Mold stress simulation test method and system based on structure performance of injection molded part

Abstract

This application provides a mold stress simulation testing method and system based on the structural performance of injection molded parts, belonging to the field of computer simulation testing technology. First, it obtains a set of structural performance parameters of the injection molded part and a set of mold operating condition parameters. Then, it analyzes the structural performance parameters of the injection molded part to generate a network of internal stress transmission trunk and branch paths, and maps stress loading boundary conditions to the mold operating condition parameters accordingly. The mapped set of stress loading boundary conditions is input into a mold stress field distribution simulation model for iterative calculation, outputting data on the surface stress of the mold cavity, internal pressure, and locking force field distribution of the closed contact surface. Finally, the calculation results are reverse-mapped to obtain predicted data on residual stress, warpage deformation, and shrinkage indentation distribution after demolding of the injection molded part. This invention can accurately predict the quality of injection molded parts and optimize mold design.

Description

Technical Field

[0001] This application relates to the field of computer simulation testing technology, and more specifically, to a mold stress simulation testing method and system based on the structural performance of injection molded parts. Background Technology

[0002] Traditional methods often face numerous challenges in evaluating the relationship between the structural performance of injection molded parts and mold stress. On the one hand, considerations of the structural performance parameters of injection molded parts are typically limited to simple geometric dimensions and surface quality, making it difficult to comprehensively and deeply analyze the complex stress transmission mechanisms within the part. For example, key parameters such as the geometric configuration distribution, rib layout, and gate location of the injection molded part have a significant impact on the stress transmission path within the part, but traditional methods lack a systematic study of the relationship between these parameters and the stress transmission path.

[0003] On the other hand, mold operating parameters such as mold temperature zone control, pressure segmented loading, and locking sequence control are often set based on empirical values ​​or simple experimental data in traditional methods, which cannot be accurately mapped to the stress loading boundary conditions on the mold cavity surface. This makes it difficult to accurately simulate the stress conditions of the mold during actual production in mold stress simulation tests, and consequently, it is impossible to accurately predict quality problems such as residual stress, warpage, and shrinkage marks after the injection molded part is demolded. Summary of the Invention

[0004] In view of this, the purpose of this application is to provide a mold stress simulation test method and system based on the structural performance of injection molded parts.

[0005] According to a first aspect of this application, a mold stress simulation testing method based on the structural performance of injection molded parts is provided, the method comprising: Obtain a set of structural performance parameters for injection molded parts and a set of mold operating parameters. The set of structural performance parameters for injection molded parts includes geometric configuration distribution parameters, rib layout parameters, and gate location parameters. The set of mold operating parameters includes mold temperature zone control parameters, mold pressure segmented loading parameters, and mold locking timing control parameters. The set of structural performance parameters of the injection molded part is subjected to stress transmission path analysis processing to generate a main stress transmission path network and a branch stress transmission path network inside the injection molded part. The main stress transmission path network inside the injection molded part includes a first set of stress transmission channels extending along the main direction of the geometric configuration of the injection molded part. The branch stress transmission path network inside the injection molded part includes a second set of stress transmission channels extending from the branch path network to the local feature area of ​​the injection molded part. Based on the main path network of stress transmission inside the injection molded part and the branch path network of stress transmission inside the injection molded part, the set of mold working condition parameters is subjected to stress loading boundary condition mapping to obtain the set of stress loading boundary conditions for each spatial region on the surface of the mold cavity. The set of stress loading boundary conditions includes a first set of stress loading boundary conditions corresponding to the end of the first set of stress transmission channels and a second set of stress loading boundary conditions corresponding to the end of the second set of stress transmission channels. The set of stress loading boundary conditions is input into the mold stress field distribution simulation model for iterative calculation of stress field distribution, and the results of the iterative calculation of stress field distribution are output. The mold stress field distribution simulation model uses the set of stress loading boundary conditions as driving boundary conditions and the main path network and branch path network of stress transmission inside the injection molded part as stress propagation constraints. The results of the iterative calculation of stress field distribution independently include stress field distribution data on the surface of the mold cavity, pressure field distribution data inside the mold cavity, and locking force field distribution data on the mold closed contact surface. The stress field distribution iterative calculation results are subjected to inverse mapping processing of injection molded part structural performance to obtain the inverse mapping results of injection molded part structural performance. The inverse mapping results of injection molded part structural performance include the predicted data of residual stress distribution after demolding, the predicted data of warpage deformation distribution after demolding, and the predicted data of shrinkage indentation distribution after demolding.

[0006] According to a second aspect of this application, a mold stress simulation testing system based on the structural performance of injection molded parts is provided. The mold stress simulation testing system based on the structural performance of injection molded parts includes a machine-readable storage medium and a processor. The machine-readable storage medium stores machine-executable instructions. When the processor executes the machine-executable instructions, the mold stress simulation testing system based on the structural performance of injection molded parts implements the aforementioned mold stress simulation testing method based on the structural performance of injection molded parts.

[0007] Based on any of the above aspects, the technical effect of this application is as follows: By acquiring the set of structural performance parameters of the injection molded part and the set of mold operating parameters, the stress transmission path of the injection molded part is analyzed using the set of structural performance parameters. This generates a main path network and branch path network for stress transmission within the injection molded part, revealing the stress transmission pattern within the part. Based on this stress transmission path network, stress loading boundary conditions are mapped onto the set of mold operating parameters, resulting in more accurate stress loading boundary conditions for each spatial region on the mold cavity surface. This effectively simulates the stress conditions of the mold in actual production. The set of stress loading boundary conditions is then input into the mold stress field distribution simulation model for iterative calculation. The output iterative calculation results include stress field distribution data on the mold cavity surface, pressure field distribution data inside the mold cavity, and locking force field distribution data on the mold closed contact surface, reflecting the stress state of the mold under different operating conditions. Finally, the iterative calculation results of the stress field distribution are subjected to inverse mapping of the injection molded part's structural performance. The resulting inverse mapping results accurately predict the residual stress distribution, warpage distribution, and shrinkage mark distribution after demolding, significantly improving the efficiency and quality stability of injection molding production. Attached Figure Description

[0008] Figure 1 A schematic flowchart of the mold stress simulation test method based on the structural performance of injection molded parts provided in the embodiments of this application is shown. Figure 2 This illustration shows a schematic diagram of the component structure of a mold stress simulation test system based on the structural performance of injection molded parts, provided in an embodiment of this application, for implementing the above-described mold stress simulation test method based on the structural performance of injection molded parts. Detailed Implementation

[0009] Figure 1 This paper illustrates a flowchart of a mold stress simulation testing method and system based on the structural performance of injection molded parts, as provided in an embodiment of this application. The detailed steps include: Step S110: Obtain the set of structural performance parameters of the injection molded part and the set of mold operating parameters. The set of structural performance parameters of the injection molded part includes the geometric configuration distribution parameters of the injection molded part, the rib layout parameters of the injection molded part, and the gate position parameters of the injection molded part. The set of mold operating parameters includes the mold temperature zone control parameters, the mold pressure segmented loading parameters, and the mold locking timing control parameters.

[0010] The geometric configuration distribution parameters of the injection molded part include triangular facet mesh data of the outer surface of the injection molded part, wall thickness distribution field data, extension lines of the main direction of the geometric configuration, and extension lines of the secondary direction of the geometric configuration. The wall thickness distribution field data is stored in the form of a wall thickness scalar field on the volume grid nodes, with the node wall thickness numerical code T_i. The main direction extension line is a spatial curve along the length direction of the injection molded part, consisting of a continuous sequence of spatial point coordinates; the secondary direction extension line is a spatial curve along the width direction. The rib layout parameters of the injection molded part include the rib extension direction line, the rib height distribution curve H(s), the rib root fillet radius distribution curve R(s), and the connection boundary profile of the rib and the base body connection area. The rib extension direction line is the spatial trajectory of the rib neutral axis; the connection boundary profile consists of a closed sequence of spatial point coordinates. The gate location parameters of the injection molded part include the gate location coordinates (Xg, Yg, Zg), the gate shape profile point sequence, and the gate direction vector (Vx, Vy, Vz).

[0011] The mold temperature zone control parameters include the spatial boundaries of each temperature control zone and the corresponding temperature loading values. The mold pressure segment loading parameters include the time interval [t1_k, t2_k] of each pressure loading stage and the corresponding pressure loading values. The mold locking timing control parameters include the start time T_m, duration D_m, and corresponding locking force loading values ​​of each locking force loading stage.

[0012] Step S120: Perform force transmission path analysis on the set of structural performance parameters of the injection molded part to generate a main stress transmission path network and a branch path network inside the injection molded part. The main path network contains a set of first stress transmission channels extending along the main direction of the geometric configuration; the branch path network contains a set of second stress transmission channels extending from the bifurcation of the main path to local feature regions.

[0013] Step S121: Analyze the main direction extension lines and secondary direction extension lines in the geometric configuration distribution parameters of the injection molded part, and generate an initial set of main trunk transfer paths extending along the main direction of the geometric configuration based on the main direction extension lines. The path direction of each initial main trunk transfer path maintains a spatial parallel relationship with the direction of the main direction extension line.

