18 mm; and in the final deceleration segment, the speed linearly decreases to 0 within a stroke of 18 to 20 mm. The support reaction force of the lower punch is controlled by the hydraulic system, and in this embodiment, it is set to a constant 6 kN, which always pushes the powder upward during the pressing process. The lateral constraint stiffness value of the female die depends on the die frame structure. In this embodiment, the female die is fixed in a rigid die sleeve, and its radial stiffness is equivalent to 200 kN / mm, which is determined by finite element modal analysis. The termination position parameter of the pressing stroke is set to move downward 22 mm from the initial contact surface between the upper punch and the powder. Considering the overpressure, the upper punch displacement is 22 mm when the pressing actually ends, while the lower punch remains stationary. These boundary conditions are applied as time-load curves to the corresponding boundary surfaces of the simulation model.

[0065] Furthermore, S2 also includes:

[0066] Determine the total number of simulation mesh elements required for each candidate configuration. Using the geometric centroid of each candidate configuration as the reference point, establish a unified element coordinate reference system for the simulation mesh space corresponding to the three-dimensional solid model of each candidate configuration. In this element coordinate reference system, calibrate the coordinate reference value of each simulation mesh element and distribute the calibrated coordinate reference value of each simulation mesh element to the calculation nodes corresponding to each local mold region divided by the candidate configuration.

[0067] The specific implementation method for establishing a unified element coordinate reference system in the simulation mesh space corresponding to the 3D solid model of each candidate configuration is as follows:

[0068] Using the projection point of the geometric centroid of each candidate configuration in three-dimensional space as the origin, and the direction of the principal inertial axis of the three-dimensional solid model of the candidate configuration as the coordinate axis direction, an orthogonal right-handed coordinate system is established as the unit coordinate reference system of the candidate configuration. In this unit coordinate reference system, each simulation mesh unit is assigned a unique coordinate reference value triplet according to the topological connection relationship of the simulation mesh unit.

[0069] In this embodiment, the specific steps for establishing a unified unit coordinate reference system are as follows: First, the volume centroid of the 3D solid model of each candidate configuration is calculated by discretizing the solid into tetrahedrons and integrating. For example, for a certain candidate configuration, the centroid coordinates are calculated to be (15.234, 8.671, -2.345) mm. Then, the inertia tensor matrix of the solid is calculated, and the three principal inertial axis direction vectors are obtained through eigenvalue decomposition: (0.707, 0.707, 0), (-0.707, 0.707, 0), and (0, 0, 1), which are mutually orthogonal. A right-handed coordinate system is established with the centroid as the origin and these three directions as the X, Y, and Z axes. Within this coordinate system, each unit is assigned a coordinate reference value triplet in ascending order of its mesh unit number. The specific approach is as follows: Extract the centroid value of each element in the global coordinate system, and transform it to the new coordinate system through rotation and translation. The transformation formula is the product of the coordinate difference and the rotation matrix, but in practice, the coordinate transformation function of the simulation software is directly called. After transformation, for example, the new coordinates of element number 1 are (2.345, -1.234, 0.567), which is used as the coordinate reference value for that element. The reference values ​​of all elements are stored in a lookup table, where each row includes the element number and three coordinate values.

[0070] Furthermore, in S2, the specific method for distributing the coordinate reference values ​​of each calibrated simulation mesh element to the computation nodes corresponding to each local mold region divided by the candidate configuration is as follows:

[0071] Based on the spatial occupancy of each local mold region in the three-dimensional solid model, the coordinate reference values ​​of the simulation mesh elements falling within the spatial occupancy range are packaged into a data packet. This data packet is then sent to the computing node corresponding to the local mold region through a parallel computing communication interface. Each computing node stores the data packet upon receiving it and returns an acknowledgment signal.