[0014] Extract the spatial point coordinate sequence Lm={P1, P2, ..., Pn} of the main direction extension line from the geometric configuration distribution parameters of the injection molded part. Taking each point Pj on Lm as the starting point of the path, extend it into the injection molded part along the tangent direction Tj at Pj, with an extension step size of a preset value S0. The termination condition is that the path touches the boundary of the injection molded part or intersects with the secondary direction extension line. Each initial main trunk transmission path is recorded as a spatial point coordinate sequence Qj={Qj1, Qj2, ..., Qjm}, and the set of all paths constitutes the set S0.

[0015] Step S122: Based on the extension direction line of the stiffener and the connection boundary contour of the connection area between the stiffener and the matrix, select the target main trunk transmission path that has a spatial intersection relationship with the extension direction line of the stiffener from the initial main trunk transmission path set, and mark it as the stiffener-associated main trunk transmission path.

[0016] For each path Qj in S0, traverse all stiffeners. Extract the extension direction lines Lr={R1, R2, ..., Rp} and the connecting boundary contours Bd={B1, B2, ..., Bq} of the current stiffener. Calculate the Euclidean distance between each point on Qj and each point on Lr. If there is a distance less than the threshold Dc, it is determined that there is a spatial intersection relationship. Alternatively, determine whether Qj passes through the spatial surface enclosed by Bd: calculate the minimum signed distance between Qj and Bd. If the sign changes from positive to negative and the absolute value is less than the threshold Dt, it is determined that it passes through. If either condition is met, Qj is selected as path Ra.

[0017] Step S123: Extract the spatial intersection coordinates of the main stress transmission path associated with the stiffener and the extension direction line of the stiffener. Using the spatial intersection coordinates as the starting point of the stress transmission path bifurcation, extend along the extension direction line of the stiffener away from the main stress transmission path associated with the stiffener to generate an initial set of branch transmission paths connected to the main stress transmission path associated with the stiffener.

[0018] For each path Ra and its corresponding stiffener extension direction line Lr, calculate the spatial intersection point Pc of Ra and Lr. The calculation method is as follows: for each point A_i on Ra and each point B_j on Lr, calculate the shortest spatial distance between line segment A_iA_{i+1} and line segment B_jB_{j+1}. If the distance is less than Dc, then take the midpoint of the two closest points on the two line segments as Pc. Starting from Pc, extend along Lr in a direction away from Ra (i.e., from Pc to B_{j+1}), with an extension step size of Sb. The termination condition is when the path touches the end point of the stiffener. Each initial branch path is recorded as a point sequence Br_c = {C_c1, C_c2, ..., C_cr}, and the set of all branch paths constitutes set B0.

[0019] Step S124: Extend the end of each initial branch transfer path in the initial branch transfer path set to the coordinates of the end point of the rib plate, and adjust the direction of the end of the initial branch transfer path at the coordinates of the end point of the rib plate according to the boundary of the gate adjacent area, so that it points to the local feature area of ​​the injection molded part within the boundary of the gate adjacent area, thereby generating the final branch transfer path set.

[0020] For each path Br_c in B0, extract the coordinates of the end point Pe (i.e., the last point of Lr) of the corresponding rib. Starting from the last point Ce of Br_c, continue extending along the direction (Pe-Ce) until reaching Pe. Then, extract the gate proximity region boundary Gn from the injection molded part gate location parameters. This gate proximity region boundary consists of a sequence of spatial point coordinates. Calculate the direction vector Va from Pe to the nearest point within Gn, and linearly interpolate the direction of the branch path end from the original direction to the Va direction: the interpolation point position Px = Pe + α·Va, where α increases from 0 to the coefficient α0. Generate the final branch propagation path Bf_c, whose point sequence contains all points from Pc to Pe and then to the adjusted end. The set of all Bf_c constitutes the set Bf.

[0021] Step S125: Perform path network topology connection processing on the main stress transmission path associated with the stiffener plate and other main stress transmission paths in the initial set of main stress transmission paths, except for the main stress transmission path associated with the stiffener plate, to form a main stress transmission path network for the internal stress transmission of the injection molded part, in which multiple main stress transmission paths are interconnected through path intersections. Each main stress transmission path corresponds to a first stress transmission channel in the first set of first stress transmission channels.

[0022] Extract all paths Ra obtained in step S122, and other paths Ro not marked as Ra in set S0 obtained in step S121. Detect spatial intersections between these paths: for any two paths U and V, calculate the nearest point pair between them. If the nearest distance is less than a threshold Te, mark the midpoint of the pair as an intersection point Pj. On each path, divide the path into several segments according to Pj, with each segment connecting two Pj or one Pj and an end point. Organize all segments into a graph structure Gt=(Vt, Et) according to their connection relationships, where Vt is the set of all Pj and end points, and Et is the set of segments connecting these points. This Gt is the main stress transmission path network inside the injection molded part, with each edge in the graph corresponding to a channel in the first stress transmission channel set. Gt is stored in the form of an adjacency list, where each node contains a node number and three-dimensional coordinates, and each edge contains a start point number, an end point number, and a path point sequence.

[0023] Step S126: Perform path network topology connection processing on each final branch transmission path in the final branch transmission path set to form an internal stress transmission branch path network for the injection molded part, in which multiple branch transmission paths are connected to the main stress transmission path network inside the injection molded part through the coordinates of spatial intersection points. Each branch transmission path corresponds to a second stress transmission channel in the second stress transmission channel set.

[0024] Extract all paths Bf_c obtained in step S124. For each Bf_c, its starting point is the spatial intersection point Pc calculated in step S123, which also belongs to the node set Vt in the backbone path network Gt. Connect Bf_c as a branch path to the corresponding Pc node in Gt. Detect spatial intersection points between different branch paths: For any two branch paths D and E, calculate the closest point pair between them. If the closest distance is less than Te, mark the midpoint of the point pair as the branch intersection point Pb. On each branch path, divide the path into several segments according to Pb. Organize all branch path segments into a graph structure Gb=(Vb, Eb) according to the connection relationship, where Vb is the set of all Pb and the end points of the branch paths, and Eb is the set of segments connecting the above points. At the same time, establish the connection relationship between Gb and Gt: For each branch path starting point Pc connected to the backbone network, record the corresponding backbone network node ID in Gb. Gb represents the stress transmission branch path network within the injection molded part. Each edge in the diagram corresponds to a channel in the second set of stress transmission channels. Gb is stored in the form of an adjacency list. Each node contains a node number, three-dimensional coordinates, and the ID of the main node it connects to (if it exists). Each edge contains a start point number, an end point number, and a sequence of path points.

[0025] Step S130: Perform stress loading boundary condition mapping processing on the mold working condition parameter set according to the main stress transmission path network and the branch stress transmission path network inside the injection molded part, to obtain the stress loading boundary condition set of each spatial region on the surface of the mold cavity. The stress loading boundary condition set includes a first stress loading boundary condition subset corresponding to the end of the first stress transmission channel set and a second stress loading boundary condition subset corresponding to the end of the second stress transmission channel set.

[0026] Based on the main stress transmission path network Gt generated in step S125 and the branch path network Gb generated in step S126, stress loading boundary condition mapping processing is performed on the set of mold working condition parameters obtained in step S110.

[0027] Step S131: Extract the position coordinates of the end point of each first stress transmission channel in the stress transmission backbone path network inside the injection molded part, and perform a coordinate projection mapping operation in the spatial coordinate system of the mold cavity surface according to the position coordinates of the end point of the channel to obtain the boundary of the first projection area of ​​the mold cavity surface corresponding to the position coordinates of the end point of each first stress transmission channel.

[0028] Traverse each edge in the edge set Et of the backbone path network Gt, and obtain the coordinates P_end (i.e., the coordinates of the terminating node of the edge) of the first stress transfer channel corresponding to each edge. Transform the coordinates P_end from the geometric space coordinate system of the injection molded part to the space coordinate system of the mold cavity surface. The transformation method is to obtain the spatial mapping matrix M_map between the outer surface of the injection molded part and the surface of the mold cavity. This spatial mapping matrix describes the one-to-one correspondence between the points on the surface of the injection molded part and the points on the surface of the mold cavity. M_map is a 3x3 transformation matrix. Calculate the projection point coordinates P_proj = M_map·P_end. With P_proj as the center point, generate the boundary of the projection region in the space coordinate system of the mold cavity surface. The generation method is to draw a circle with P_proj as the center and a preset radius R_proj as the radius in the tangent plane of the mold cavity surface, and discretize the circle into N points to obtain a closed spatial point sequence Boundary_proj. The Boundary_proj is the boundary of the first projected region on the mold cavity surface corresponding to the end of the current first stress transmission channel. Repeat the above operation for all first stress transmission channels to obtain the set of first projected region boundaries, Set_B1.

[0029] Step S132: Extract the position coordinates of the channel end point of each second stress transmission channel in the stress transmission branch path network inside the injection molded part, and perform a coordinate projection mapping operation in the spatial coordinate system of the mold cavity surface according to the position coordinates of the channel end point to obtain the boundary of the second projection area of ​​the mold cavity surface corresponding to the position coordinates of the channel end point of each second stress transmission channel.

[0030] Traverse each edge in the edge set Eb of the branch path network Gb, and obtain the coordinates P_end_b (i.e., the coordinates of the terminating node of the edge) of the second stress transfer channel corresponding to each edge. Using the same coordinate projection mapping operation as in step S131, project P_end_b onto the spatial coordinate system of the mold cavity surface, obtaining the projection point coordinates P_proj_b = M_map·P_end_b. With P_proj_b as the center point and a preset radius R_proj_b as the radius, generate a closed spatial point sequence Boundary_proj_b in the tangent plane of the mold cavity surface. Repeat the above operation for all second stress transfer channels to obtain the boundary set Set_B2 of the second projection region.