[0072] In this embodiment, the specific method for distributing coordinate reference values ​​to the computational nodes of the local mold area is as follows: First, the local area is divided according to its spatial occupancy. For example, the entire mold is divided axially into an upper section (z>5 mm), a middle section (-5≤z≤5 mm), and a lower section (z<-5 mm), and radially into an inner and outer ring, forming four regions in total. For each region, all mesh elements are traversed. If the centroid coordinates of an element satisfy the spatial conditions of that region, the element number and its coordinate reference value are added to the data packet for that region. The data packet uses a compact binary format. The first 4 bytes are the region identifier, the next 4 bytes are the number of elements in that region, and the following bytes are the 4-byte integer number and 4-byte floating-point numbers for each element's three coordinate values, totaling 16 bytes per element. The data packet is sent to the corresponding computational node via the MPI_Send function. Each computational node runs an independent simulation process. After sending, the master node calls MPI_Recv to wait for an acknowledgment signal. Each node stores the received data in its local memory and returns an integer value of 1 to indicate acknowledgment. If no acknowledgment is received within a timeout period, the packet is resent up to three times.

[0073] Furthermore, S3 also includes:

[0074] Obtain the actual spatial coordinate values ​​of each simulation mesh unit within the local mold area, calculate the spatial position deviation between the actual spatial coordinate value of each simulation mesh unit and the coordinate reference value in the unit coordinate reference system of the candidate configuration to which the local mold area belongs, select the simulation mesh unit corresponding to the coordinate reference value with the smallest spatial position deviation, and set the simulation mesh unit as the optimized reference unit of the local mold area.

[0075] The method for calculating the spatial position deviation between the actual spatial coordinate value of each simulation mesh element and the coordinate reference value of each element in the candidate configuration to which the local mold region belongs is as follows:

[0076] Calculate the Euclidean distance between the actual spatial coordinates of each simulation grid cell and each coordinate reference value in the cell coordinate reference system, and use the calculated Euclidean distance as the spatial position deviation between the simulation grid cell and the coordinate reference value.

[0077] In this embodiment, the specific implementation of calculating the spatial position deviation is as follows: For a certain simulation mesh cell in a local mold region, let its actual spatial coordinates be x_real, y_real, z_real, and each coordinate reference value stored in the coordinate reference system of the cell in this region is x_base_i, y_base_i, z_base_i, where i ranges from 1 to N and represents the total number of cells in this region; calculate the Euclidean distance between them, specifically by calculating Δx = x_real - x_base_i and Δy = y_real - x_base_i respectively. The distance is calculated by first determining the base coordinates of a given unit, then calculating the sum of squares of the three differences, and finally taking the square root. In this embodiment, the actual coordinates of a certain unit are 12.345, 6.789, 3.210, and the first base value in the base value table is 12.300, 6.800, 3.200. The differences are 0.045, -0.011, and 0.010, and the sum of squares is 0.002025 + 0.000121 + 0.0001 = 0.002246, which gives a square root of 0.0474 mm. The distances between this unit and all base values ​​are calculated sequentially, and the minimum distance value and its corresponding base unit number are recorded. For example, if the minimum distance is 0.002 mm, corresponding to base unit number 5678, then this unit is selected as the optimized base unit.

[0078] Furthermore, S4 also includes:

[0079] The positioning coordinates of the simulation mesh units other than the optimized reference unit in each local mold area are adjusted step by step to approach the coordinate reference value of the optimized reference unit. After the step-by-step approximation adjustment is completed, the adjusted positioning coordinate data of all simulation mesh units in each local mold area and the 3D solid model mesh unit affiliation information of the candidate configuration to which the local mold area belongs are summarized and sent to the global integration module.

[0080] The specific implementation method for adjusting the positioning coordinates of the simulation mesh elements other than the optimized reference element within each local mold area to gradually approximate the coordinate reference value of its optimized reference element is as follows:

[0081] Determine the deviation vector between the current positioning coordinates of each simulation mesh cell (excluding the optimized reference cell) within the local mold area and the coordinate reference value of the optimized reference cell. Decompose the deviation vector into several levels of approximation sub-vectors according to a preset approximation step size ratio coefficient. In each level of approximation operation, add an approximation sub-vector to the positioning coordinates of the simulation mesh cell until the remaining deviation between the positioning coordinates of the simulation mesh cell and the coordinate reference value of the optimized reference cell is less than a preset deviation threshold.