[0031] Step S133: Analyze the mold temperature zoning control parameters in the mold working condition parameter set, obtain the spatial boundaries of each temperature control zone contained in the mold temperature zoning control parameters and the temperature loading value corresponding to each temperature control zone, perform spatial overlap detection processing on the spatial boundaries of each temperature control zone with the first projection area boundary and the second projection area boundary of the mold cavity surface, respectively, and generate a first temperature loading boundary condition subset corresponding to the first projection area boundary of the mold cavity surface and a second temperature loading boundary condition subset corresponding to the second projection area boundary of the mold cavity surface.

[0032] Extract the spatial boundary T_zone_boundary_w and the corresponding temperature loading value T_val_w of each temperature control zone from the mold temperature zoning control parameters obtained in step S110. For each first projection zone boundary Boundary_proj in Set_B1, calculate its spatial overlap area with each T_zone_boundary_w. The spatial overlap area detection method is as follows: determine whether the area enclosed by Boundary_proj intersects with the area enclosed by T_zone_boundary_w. If they intersect, calculate the overlap area ratio. When the overlap area ratio exceeds a preset threshold O_t, assign the temperature loading value T_val_w corresponding to T_zone_boundary_w to Boundary_proj. All temperature loading values ​​assigned to the first projection zone boundaries constitute the first temperature loading boundary condition subset Set_T1. Similarly, for each second projection zone boundary Boundary_proj_b in Set_B2, calculate its spatial overlap area with each T_zone_boundary_w, assign temperature loading values, and constitute the second temperature loading boundary condition subset Set_T2.

[0033] Step S134: Analyze the mold pressure segmented loading parameters in the mold working condition parameter set, obtain the time interval of each pressure loading stage and the pressure loading value corresponding to each pressure loading stage contained in the mold pressure segmented loading parameters, perform time overlap interval detection processing on the time interval of each pressure loading stage and the time axis of the injection molding process, and generate a first pressure loading boundary condition subset corresponding to the boundary of the first projection area of ​​the mold cavity surface and a second pressure loading boundary condition subset corresponding to the boundary of the second projection area of ​​the mold cavity surface.

[0034] Extract the time intervals [t1_k, t2_k] and corresponding pressure loading values ​​P_val_k for each pressure loading stage from the mold pressure segmented loading parameters obtained in step S110. Define the total time axis of the injection molding process as [0, T_total]. For each first projection region boundary Boundary_proj in Set_B1, calculate the time overlap interval between each time interval [t1_k, t2_k] and the total time axis. When the length of the time overlap interval is greater than the preset threshold O_p, assign the pressure loading value P_val_k to the time period of Boundary_proj within the time overlap interval. All pressure loading values ​​assigned to the first projection region boundaries and their corresponding time periods constitute the first pressure loading boundary condition subset Set_P1. Similarly, for each second projection region boundary Boundary_proj_b in Set_B2, assign pressure loading values ​​to form the second pressure loading boundary condition subset Set_P2.

[0035] Step S135: Analyze the mold locking timing control parameters in the mold working condition parameter set, obtain the start time and duration of each locking force loading stage contained in the mold locking timing control parameters, as well as the locking force loading value corresponding to each locking force loading stage, generate a locking force loading time window based on the start time and duration, perform time overlap interval detection processing on the locking force loading time window and the time axis of the injection molding process, and generate a first locking force loading boundary condition subset corresponding to the boundary of the first projection area of ​​the mold cavity surface and a second locking force loading boundary condition subset corresponding to the boundary of the second projection area of ​​the mold cavity surface.

[0036] Extract the start time T_m, duration D_m, and corresponding locking force loading value F_val_m for each locking force loading stage from the mold locking timing control parameters obtained in step S110. For each locking force loading stage, generate a locking force loading time window [T_m, T_m+D_m]. For each first projection region boundary Boundary_proj in Set_B1, calculate the time overlap interval between each time window [T_m, T_m+D_m] and the total time axis. When the length of the time overlap interval is greater than a preset threshold O_f, assign the locking force loading value F_val_m to the time period within the time overlap interval of the Boundary_proj. All locking force loading values ​​assigned to the first projection region boundaries and their corresponding time periods constitute the first locking force loading boundary condition subset Set_F1. Similarly, for each second projection region boundary Boundary_proj_b in Set_B2, assign locking force loading values ​​to form the second locking force loading boundary condition subset Set_F2.

[0037] Step S136: Perform boundary condition association and combination processing on the first temperature loading boundary condition subset, the first pressure loading boundary condition subset, and the first locking force loading boundary condition subset to generate a first stress loading boundary condition subset corresponding to the end of the first stress transfer channel set. The first stress loading boundary condition subset includes the temperature loading boundary condition, pressure loading boundary condition, and locking force loading boundary condition defined synchronously on the time axis.

[0038] The Set_T1 obtained in step S133, Set_P1 obtained in step S134, and Set_F1 obtained in step S135 are associated and combined. The association and combination method is as follows: for each first projection region boundary Boundary_proj, its corresponding temperature loading value (from Set_T1), pressure loading value and its time period (from Set_P1), and locking force loading value and its time period (from Set_F1) are synchronized and aligned on the time axis. Specifically, the total time axis is discretized into N time steps Δt. At each time step, the boundary condition corresponding to Boundary_proj contains three components: temperature value T(t), pressure value P(t), and locking force value F(t). Among them, T(t) is a piecewise constant function on the time axis, and P(t) and F(t) are piecewise constant functions or piecewise linear functions. The set of triples (T, P, F) of all Boundary_proj at each time step constitutes the first stress loading boundary condition subset Set_BC1. The data structure of Set_BC1 is as follows: For each first projection region boundary, a time series array is stored, and each element of the array contains a timestamp, temperature value, pressure value, and locking force value.

[0039] Step S137: Perform boundary condition association and combination processing on the second temperature loading boundary condition subset, the second pressure loading boundary condition subset, and the second locking force loading boundary condition subset to generate a second stress loading boundary condition subset corresponding to the end of the second stress transmission channel set. The second stress loading boundary condition subset includes the temperature loading boundary condition, pressure loading boundary condition, and locking force loading boundary condition defined synchronously on the time axis.

[0040] The Set_T2 obtained in step S133, Set_P2 obtained in step S134, and Set_F2 obtained in step S135 are associated and combined. The combination method is the same as in step S136: for each second projection region boundary Boundary_proj_b, its corresponding temperature loading value, pressure loading value and its time period, and locking force loading value and its time period are synchronously aligned on the time axis, forming a triplet (T_b(t), P_b(t), F_b(t)) at each discrete time step. The set of triplets of all Boundary_proj_b at each time step constitutes the second stress loading boundary condition subset Set_BC2.

[0041] Step S138: Combine the first set of stress loading boundary conditions and the second set of stress loading boundary conditions into a set of stress loading boundary conditions for each spatial region on the surface of the mold cavity.

[0042] The Set_BC1 obtained in step S136 and the Set_BC2 obtained in step S137 are merged to obtain the complete set of stress loading boundary conditions, Set_BC = Set_BC1∪Set_BC2. The data structure of Set_BC is as follows: it contains the mapping relationship between multiple spatial region boundaries (i.e., the first projection region boundary and the second projection region boundary) and time series boundary conditions. Each spatial region boundary corresponds to a time series array, and each element of the array contains a timestamp, temperature value, pressure value, and locking force value.

[0043] Step S140: Input the set of stress loading boundary conditions into the mold stress field distribution simulation model for iterative calculation of stress field distribution, and output the iterative calculation results of stress field distribution. The mold stress field distribution simulation model uses the set of stress loading boundary conditions as driving boundary conditions and the main path network and branch path network of stress transmission inside the injection molded part as stress propagation constraints. The iterative calculation results of stress field distribution independently include stress field distribution data on the surface of the mold cavity, pressure field distribution data inside the mold cavity, and locking force field distribution data on the mold closed contact surface.

[0044] Input the set of stress loading boundary conditions Set_BC generated in step S130 into the mold stress field distribution simulation model for iterative calculation of stress field distribution.

[0045] Step S141: Load the first set of stress loading boundary conditions from the set of stress loading boundary conditions to the input boundary condition interface of the mold stress field distribution simulation model. The mold stress field distribution simulation model applies the first driving boundary condition to the spatial region corresponding to the end of the first stress transmission channel set on the mold cavity surface according to the temperature loading boundary condition, pressure loading boundary condition and locking force loading boundary condition synchronously defined in the first set of stress loading boundary conditions.

[0046] The mold stress field distribution simulation model reads the boundary of each first projected region, Boundary_proj, and its corresponding temperature time series T(t), pressure time series P(t), and locking force time series F(t) from Set_BC1. In the spatial discrete mesh of the simulation model, the set of mesh cells Mesh_cells_proj covered by Boundary_proj is located. For each mesh cell in Mesh_cells_proj, at each time step, a temperature load T(t) is applied as a thermal boundary condition, a pressure load P(t) as a mechanical boundary condition, and a locking force load F(t) as a constraint boundary condition. These three loads are applied simultaneously at the same time step, forming the first driving boundary condition.

[0047] Step S142: Load the second stress loading boundary condition subset from the stress loading boundary condition set to the input boundary condition interface of the mold stress field distribution simulation model. Based on the coordinated loading sequence of temperature loading value, pressure loading value and locking force loading value contained in each second stress loading boundary condition in the second stress loading boundary condition subset on the time axis, apply the second driving boundary condition on the spatial region corresponding to the end of the second stress transmission channel set on the mold cavity surface.

[0048] The mold stress field distribution simulation model reads the boundary of each second projection region, Boundary_proj_b, and its corresponding temperature time series T_b(t), pressure time series P_b(t), and locking force time series F_b(t) in Set_BC2. In the spatial discrete grid of the simulation model, the set of mesh cells covered by Boundary_proj_b, Mesh_cells_proj_b, is located. For each mesh cell in Mesh_cells_proj_b, at each time step, a temperature load T_b(t), a pressure load P_b(t), and a locking force load F_b(t) are applied to form the second driving boundary conditions.

[0049] Step S143: The main path network of stress transmission inside the injection molded part is used as the main propagation channel constraint for stress wave propagation inside the mold cavity, and the branch path network of stress transmission inside the injection molded part is used as the branch propagation channel constraint for stress wave propagation inside the mold cavity. The main propagation channel constraint limits the propagation speed priority and propagation amplitude attenuation rate of stress wave in the spatial direction corresponding to the first set of stress transmission channels. The branch propagation channel constraint limits the propagation speed priority and propagation amplitude attenuation rate of stress wave in the spatial direction corresponding to the second set of stress transmission channels.

[0050] The mold stress field distribution simulation model reads the main path network Gt generated in step S125 and the branch path network Gb generated in step S126. For each edge in Gt (corresponding to the first stress transmission channel), the stress wave propagation speed priority coefficient α_trunk_edge = 1.0 + λ_edge is defined along the edge direction, where λ_edge is the speed enhancement coefficient calculated based on the path segment length and curvature; the stress wave propagation amplitude attenuation rate β_trunk_edge = β0 is defined along the edge direction. (1+μ_edge), where μ_edge is the attenuation enhancement coefficient calculated based on the node density on the path segment. For each edge in Gb (corresponding to the second stress transmission channel), the stress wave propagation velocity priority coefficient along the direction of that edge is defined as α_branch_edge=0.6+λ_branch, where λ_branch is a coefficient calculated based on the angle between the branch path and the main path; the stress wave propagation amplitude attenuation rate along the direction of that edge is defined as β_branch_edge=β0. (1.5+μ_branch), where μ_branch is the attenuation coefficient calculated based on whether the end of the branch path is connected to a local feature region.

[0051] Step S144: Using the first driving boundary condition and the second driving boundary condition as stress wave excitation sources, and the main propagation channel constraint and the branch propagation channel constraint as stress wave propagation path restriction conditions, perform iterative calculation of the stress wave propagation process in the internal space of the mold cavity. In each iteration calculation, the stress wave starts from the spatial position where the first driving boundary condition and the second driving boundary condition are located, and propagates layer by layer into the internal space of the mold cavity along the spatial direction defined by the main propagation channel constraint and the branch propagation channel constraint.

[0052] Initialize the stress wave field on the discrete mesh inside the mold cavity. Mark the mesh elements in `Mesh_cells_proj` and `Mesh_cells_proj_b` defined in steps S141 and S142 as excitation source elements. At each time step, the excitation source elements generate initial stress waves based on the applied temperature, pressure, and locking force loads. The stress waves originate from the excitation source elements, and their propagation direction priority is determined according to the propagation speed priority coefficient defined in step S143: preferentially propagating along the main propagation channel constraint (i.e., the direction of the Gt edge), secondly along the branch propagation channel constraint (i.e., the direction of the Gb edge), and finally spreading in other directions. During propagation, the energy amplitude of the stress wave attenuates according to the attenuation rate β defined in step S143. Iterative calculations are performed layer by layer: after the propagation of the k-th layer is completed, the starting node and propagation direction of the (k+1)-th layer are determined based on the arrival time and energy amplitude of the stress waves at each mesh node in the k-th layer.

[0053] Step S145: During each iteration of the calculation, record the arrival time, propagation direction, and characteristic parameters of the stress wave on each spatial grid node inside the mold cavity. Based on the arrival time, propagation direction, and characteristic parameters of the stress wave, calculate and update the stress tensor component values ​​on each spatial grid node using the stress wave propagation model.

[0054] At each spatial grid node, the arrival time of the stress wave, t_arrival, is recorded. The stress wave propagation direction vector, D_prop, is also recorded; this vector is the unit direction vector of the line connecting the current node and the previous node along the stress wave propagation path. Stress wave characteristic parameters are also recorded, including the stress wave amplitude A_wave, stress wave frequency f_wave, and stress wave waveform type (longitudinal or transverse wave). The stress wave propagation model calculates the stress tensor components based on the elastic dynamics equations: for longitudinal waves, the normal component of the stress tensor is σ_nn = ρ. c_p^2 ε_nn, where ρ is the material density, c_p is the longitudinal wave velocity, and ε_nn is the normal strain; for transverse waves, the tangential component of the stress tensor σ_nt = ρ c_s^2 ε_nt, where c_s is the transverse wave velocity and ε_nt is the tangential strain. For a mesh node simultaneously affected by both P-waves and S-waves, the stress tensor components are the superposition of the contributions from each wave: σ_ij_total = Σ(σ_ij_p) + Σ(σ_ij_s). The updated stress tensor component values ​​are stored in the stress tensor array of each mesh node, which contains six independent components: σ_xx, σ_yy, σ_zz, σ_xy, σ_xz, and σ_yz.

[0055] Step S146: Stop the iterative calculation of the stress wave propagation process when the preset iteration termination condition is reached. The iteration termination condition is that the stress wave propagates to all spatial grid nodes inside the mold cavity and the change in the stress tensor component value on each spatial grid node is lower than the preset change threshold.

[0056] After each iteration step, two termination conditions are checked. The first termination condition is that the stress wave arrival time t_arrival for all spatial grid nodes has been recorded and is not the initial value, meaning the stress wave has covered the entire internal space of the mold cavity. The second termination condition is that for each spatial grid node, the relative change of each component of the stress tensor between two consecutive iteration steps is calculated as Δσ_ij = |σ_ij_new - σ_ij_old| / |σ_ij_old|. If Δσ_ij for all components of all nodes is less than a preset threshold δ_stop (e.g., δ_stop is 0.001), then the second termination condition is met. The iteration calculation stops when both conditions are met simultaneously.

[0057] Step S147: After stopping the iterative calculation, output the set of stress tensor component values ​​on each spatial grid node on the surface of the mold cavity as the stress field distribution data on the surface of the mold cavity, output the set of pressure scalar values ​​on each spatial grid node inside the mold cavity as the pressure field distribution data inside the mold cavity, and output the set of locking force vector component values ​​on each spatial grid node on the closed contact surface of the mold as the locking force field distribution data on the closed contact surface of the mold.

[0058] From the final iteration results, extract the set of mesh nodes Surface_nodes for the mold cavity surface. Each node contains six components of the stress tensor: σ_xx_s, σ_yy_s, σ_zz_s, σ_xy_s, σ_xz_s, and σ_yz_s. The above data of all nodes constitute the stress field distribution data Data_Stress_surface for the mold cavity surface. Extract the pressure scalar value P_node for all mesh nodes inside the mold cavity (including surface nodes and internal nodes). The pressure value of each node is calculated based on the spherical stress component of the stress tensor at that node: P_node=(σ_xx+σ_yy+σ_zz) / 3. The pressure values ​​of all nodes constitute the pressure field distribution data Data_Pressure for the mold cavity interior. Extract the set of mesh nodes Contact_nodes for the mold closed contact surface. Each node contains locking force vector components F_x, F_y, and F_z. The locking force vector is calculated based on the normal stress and tangential stress of the contact surface node: normal locking force component F_normal=σ_nn A_node, tangential locking force component F_tangential=σ_nt A_node, where A_node represents the contact area represented by the node, and the locking force vector components of all contact surface nodes constitute the locking force field distribution data of the mold closed contact surface, Data_Force_contact.

[0059] Step S150: Perform reverse mapping processing on the stress field distribution iterative calculation results to obtain the reverse mapping results of the injection molded part structure performance. The reverse mapping results of the injection molded part structure performance include the predicted data of residual stress distribution after demolding, the predicted data of warpage deformation distribution after demolding, and the predicted data of shrinkage indentation distribution after demolding.

[0060] Perform reverse mapping processing on the stress field distribution iterative calculation results output in step S140 for the injection molded part structural performance.

[0061] Step S151: The stress field distribution data of the mold cavity surface is back-projected from the spatial coordinate system of the mold cavity surface to the geometric spatial coordinate system of the injection molded part to generate the surface stress field distribution data before demolding corresponding to each spatial position point on the surface of the injection molded part. The surface stress field distribution data before demolding includes the surface stress tensor component values ​​at each spatial position point on the surface of the injection molded part.