[0082] In this embodiment, the specific implementation of the step-by-step approximation adjustment is as follows: Let the current positioning coordinates of a certain non-reference unit be P_cur, for example (2.100, 4.200, 1.500), and the coordinate reference value of its corresponding optimized reference unit be P_base, for example (2.000, 4.000, 1.400); First, calculate the deviation vector D = P_base - P_cur = (-0.100, -0.200, -0.100); Set the approximation step size ratio coefficient to 0.15, that is, adjust 15% of the deviation vector at each level. The first-level adjustment Δ1 = D × 0.15 = (-0.015, -0.030, -0.015), updating P_cur to P_cur + Δ1 = (2.085, 4.170, 1.485); at this point, the remaining deviation D_rem = P_base - P_new = (-0.085, -0.170, -0.085); the second level takes 15% of the remaining deviation, i.e., (-0.01275, -0.0255, -0.01275), and updates the coordinates; this process is repeated until the magnitude of the remaining deviation vector, i.e., the square root of the sum of the squares of each component, is less than the preset deviation threshold of 0.003 mm. In this embodiment, after 12 levels of adjustment, the magnitude of the remaining deviation is 0.0028 mm, satisfying the stopping condition, and the final coordinates are (2.0003, 4.0001, 1.3999). All non-reference elements independently undergo the above cyclic approximation operation according to their respective deviation vectors, and the number of approximation levels may be different for each element.

[0083] Furthermore, S5 also includes:

[0084] Perform a collaborative fusion operation on the positioning coordinate data of all local mold regions of each candidate configuration. Based on the collaborative fusion result, recalculate the update amount of the coordinate reference value of each simulation mesh unit in the unit coordinate reference system in the three-dimensional solid model of each candidate configuration. Superimpose the update amount onto the current coordinate reference value of each simulation mesh unit to obtain the updated coordinate reference value of each simulation mesh unit.

[0085] The specific implementation method for performing collaborative fusion operation on the positioning coordinate data of all local mold regions of each candidate configuration is as follows:

[0086] Using the common boundary simulation mesh element between each local mold area as the fusion reference point, the positioning coordinate data of adjacent local mold areas at the common boundary are weighted and averaged. The original positioning coordinate data of each adjacent local mold area are replaced with the weighted averaged coordinate value. The global consistency verification of all fused positioning coordinate data is performed according to the topology consistency constraint. The positioning coordinate data that passes the global consistency verification is taken as the collaborative fusion result.

[0087] In this embodiment, the specific implementation of the collaborative fusion operation is as follows: After receiving the adjusted coordinate data of all local regions, the global integration module first establishes a global node index table, marking the nodes shared by different regions, i.e., the nodes located on the region boundaries; for example, the common boundary between the upper punch region and the female mold wall region is a ring of nodes, totaling 256 nodes; for each common boundary node, the coordinate values ​​from the two regions are obtained. Assuming the node number is N1001, the coordinates from region A are (10.123, 5.678, 2.345), and the coordinates from region B are (10.110, 5.690, 2.340); the total number of units in region A is calculated, e.g., 50000, and the total number of units in region B is calculated, e.g., 30000, with weights of 50000 / 80000 = 0.625 and 30000 / 80000 = 0.625 respectively. 80000 = 0.375; the new coordinates after weighted averaging are (10.123×0.625+10.110×0.375,5.678×0.625+5.690×0.375,2.345×0.625+2.340×0.375)=(10.118125,5.6825,2.343125); this weighted average is performed on all common boundary nodes; afterwards, for internal non-common nodes, the coordinates remain unchanged; after completing the weighted average, a global consistency check is performed: the volume of each grid cell is recalculated, and it is checked whether there are negative volumes or abrupt changes in volume where the volume ratio of adjacent cells is greater than 3; in this embodiment, it is found that all cell volumes are positive, and the maximum volume ratio is 2.8, which passes the check; the coordinate dataset that passes the check is used as the final collaborative fusion result.

[0088] Furthermore, S6 also includes:

[0089] The powder forming and pressing process simulation was run again on the reconstructed simulation grid system to obtain updated pressing and densification distribution field data. The mold surface deformation index and density distribution uniformity index corresponding to each candidate configuration were extracted from the updated pressing and densification distribution field data. All candidate configurations were comprehensively ranked according to the mold surface deformation index and density distribution uniformity index. The candidate configuration with the highest comprehensive ranking was output as the optimized design result of the powder metallurgy mold.