[0062] Obtain the Data_Stress_surface output from step S147, which contains the stress tensor components of each mesh node on the mold cavity surface. Obtain the inverse matrix M_map_inv of the spatial mapping matrix M_map used in step S131. For each mesh node N_mold on the mold cavity surface, its coordinates are P_mold, and its stress tensor is σ_mold. Calculate the coordinates of the corresponding point on the injection molded part surface: P_part = M_map_inv·P_mold. Directly assign σ_mold to point P_part to obtain the pre-demolding stress tensor σ_before_demold at point P_part on the injection molded part surface. Perform the above mapping on all mesh nodes on the mold cavity surface to obtain the pre-demolding surface stress field distribution data Data_Stress_before at all spatial locations on the injection molded part surface.

[0063] Step S152: The pressure field distribution data inside the mold cavity is back-projected from the internal spatial coordinate system of the mold cavity to the geometric spatial coordinate system of the injection molded part, generating internal pressure field distribution data before demolding corresponding to each spatial location point inside the injection molded part. The internal pressure field distribution data before demolding includes the internal pressure scalar value at each spatial location point inside the injection molded part.

[0064] Obtain the Data_Pressure output from step S147, which contains the pressure scalar values ​​of each mesh node inside the mold cavity. Establish the mapping matrix M_vol_map between the injection molded part volume mesh and the mold cavity volume mesh. For each mesh node N_cavity inside the mold cavity, its coordinates are P_cavity, and its pressure value is P_val. Calculate the coordinates of the corresponding point inside the injection molded part: P_part_vol = M_vol_map_inv · P_cavity, where M_vol_map_inv is the inverse matrix of M_vol_map. Assign P_val to the point P_part_vol to obtain the pre-demolding pressure value at point P_part_vol inside the injection molded part. Perform the above mapping on all mesh nodes inside the mold cavity to obtain the pre-demolding internal pressure field distribution data Data_Pressure_before for all spatial locations inside the injection molded part.

[0065] Step S153: Project the locking force field distribution data of the mold closed contact surface from the mold closed contact surface spatial coordinate system to the injection part parting surface region corresponding to the mold closed contact surface in the injection part geometric spatial coordinate system, and generate the pre-demolding parting surface locking force distribution data corresponding to each spatial position point in the injection part parting surface region. The pre-demolding parting surface locking force distribution data includes the locking force vector component values ​​at each spatial position point in the injection part parting surface region.

[0066] Obtain the Data_Force_contact output from step S147, which contains the locking force vectors (F_x, F_y, F_z) of each mesh node on the mold closed contact surface. Obtain the spatial mapping matrix M_contact_map between the mold closed contact surface and the injection molded part parting surface. For each mesh node N_contact on the mold closed contact surface, its coordinates are P_contact, and its locking force vector is F_contact. Calculate the coordinates of the corresponding point on the injection molded part parting surface: P_part_line = M_contact_map_inv·P_contact. Assign F_contact to the P_part_line point to obtain the pre-demolding locking force vector at the P_part_line point on the injection molded part parting surface. Perform the above mapping on all mesh nodes of the mold closed contact surface to obtain the pre-demolding parting surface locking force distribution data Data_Force_before for all spatial locations in the injection molded part parting surface region.

[0067] Step S154: Construct a coupled distribution field for the injection molded part before demolding based on the surface stress field distribution data, the internal pressure field distribution data, and the parting surface locking force distribution data before demolding. Each spatial location point in the coupled distribution field simultaneously contains the surface stress tensor component value, the internal pressure scalar value, and the parting surface locking force vector component value.

[0068] The Data_Stress_before obtained in step S151, the Data_Pressure_before obtained in step S152, and the Data_Force_before obtained in step S153 are fused in the geometric space coordinate system of the injection molded part. For any spatial location point P_part in the geometric space of the injection molded part, if the point exists simultaneously within the domains of the surface stress field, the internal pressure field, and the parting surface locking force field, a coupled data tuple Coupled(P_part) = (σ_surface(P_part), P_pressure(P_part), F_line(P_part)). If the point exists only in part of the field, the missing component is set to zero. The coupled data tuples of all spatial location points constitute the coupled distribution field Field_Coupled before demolding of the injection molded part.

[0069] Step S155: Perform demolding boundary release processing on the coupled distribution field, remove the locking force vector component values ​​at the spatial location points of the parting surface area of ​​the injection molded part, and generate a transient redistribution sequence of the three-dimensional stress-pressure field of the injection molded part during demolding. The transient redistribution sequence of the three-dimensional stress-pressure field of the injection molded part during demolding includes the stress-pressure field distribution state at the beginning of demolding, the stress-pressure field distribution state at the middle of demolding, and the stress-pressure field distribution state at the end of demolding.

[0070] Perform demolding boundary release processing on Field_Coupled. Define the demolding process time interval [0, T_eject], where 0 is the demolding start time and T_eject is the demolding end time. At the demolding start time t=0, keep the locking force vector component F_line in Field_Coupled unchanged. As the demolding process progresses, the locking force vector component gradually decreases. Define the locking force release function f_release(t)=1-(t / T_eject)^γ, where γ is the release exponent (γ>1 indicates a slow release followed by a fast release, γ<1 indicates a fast release followed by a slow release). At the midpoint of demolding t_mid=T_eject / 2, the locking force vector component becomes F_line_mid=f_release(t_mid). F_line. At the end of demolding, t=T_eject, the locking force vector component drops to zero. For each time point, the stress-pressure field distribution is recalculated based on the released locking force: the stress field redistribution follows the elastic rebound equation, and the pressure field redistribution follows the fluid pressure release equation. The stress-pressure field distribution states at the start, middle, and end of demolding are recorded as a sequence Seq_redistribute=[State_0, State_mid, State_end].

[0071] Step S156: Based on the surface stress tensor component values ​​in the stress-pressure field distribution state at the end of demolding, extract the residual stress tensor component values ​​that have not been released at each spatial location point on the surface of the injection molded part, perform spatial distribution set processing on the residual stress tensor component values ​​at each spatial location point on the surface of the injection molded part, and generate the residual stress distribution prediction data after demolding of the injection molded part.

[0072] The surface stress tensor distribution at the state_end at the end of demolding is extracted from Seq_redistribute. For each spatial point P_surface on the surface of the injection molded part, the corresponding stress tensor in State_end is σ_residual(P_surface), which represents the residual stress that was not released after demolding. σ_residual(P_surface) contains six independent components: σ_r_xx, σ_r_yy, σ_r_zz, σ_r_xy, σ_r_xz, and σ_r_yz. The residual stress tensor components of all surface points are organized into a distribution field according to their spatial location to obtain the predicted residual stress distribution data Data_Residual_Stress after demolding of the injection molded part. This data is stored in the form of a mesh node-tensor mapping table, with each entry containing the node coordinates and the values ​​of the six stress components.

[0073] Step S157: Based on the surface stress tensor component values ​​and internal pressure scalar values ​​in the stress-pressure field distribution state at the end of demolding, calculate the stress gradient vector direction and pressure gradient vector direction at each spatial location point on the surface of the injection molded part. Use the vector difference between the stress gradient vector direction and the pressure gradient vector direction as the warping deformation driving direction vector at each spatial location point on the surface of the injection molded part. Perform spatial distribution set processing on the warping deformation driving direction vector at each spatial location point on the surface of the injection molded part to generate warping deformation distribution prediction data after demolding of the injection molded part.

[0074] Extract the surface stress tensor σ_residual(P_surface) and the internal pressure scalar P_residual(P_interior) from State_end. For each spatial point P_surface on the injection molded part surface, calculate the stress gradient along the surface normal direction at that point: select three neighboring points P1, P2, and P3 within the neighborhood of P_surface, obtain the normal stress component σ_nn at each point, and calculate the gradient vector using the finite difference method. Simultaneously, calculate the internal pressure gradient corresponding to that point: penetrate deep into the injection molded part along the surface normal direction, take the pressure values ​​at three depth points d1, d2, and d3, and calculate the pressure gradient vector. Calculate the warping deformation driving direction vector V_warp = G_stress - G_pressure, which is the vector difference between the stress gradient vector and the pressure gradient vector. Calculate V_warp for all surface points to obtain the warping deformation driving direction vector distribution field Data_Warp_Vector, which is stored in the form of a mesh node-vector mapping table. Each entry contains the node coordinates and the three directional components of the vector.

[0075] Step S158: Based on the normal stress component value in the surface stress tensor component value and the pressure gradient change rate in the internal pressure scalar value in the stress-pressure field distribution state at the end of demolding, identify the target spatial location points where the normal stress component value exceeds the normal stress threshold and the target spatial region boundary where the pressure gradient change rate exceeds the pressure gradient change rate threshold at each spatial location point on the surface of the injection molded part. Mark the intersection area of ​​the target spatial location point and the target spatial region boundary as the shrinkage indentation area, and generate the shrinkage indentation distribution prediction data after demolding of the injection molded part.

[0076] Extract the surface normal stress component σ_nn(P_surface) and the rate of change of the internal pressure gradient from State_end. For each surface point P_surface, determine whether σ_nn(P_surface) is greater than the normal stress threshold THK_sigma. Simultaneously, for the corresponding internal region (points within a distance d_check along the normal direction), calculate the rate of change of the pressure gradient. judge Is it greater than the pressure gradient change rate threshold THK_grad? Surface points satisfying σ_nn > THK_sigma are marked as candidate points (Set_candidate). The boundary of the internal region is projected onto the surface, and the resulting region boundary is marked as the candidate boundary, `Boundary_candidate`. The intersection of the regions enclosed by `Set_candidate` and `Boundary_candidate` is calculated; the surface points within this intersection region constitute the shrinkage dent region, `Set_shrink`. The set of boundary coordinates of all shrinkage dent regions is output to obtain the predicted shrinkage dent distribution data, `Data_Shrink`, after demolding of the injection molded part. This data is stored in the form of a region boundary-attribute mapping table, where each entry contains a sequence of region boundary points and a dent type identifier.

[0077] Step S210: Analyze the boundaries of the wall thickness distribution region and the wall thickness gradient transition region of the injection molded part in the geometric configuration distribution parameters of the injection molded part, and generate a spatial distribution topology map of the wall thickness of the injection molded part. The spatial distribution topology map of the wall thickness of the injection molded part includes the boundaries of the thin-walled region, the thick-walled region, and the wall thickness gradient transition region of the injection molded part.

[0078] Wall thickness distribution field data T_i is extracted from the geometric configuration distribution parameters of the injection molded part. The T_i of each grid node is compared with preset thin-wall thickness thresholds T_thin and T_thick thickness thresholds. Nodes with T_i less than T_thin are marked as thin-wall nodes, nodes with T_i greater than T_thick are marked as thick-wall nodes, and nodes with T_i between T_thin and T_thick are marked as transition nodes. A region growing algorithm is used to cluster the marked nodes: taking any unprocessed thin-wall node as a seed point, the algorithm expands to adjacent nodes, grouping adjacent thin-wall nodes into the same thin-wall region, until a non-thin-wall node is encountered at the boundary. The coordinate sequence of the boundary nodes of this thin-wall region is recorded as the thin-wall region boundary. Similarly, thick-wall nodes and transition nodes are clustered to obtain the thick-wall region boundary and the wall thickness gradient transition region boundary, respectively. The above three types of region boundaries are integrated into graph structure data, with each region as a graph node and the adjacency relationship between regions as graph edges, forming a topological graph of the spatial distribution of the injection molded part's wall thickness.

[0079] Step S220: Analyze the rib extension direction line, rib height distribution curve, and rib root radius distribution curve in the rib layout parameters of the injection molded part to generate a spatial layout topology diagram of the injection molded part rib. The spatial layout topology diagram of the injection molded part rib includes the spatial trajectory of the rib extension direction line in the geometric coordinate system of the injection molded part, the height change trajectory of the rib height distribution curve along the rib extension direction line, and the radius change trajectory of the rib root radius distribution curve along the rib extension direction line.

[0080] Extract the extension direction line Lr, height distribution curve H(s), and root fillet radius distribution curve R(s) of each rib from the rib layout parameters of the injection molded part. For each rib, the spatial point coordinate sequence of its extension direction line Lr is recorded as the spatial trajectory of the rib as a point sequence Lr_points. The values ​​of the rib height distribution curve H(s) at each point in Lr_points are recorded as height numerical sequences and associated with the corresponding points in Lr_points. The values ​​of the root fillet radius distribution curve R(s) at each point in Lr_points are recorded as radius numerical sequences and also associated with the corresponding points in Lr_points. Organize the above information of all ribs into graph structure data: each rib is a graph node, and the spatial connection relationship between ribs is a graph edge, forming a topological graph of the spatial layout of the injection molded part ribs.

[0081] Step S230: Analyze the gate opening position coordinates, gate opening shape contour, and gate opening direction vector in the gate position parameters of the injection molded part to generate a spatial distribution topology map of the injection molded part gate. The spatial distribution topology map of the injection molded part gate includes a local coordinate system of the gate with the opening position coordinates as the origin, a set of contour point coordinates of the gate opening shape contour in the local coordinate system, and the direction vector components of the gate opening direction vector in the local coordinate system.

[0082] Extract the gate opening position coordinates (Xg, Yg, Zg), the gate opening shape contour point sequence (Shape_points), and the gate opening direction vector (Vx, Vy, Vz) from the gate position parameters of the injection molded part. Construct a local coordinate system for the gate, with the opening position coordinates (Xg, Yg, Zg) as the origin and the opening direction vector (Vx, Vy, Vz) as the positive Z-axis. The X and Y axes of this local coordinate system lie in the gate opening plane. The X-axis direction is parallel to the tangent direction of the main geometric direction extension line of the injection molded part at the opening position, and the Y-axis direction is determined by the right-hand rule. Transform the gate opening shape contour point sequence from the global coordinate system to the local coordinate system to obtain the local coordinate sequence (Shape_points_local). Represent the gate opening direction vector as (0, 0, 1) in the local coordinate system. The above information constitutes the spatial distribution topology map of the injection molded part's gate.

[0083] Step S240: Perform topology overlay and fusion processing on the topology map of the wall thickness spatial distribution of the injection molded part, the topology map of the rib spatial layout of the injection molded part, and the topology map of the gate spatial distribution of the injection molded part to generate a spatial topology fusion map of the structural performance parameters of the injection molded part. The spatial topology fusion map of the structural performance parameters of the injection molded part contains wall thickness category identifier, rib layout association identifier, and gate distance field value at each spatial location point in the geometric coordinate system of the injection molded part.

[0084] The wall thickness spatial distribution topology map generated in step S210, the rib plate spatial layout topology map generated in step S220, and the gate spatial distribution topology map generated in step S230 are superimposed and fused. For any spatial point P(x, y, z) in the geometric coordinate system of the injection molded part, firstly, it is determined which wall thickness region the point is located in, and the corresponding wall thickness category identifier is assigned to the point. The wall thickness category identifier has a value of 0 for thin-walled region, 1 for transition region, and 2 for thick-walled region. Secondly, the shortest distance from the point to the nearest rib plate extension direction line is calculated. If the distance is less than the rib plate association distance threshold D_rib, the rib plate layout association identifier of the point is assigned a value of 1; otherwise, it is assigned a value of 0. Then, the Euclidean distance from the point to the gate opening position coordinates is calculated, and this distance value is assigned to the point as the gate distance field value. For all nodes of the injection molded part volume mesh, the three attribute values ​​are calculated in the above manner to form a multi-dimensional attribute field, which is the spatial topology fusion map of the injection molded part's structural performance parameters.

[0085] Step S250: Based on the wall thickness category identifier and rib layout association identifier in the spatial topology fusion diagram of the injection molded part structural performance parameters, identify the target spatial location point whose wall thickness category identifier is the boundary of the thin-walled region and whose rib layout association identifier is connected to the radius distribution curve of the rib root, and mark the target spatial location point as the set of stress transmission path starting points.

[0086] Traverse all mesh nodes in the spatial topology fusion diagram of the injection molded part's structural performance parameters generated in step S240. For each node, check if its wall thickness category identifier is 0 and if its stiffener layout association identifier is 1. For nodes that satisfy both conditions, further determine if the node is located within the region described by the stiffener root fillet radius distribution curve R(s): calculate the distance from the node to the nearest point on the stiffener root fillet radius distribution curve. If this distance is less than the fillet radius value R(s) at that point multiplied by a preset coefficient K_fillet, then the node is identified as a target spatial location point. All nodes identified as target spatial location points constitute the stress transfer path starting point set Set_SP.

[0087] Step S260: Based on the gate distance field value in the spatial topology fusion diagram of the injection molded part's structural performance parameters, select a subset of near-gate stress transmission starting points with gate distance field values ​​less than the distance field threshold and a subset of far-gate stress transmission starting points with gate distance field values ​​greater than or equal to the distance field threshold from the set of stress transmission path starting points. The subset of near-gate stress transmission starting points is used to preferentially construct the main transmission path in the internal stress transmission backbone path network of the injection molded part, and the subset of far-gate stress transmission starting points is used to secondarily construct the backbone transmission path in the internal stress transmission backbone path network of the injection molded part.

[0088] For each starting point in the set of stress transfer path starting points Set_SP generated in step S250, the corresponding gate distance field value Dist_gate is read from the topology fusion diagram of the injection molded part's structural performance parameter space. Dist_gate is compared with a preset distance field threshold D_gate_th. If Dist_gate is less than D_gate_th, the starting point is assigned to the near-gate stress transfer starting point subset Set_NG. If Dist_gate is greater than or equal to D_gate_th, the starting point is assigned to the far-gate stress transfer starting point subset Set_FG. When generating the initial main transfer path in step S121, the path extension is preferentially performed using points in Set_NG as the path starting point. After all points in Set_NG have been processed, the path extension is then performed using points in Set_FG as the path starting point.

[0089] Step S310: Analyze the spatial boundaries of each temperature control zone in the mold temperature zone control parameters and the temperature loading time curve corresponding to each temperature control zone. The temperature loading time curve includes the temperature loading start time point, the temperature loading end time point, and the temperature linear rise slope between the temperature loading start time point and the temperature loading end time point.

[0090] The spatial boundaries T_zone_boundary_w of each temperature control zone and the corresponding temperature loading time curves are extracted from the mold temperature zone control parameters. For each temperature control zone, its temperature loading time curve is defined by three parameters: the temperature loading start time point T_st_w, the temperature loading end time point T_ed_w, and the temperature linear rise slope K_temp_w. The temperature linear rise slope K_temp_w is calculated as follows: K_temp_w equals the end temperature value minus the start temperature value, divided by the end time point minus the start time point. Before the temperature loading start time point, the temperature remains at the initial temperature T_initial; between the temperature loading start time point and the end time point, the temperature changes linearly with time; after the temperature loading end time point, the temperature remains at the end temperature value T_val_end.