[0090] The specific method for extracting the mold surface deformation index corresponding to each candidate configuration from the updated compacted distribution field data is as follows:

[0091] On the reconstructed simulation mesh system, all surface simulation mesh elements corresponding to the mold surface in the three-dimensional solid model are identified. The spatial displacement of each surface simulation mesh element at the moment of pressing is extracted. The difference between the maximum and minimum values ​​of the spatial displacement of all surface simulation mesh elements is calculated, and this difference is used as the mold surface deformation index corresponding to the candidate configuration.

[0092] In this embodiment, the specific method for extracting the mold surface deformation index is as follows: On the reconstructed simulation mesh system, the surface simulation mesh units are identified by searching the unit surfaces in the 3D solid model that are in contact with the mold surface. Specifically, the geometric surface ID of the mold surface is defined before the simulation, and after the pressing simulation is completed, the displacement fields of all nodes are output. All surface units, such as the cavity tooth surface, have a total of 2450 hexahedral units. For each unit, the displacement values ​​of all its nodes along the Z-axis pressing direction at the pressing termination time are extracted, and the average displacement of the eight nodes of the unit is taken as the surface displacement of the unit. Then, the maximum and minimum values ​​of the displacements of all surface units are counted. In this embodiment, the maximum value appears at the tooth tip at 0.52 mm, and the minimum value appears at the tooth root at 0.38 mm, with a difference of 0.14 mm. This difference is recorded as the mold surface deformation index of the candidate configuration. If there are multiple surfaces, such as the upper punch contact surface and the lower punch contact surface, the difference of each is calculated separately, and the maximum value is taken as the final index.

[0093] Furthermore, in step S6, the specific method for extracting the density distribution uniformity index corresponding to each candidate configuration from the updated compressed densification distribution field data is as follows:

[0094] The local relative density value of each simulation mesh cell at the moment of compression completion is extracted on the reconstructed simulation mesh system. The arithmetic mean and standard deviation of the local relative density values ​​of all simulation mesh cells are calculated. The ratio of the standard deviation to the arithmetic mean is used as the density distribution uniformity index corresponding to the candidate configuration.

[0095] In this embodiment, the specific method for extracting the density distribution uniformity index is as follows: From the updated simulation results, read the local relative density values ​​of all volume elements, for example, a total of 180,000 elements, at the end of the compression. Each element outputs a floating-point number between 0 and 1. These values ​​are stored in an array, and the arithmetic mean is calculated: sum all the relative density values ​​and divide by the total number of elements. In this embodiment, the sum is 173,700, and the average is 0.965. Next, the standard deviation is calculated: For each element, calculate the square of the difference between its relative density and the average, sum them up, divide by the total number of elements, and then take the square root. In this embodiment, the sum of squares is 0.05832, divided by 180,000, resulting in 0.000000324, and the square root is 0.018. Then, divide the standard deviation by the average to obtain the ratio 0.018 / 0.965≈0.01865, which is the density distribution uniformity index. The smaller the value, the more uniform the density distribution; if all the relative density values ​​of a certain configuration are equal, the standard deviation is zero and the index is zero, but this is impossible in practice; in this embodiment, the index is used as the second priority criterion for comprehensive ranking.

[0096] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0097] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units. The above are merely embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

[0098] The specific embodiments of the invention have been described in detail above, but they are only examples, and this application is not limited to the specific embodiments described above. For those skilled in the art, any equivalent modifications or substitutions to the invention are also within the scope of this application. Therefore, all equivalent changes, modifications, and improvements made without departing from the spirit and principles of this application should be covered within the scope of this application.

Claims

1. A method for optimizing the design of powder metallurgy molds based on three-dimensional modeling of powder forming simulation, characterized in that, include: S1. Import the initial geometric configuration of the powder metallurgy mold to be optimized into the powder forming simulation environment, and select several candidate configurations from the candidate mold configuration set whose pressing and densification distribution field data meet the preset uniformity criterion. S2, perform simulation meshing operation on the three-dimensional solid model corresponding to each candidate configuration selected, and distribute the coordinate reference values ​​of each calibrated simulation mesh unit to the calculation nodes corresponding to each local mold region divided by the candidate configuration. S3, for each local mold region in each candidate configuration, determine the optimized reference element for that local mold region; S4. For each local mold region of the optimized baseline unit in each candidate configuration, perform local positioning parameter densification and refinement processing, and summarize and send the adjusted positioning coordinate data of all simulation mesh units in each local mold region and the mesh unit affiliation information of the candidate configuration to which the local mold region belongs to to the global integration module. S5, the global integration module receives the positioning coordinate data and grid cell affiliation information of all local mold regions from each candidate configuration, and obtains the updated coordinate reference values ​​of each simulation grid cell. S6 reconstructs the simulation mesh system of the three-dimensional solid model of each candidate configuration based on the updated coordinate reference values ​​of each simulation mesh unit, and outputs the optimized design results as powder metallurgy molds.