[0091] Step S320: Analyze the time interval of each pressure loading stage in the mold pressure segment loading parameters and the pressure loading spatial distribution pattern corresponding to each pressure loading stage. The pressure loading spatial distribution pattern includes the coordinates of the pressure loading center point, the diffusion rate parameter of the pressure loading from the center point to the edge, and the diffusion attenuation index of the pressure loading from the center point to the edge.

[0092] The time intervals [t1_k, t2_k] and corresponding spatial distribution patterns of pressure loading for each pressure loading stage are extracted from the segmented pressure loading parameters of the mold. For each pressure loading stage, its spatial distribution pattern is defined by three parameters: the coordinates of the pressure loading center point P_center_k, the pressure diffusion velocity parameter V_diffuse_k, and the pressure diffusion attenuation index Alpha_k. Within the time interval [t1_k, t2_k], the pressure value at any point P_surface on the mold cavity surface is determined as follows: first, the distance d_surface from P_surface to P_center_k is calculated, and then the time t_diffuse required for the pressure wave to propagate from P_center_k to P_surface is calculated, which is equal to d_surface divided by V_diffuse_k. If the current time t is greater than or equal to t1_k plus t_diffuse, then the pressure value at that point is P_stage_k multiplied by a power of Alpha_k minus d_surface divided by the maximum influence radius D_max; if the current time t is less than t1_k plus t_diffuse, then the pressure value at that point is zero.

[0093] Step S330: Analyze the start time and duration of each locking force loading stage in the mold locking timing control parameters, as well as the locking force loading direction vector corresponding to each locking force loading stage. The locking force loading direction vector includes the normal component of the locking force in the normal direction of the mold closed contact surface and the tangential component of the locking force in the tangential direction of the mold closed contact surface.

[0094] The start time T_m, duration D_m, and corresponding locking force loading direction vector for each locking force loading stage are extracted from the mold locking timing control parameters. For each locking force loading stage, its locking force loading direction vector contains two components: the normal component F_normal_m and the tangential component F_tangential_m. The formula for calculating the modulus of the locking force loading direction vector is |F_m|=sqrt(F_normal_m^2+F_tangential_m^2). Within the time window [T_m, T_m+D_m], the relationship between the locking force loading value F_m(t) and time is as follows: when t belongs to [T_m, T_m+D_m / 2], F_m(t)=2(t-T_m)F_m / D_m; when t belongs to [T_m+D_m / 2, T_m+D_m], F_m(t)=2(T_m+D_m-t)F_m / D_m.

[0095] Step S340: Align the temperature loading time curve, the pressure loading spatial distribution pattern, and the locking force loading direction vector along the time axis to generate a joint distribution field. The joint distribution field at each time point includes the correspondence between the spatial boundary of the temperature control area and the temperature loading value, the correspondence between the coordinates of the pressure loading center point and the pressure diffusion velocity parameter and the pressure diffusion attenuation index, and the values ​​of the normal and tangential components of the locking force loading direction vector.

[0096] Align the temperature loading time curve obtained in step S310, the pressure loading spatial distribution pattern obtained in step S320, and the locking force loading direction vector obtained in step S330 on the same time axis. Define a total time axis [0, T_total], and discretize the time axis with a time step Δt_discrete to obtain the time point sequence t0, t1, ..., tN. At each discrete time point tk, record the following information: for each temperature control region, record the temperature loading value T_val_w(tk) at that time point; for each pressure loading stage, record the coordinates of the pressure loading center point P_center_k, the diffusion velocity parameter V_diffuse_k, and the diffusion decay index Alpha_k; for each locking force loading stage, record the normal component F_normal_m(tk) and the tangential component F_tangential_m(tk) of the locking force loading direction vector. Organize the above information at all time points into a joint distribution field Field_Joint.

[0097] Step S350: Input the joint distribution field into the mold response characteristic prediction model to perform mold cavity surface response characteristic prediction calculation. The mold response characteristic prediction model takes the joint distribution field as input excitation, and uses the thermal conductivity coefficient matrix, elastic modulus tensor and Poisson's ratio distribution field of the mold material as mold intrinsic property constraints. It outputs the thermal expansion displacement, elastic deformation displacement and viscoelastic creep at different time points of each spatial position on the mold cavity surface.

[0098] The joint distribution field Field_Joint generated in step S340 is input into the mold response characteristic prediction model, which adopts a finite element analysis framework. The thermal conductivity coefficient matrix K_thermal of the mold material is a 3x3 symmetric positive definite matrix, the elastic modulus tensor E_elastic is a 6x6 symmetric matrix, and the Poisson's ratio distribution field ν_field is stored in the form of a scalar field on each grid node. For each time point tk and each spatial location point P_surface on the mold cavity surface, the model first calculates the thermal expansion displacement based on the temperature loading value in the joint distribution field. The thermal expansion displacement is equal to the thermal expansion coefficient multiplied by the temperature loading value and then multiplied by the original size. Secondly, the elastic deformation displacement is calculated based on the pressure loading value. The elastic deformation displacement is equal to the pressure value multiplied by the force area divided by the elastic modulus and then divided by the force area and then multiplied by the original size. Then, the viscoelastic creep is calculated based on the locking force loading value and the viscoelastic constitutive model. The viscoelastic creep is equal to the integral of the locking force over time divided by the viscosity coefficient. The model outputs three displacement fields: thermal expansion displacement field, elastic deformation displacement field, and viscoelastic creep displacement field.

[0099] Step S360: Based on the thermal expansion displacement, elastic deformation displacement, and viscoelastic creep of each spatial location point on the surface of the mold cavity at different time points, calculate the total displacement of each spatial location point on the surface of the mold cavity at different time points, and mark the target spatial location point at the target time point that exceeds the elastic limit displacement threshold of the mold material as the mold yield initiation point.

[0100] For each spatial location point P_surface on the mold cavity surface and each time point tk, calculate the total displacement D_total(tk) = |D_thermal(tk)| + |D_elastic(tk)| + |D_visco(tk)|. Obtain the elastic limit displacement threshold D_elastic_limit of the mold material, which is calculated from the material's yield strength σ_yield and elastic modulus E: D_elastic_limit = σ_yield L0 / E. If D_total(tk) is greater than D_elastic_limit, then mark the spatial location point P_surface at time point tk as the mold yield start point. Record the coordinates and time information of all marked yield start points as the mold yield start point set Set_YP.

[0101] Step S370: The set of coordinates of the mold yield initiation point in the spatial coordinate system of the mold cavity surface is used as the spatial distribution map of the mold yield initiation point. The spatial distribution map of the mold yield initiation point is spatially superimposed with the stress transmission backbone path network inside the injection molded part to identify the first mold yield initiation point subset located in the spatial region corresponding to the end of the first stress transmission channel set and the second mold yield initiation point subset located in the spatial region corresponding to the end of the second stress transmission channel set.

[0102] The coordinates of the mold yield initiation point set Set_YP generated in step S360 are extracted to form a spatial distribution map of mold yield initiation points Map_YP. Map_YP is then spatially overlaid with the backbone path network Gt generated in step S125. For each first stress transfer channel in Gt, the boundary of the first projected region Boundary_proj corresponding to the end point of that channel is obtained. It is determined whether each yield initiation point in Map_YP is located within the region enclosed by Boundary_proj. If it is located within the region, the yield initiation point is assigned to the first mold yield initiation point subset Set_YP1. Similarly, for each second stress transfer channel in the branch path network Gb, the boundary of the second projected region Boundary_proj_b corresponding to the end point of that channel is obtained. It is determined whether the yield initiation point in Map_YP is located within the region enclosed by Boundary_proj_b. If it is located within the region, the yield initiation point is assigned to the second mold yield initiation point subset Set_YP2.

[0103] Step S380: Based on the distribution density of the first mold yield initiation point subset in the corresponding spatial region at the end of the first stress transmission channel set, adjust the loading order priority of the temperature loading value, pressure loading value and locking force loading value of the first stress loading boundary condition in the first stress loading boundary condition subset on the time axis.

[0104] The distribution density ρ_yield1 of the first mold yield initiation point subset Set_YP1 within the first projected region boundary Boundary_proj is calculated as the number of yield initiation points in Set_YP1 located within Boundary_proj divided by the area enclosed by Boundary_proj. ρ_yield1 is compared with the density threshold ρ_th. If ρ_yield1 is greater than ρ_th, it indicates that the region is prone to yielding, and the load loading priority for this region needs to be reduced. For the first stress loading boundary condition corresponding to Boundary_proj, the rising slope of its temperature loading value is reduced by a preset ratio δT, the peak occurrence time of its pressure loading value is delayed by a preset time ΔT_p, and the start time of its locking force loading value is delayed by a preset time ΔT_f.

[0105] Step S390: Based on the distribution density of the second mold yield initiation point subset in the corresponding spatial region at the end of the second stress transmission channel set, adjust the loading order priority of the temperature loading value, pressure loading value and locking force loading value of the second stress loading boundary condition in the second stress loading boundary condition subset on the time axis.