2. The powder metallurgy mold optimization design method based on three-dimensional modeling of powder forming simulation according to claim 1, characterized in that, S1 also includes: A three-dimensional solid model corresponding to the initial geometric configuration is established in the simulation environment. Powder material property parameters and pressing boundary conditions are applied to the three-dimensional solid model. The powder forming pressing process simulation is run to obtain the pressing densification distribution field data corresponding to the initial geometric configuration. The initial geometric configuration is evaluated based on the pressing densification distribution field data. Several candidate configurations that satisfy the preset uniformity criterion are selected from the candidate mold configuration set. The powder material properties include the initial relative density of the powder particles, the flow stress curve data of the powder material, the friction coefficient between the powder particles and the mold surface, and the volumetric compressibility modulus of the powder body during the pressing process.

3. The powder metallurgy mold optimization design method based on three-dimensional modeling of powder forming simulation according to claim 2, characterized in that, In S1, the pressing boundary conditions include the pressing speed curve of the upper die punch, the support reaction force setting value of the lower die punch, the lateral constraint stiffness value of the female die, and the termination position parameter of the pressing stroke.

4. The powder metallurgy mold optimization design method based on three-dimensional modeling of powder forming simulation according to claim 3, characterized in that, S2 also includes: Determine the total number of simulation mesh elements required for each candidate configuration. Using the geometric centroid of each candidate configuration as the reference point, establish a unified element coordinate reference system for the simulation mesh space corresponding to the three-dimensional solid model of each candidate configuration. In this element coordinate reference system, calibrate the coordinate reference value of each simulation mesh element and distribute the calibrated coordinate reference value of each simulation mesh element to the calculation nodes corresponding to each local mold region divided by the candidate configuration. The specific implementation method for establishing a unified element coordinate reference system in the simulation mesh space corresponding to the 3D solid model of each candidate configuration is as follows: Using the projection point of the geometric centroid of each candidate configuration in three-dimensional space as the origin, and the direction of the principal inertial axis of the three-dimensional solid model of the candidate configuration as the coordinate axis direction, an orthogonal right-handed coordinate system is established as the unit coordinate reference system of the candidate configuration. In this unit coordinate reference system, each simulation mesh unit is assigned a unique coordinate reference value triplet according to the topological connection relationship of the simulation mesh unit.

5. The powder metallurgy mold optimization design method based on three-dimensional modeling of powder forming simulation according to claim 4, characterized in that, In step S2, the specific method for distributing the coordinate reference values ​​of each calibrated simulation mesh element to the computation nodes corresponding to each local mold region divided by the candidate configuration is as follows: Based on the spatial occupancy of each local mold region in the three-dimensional solid model, the coordinate reference values ​​of the simulation mesh elements falling within the spatial occupancy range are packaged into a data packet. This data packet is then sent to the computing node corresponding to the local mold region through a parallel computing communication interface. Each computing node stores the data packet upon receiving it and returns an acknowledgment signal.

6. The powder metallurgy mold optimization design method based on three-dimensional modeling of powder forming simulation according to claim 5, characterized in that, S3 also includes: Obtain the actual spatial coordinate values ​​of each simulation mesh unit within the local mold area, calculate the spatial position deviation between the actual spatial coordinate value of each simulation mesh unit and the coordinate reference value in the unit coordinate reference system of the candidate configuration to which the local mold area belongs, select the simulation mesh unit corresponding to the coordinate reference value with the smallest spatial position deviation, and set the simulation mesh unit as the optimized reference unit of the local mold area. The method for calculating the spatial position deviation between the actual spatial coordinate value of each simulation mesh element and the coordinate reference value of each element in the candidate configuration to which the local mold region belongs is as follows: Calculate the Euclidean distance between the actual spatial coordinates of each simulation grid cell and each coordinate reference value in the cell coordinate reference system, and use the calculated Euclidean distance as the spatial position deviation between the simulation grid cell and the coordinate reference value.

7. The powder metallurgy mold optimization design method based on three-dimensional modeling of powder forming simulation according to claim 6, characterized in that, S4 also includes: The positioning coordinates of the simulation mesh units other than the optimized reference unit in each local mold area are adjusted step by step to approach the coordinate reference value of the optimized reference unit. After the step-by-step approximation adjustment is completed, the adjusted positioning coordinate data of all simulation mesh units in each local mold area and the 3D solid model mesh unit affiliation information of the candidate configuration to which the local mold area belongs are summarized and sent to the global integration module. The specific implementation method for adjusting the positioning coordinates of the simulation mesh elements other than the optimized reference element within each local mold area to gradually approximate the coordinate reference value of its optimized reference element is as follows: Determine the deviation vector between the current positioning coordinates of each simulation mesh cell (excluding the optimized reference cell) within the local mold area and the coordinate reference value of the optimized reference cell. Decompose the deviation vector into several levels of approximation sub-vectors according to a preset approximation step size ratio coefficient. In each level of approximation operation, add an approximation sub-vector to the positioning coordinates of the simulation mesh cell until the remaining deviation between the positioning coordinates of the simulation mesh cell and the coordinate reference value of the optimized reference cell is less than a preset deviation threshold.

8. The powder metallurgy mold optimization design method based on three-dimensional modeling of powder forming simulation according to claim 7, characterized in that, S5 also includes: Perform a collaborative fusion operation on the positioning coordinate data of all local mold regions of each candidate configuration. Based on the collaborative fusion result, recalculate the update amount of the coordinate reference value of each simulation mesh unit in the unit coordinate reference system in the three-dimensional solid model of each candidate configuration. Superimpose the update amount onto the current coordinate reference value of each simulation mesh unit to obtain the updated coordinate reference value of each simulation mesh unit. The specific implementation method for performing collaborative fusion operation on the positioning coordinate data of all local mold regions of each candidate configuration is as follows: Using the common boundary simulation mesh element between each local mold area as the fusion reference point, the positioning coordinate data of adjacent local mold areas at the common boundary are weighted and averaged. The original positioning coordinate data of each adjacent local mold area are replaced with the weighted averaged coordinate value. The global consistency verification of all fused positioning coordinate data is performed according to the topology consistency constraint. The positioning coordinate data that passes the global consistency verification is taken as the collaborative fusion result.

9. The powder metallurgy mold optimization design method based on three-dimensional modeling of powder forming simulation according to claim 8, characterized in that, S6 also includes: The powder forming and pressing process simulation was run again on the reconstructed simulation grid system to obtain updated pressing and densification distribution field data. The mold surface deformation index and density distribution uniformity index corresponding to each candidate configuration were extracted from the updated pressing and densification distribution field data. All candidate configurations were comprehensively ranked according to the mold surface deformation index and density distribution uniformity index. The candidate configuration with the highest comprehensive ranking was output as the optimized design result of the powder metallurgy mold. The specific method for extracting the mold surface deformation index corresponding to each candidate configuration from the updated compacted distribution field data is as follows: On the reconstructed simulation mesh system, all surface simulation mesh elements corresponding to the mold surface in the three-dimensional solid model are identified. The spatial displacement of each surface simulation mesh element at the moment of pressing is extracted. The difference between the maximum and minimum values ​​of the spatial displacement of all surface simulation mesh elements is calculated, and this difference is used as the mold surface deformation index corresponding to the candidate configuration.

10. The powder metallurgy mold optimization design method based on three-dimensional modeling of powder forming simulation according to claim 9, characterized in that, In step S6, the specific method for extracting the density distribution uniformity index corresponding to each candidate configuration from the updated compressed densification distribution field data is as follows: The local relative density value of each simulation mesh cell at the moment of compression completion is extracted on the reconstructed simulation mesh system. The arithmetic mean and standard deviation of the local relative density values ​​of all simulation mesh cells are calculated. The ratio of the standard deviation to the arithmetic mean is used as the density distribution uniformity index corresponding to the candidate configuration.

Images