[0106] The distribution density ρ_yield2 of the second mold yield initiation point subset Set_YP2 within the boundary Boundary_proj_b of the second projected region is calculated as the number of yield initiation points in Set_YP2 located within that Boundary_proj_b divided by the area enclosed by Boundary_proj_b. ρ_yield2 is compared to the density threshold ρ_th. If ρ_yield2 is greater than ρ_th, adjustments are made to the second stress loading boundary conditions corresponding to that Boundary_proj_b: the rising slope of the temperature loading value is reduced, the peak occurrence time of the pressure loading value is delayed, and the initiation time of the locking force loading value is delayed. The adjustment magnitude is proportional to ρ_yield2; the larger ρ_yield2 is, the larger the adjustment magnitude.

[0107] Step S410: Extract the stress wave propagation path trajectory on each spatial grid node inside the mold cavity recorded by the simulation model of the mold stress field distribution during the iterative calculation of the stress field distribution. The stress wave propagation path trajectory includes the node number sequence of each spatial grid node that the stress wave passes through in sequence from the excitation source position.

[0108] During the iterative calculation in step S140, the mold stress field distribution simulation model records the spatial grid node numbers that each stress wave passes through sequentially after starting from the excitation source. For each excitation source, a node number sequence Seq_nodes=[N_start, N_2, N_3, ..., N_end] is generated, where N_start is the node number where the excitation source is located, and N_end is the node number where the stress wave terminates. The node number sequences corresponding to all excitation sources constitute the stress wave propagation path trajectory set Set_Paths.

[0109] Step S420: Based on the node number sequence in the stress wave propagation path trajectory, identify the target spatial grid node in which the node number sequence in the stress wave propagation path trajectory repeats cyclically, and mark the target spatial grid node as a stress wave standing wave node.

[0110] Iterate through each node number sequence in the set of stress wave propagation path trajectories, Set_Paths, generated in step S410. For each sequence, check if any node number appears repeatedly. If a node number appears twice or more in the sequence, mark that node as a stress wave standing wave node. Extract the coordinate information of all marked standing wave nodes to obtain the set of stress wave standing wave nodes, Set_StandingNodes.

[0111] Step S430: Extract the coordinate set of all target spatial mesh nodes marked as stress wave standing wave nodes in the spatial coordinate system inside the mold cavity, and generate a spatial distribution map of stress wave standing wave nodes inside the mold cavity.

[0112] Extract the node coordinates from the stress wave standing wave node set Set_StandingNodes obtained in step S420 to form a spatial distribution map of stress wave standing wave nodes inside the mold cavity, Map_Standing, which is stored in the form of a point cloud, with each point containing a node number and three-dimensional spatial coordinates.

[0113] Step S440: Perform spatial overlay analysis on the spatial distribution map of stress wave standing wave nodes inside the mold cavity and the stress transmission backbone path network inside the injection molded part to identify stress wave standing wave nodes located on the backbone transmission path in the stress transmission backbone path network inside the injection molded part, and mark the stress wave standing wave nodes located on the backbone transmission path as stress wave propagation blocking nodes.

[0114] Spatial overlay analysis is performed between the Map_Standing generated in step S430 and the backbone path network Gt generated in step S125. For each backbone propagation path in Gt (i.e., each edge of Gt), the spatial point sequence Path_points_trunk of that path is obtained. It is determined whether each standing wave node in Map_Standing is located on Path_points_trunk or whether the shortest distance to Path_points_trunk is less than the distance threshold D_match. If it is located, the standing wave node is marked as a stress wave propagation blocking node and added to the set Set_BlockNodes.

[0115] Step S450: Perform spatial overlay analysis on the spatial distribution map of stress wave standing wave nodes inside the mold cavity and the stress transmission branch path network inside the injection molded part to identify stress wave standing wave nodes located on the branch transmission path in the stress transmission branch path network inside the injection molded part, and mark the stress wave standing wave nodes located on the branch transmission path as stress wave energy dissipation nodes.

[0116] Spatial overlay analysis is performed between the Map_Standing generated in step S430 and the branch path network Gb generated in step S126. For each branch propagation path in Gb (i.e., each edge of Gb), the spatial point sequence Path_points_branch of that path is obtained. It is determined whether each standing wave node in Map_Standing is located on Path_points_branch or whether the shortest distance to Path_points_branch is less than the distance threshold D_match. If it is located, the standing wave node is marked as a stress wave energy dissipation node and added to the set Set_DissipNodes.

[0117] Step S460: Based on the spatial distribution position of the stress wave propagation blocking node in the stress transmission backbone path network inside the injection molded part, modify the stress wave propagation speed priority parameter of the backbone propagation channel constraint in the mold stress field distribution simulation model, reduce the stress wave propagation speed priority on the backbone transmission path where the stress wave propagation blocking node is located, and increase the stress wave propagation speed priority on the detour backbone transmission path that bypasses the stress wave propagation blocking node.

[0118] For each blocked node in the set of stress wave propagation blocking nodes Set_BlockNodes obtained in step S440, locate the trunk propagation path Path_trunk_blocked where the node is located. Reduce the stress wave propagation speed priority coefficient α_trunk_edge on this path by a preset ratio β_block. At the same time, search for a detour trunk propagation path Path_trunk_detour (i.e., an alternative path that does not pass through the blocked node) from upstream to downstream, and increase the stress wave propagation speed priority coefficient α_trunk_edge on this detour path by a preset ratio γ_detour. The modified speed priority coefficient is re-inputted into the trunk propagation channel constraint in step S143.

[0119] Step S470: Based on the spatial distribution position of the stress wave energy dissipation node in the stress transmission branch path network inside the injection molded part, modify the stress wave propagation amplitude attenuation rate parameter of the branch propagation channel constraint in the mold stress field distribution simulation model, increase the stress wave propagation amplitude attenuation rate on the branch transmission path where the stress wave energy dissipation node is located, and decrease the stress wave propagation amplitude attenuation rate on the branch transmission path where the non-stress wave energy dissipation node is located.

[0120] For each dissipation node in the set of stress wave energy dissipation nodes Set_DissipNodes obtained in step S450, the branch propagation path Path_branch_dissipated where the node is located is located. The stress wave propagation amplitude attenuation rate β_branch_edge on this path is increased by a preset ratio δ_dissip to promote the dissipation of stress wave energy on this path. For branch propagation paths where there are no stress wave energy dissipation nodes, the amplitude attenuation rate β_branch_edge is decreased by a preset ratio ε_preserve to maintain the propagation of stress wave energy along this path. The modified amplitude attenuation rate parameter is re-inputted into the branch propagation channel constraints in step S143.

[0121] Step S510: The residual stress tensor component values ​​at various spatial locations on the surface of the injection molded part in the residual stress distribution prediction data after demolding are classified. Based on the spatial angle between the direction of the maximum principal stress in the residual stress tensor component values ​​and the extension line of the main direction of the geometric configuration of the injection molded part, the spatial locations on the surface of the injection molded part are divided into a first set of residual stress regions where the residual stress is distributed along the main direction of the geometric configuration and a second set of residual stress regions where the residual stress is distributed along the secondary direction of the geometric configuration.

[0122] Extract the residual stress tensor σ_residual from each surface point in the residual stress distribution prediction data Data_Residual_Stress generated after demolding of the injection molded part in step S156. Calculate the direction of the maximum principal stress at each point, i.e., solve for the eigenvalues ​​and eigenvectors of the stress tensor, and take the direction of the eigenvector corresponding to the maximum eigenvalue as the direction of the maximum principal stress D_maxσ. Calculate the angle θ_main between D_maxσ and the extension direction D_main of the principal direction of the injection molded part's geometry. If θ_main is less than 45 degrees, the point is assigned to the first residual stress region set Set_R1; if θ_main is greater than or equal to 45 degrees, the point is assigned to the second residual stress region set Set_R2.

[0123] Step S520: Perform clustering processing on the warping deformation driving direction vectors at various spatial locations on the surface of the injection molded part in the warping deformation distribution prediction data after demolding. Based on the size of the vector angle between the warping deformation driving direction vectors, divide the spatial locations on the surface of the injection molded part into a third set of warping deformation regions where the warping deformation driving direction tends to be consistent and a fourth set of warping deformation regions where the warping deformation driving direction is divergent.

[0124] The warp deformation driving direction vector V_warp is extracted from the warp deformation distribution prediction data Data_Warp_Vector of the injection molded part after demolding generated in step S157. A density-based clustering algorithm is used to process the above direction vector: for any two points i and j, the angle θ_warp_ij between their direction vectors is calculated as follows: θ_warp_ij = arccos((V_warp_i·V_warp_j) / (|V_warp_i| If θ_warp_ij is less than the clustering threshold θ_cluster, the two points are grouped into the same cluster. For each cluster, the average direction of all direction vectors within the cluster is calculated. If the number of points in the cluster exceeds the preset number threshold N_cluster and the standard deviation of the direction vectors within the cluster is less than the standard deviation threshold σ_cluster, the points in the cluster are grouped into the third warped region set Set_W3; otherwise, the points in the cluster are grouped into the fourth warped region set Set_W4.

[0125] Step S530: Perform depth classification processing on the spatial boundary of the shrinkage indentation area in the predicted data of shrinkage indentation distribution after demolding of the injection molded part. Based on the difference between the normal stress component value at the spatial location point on the surface of the injection molded part within the shrinkage indentation area and the normal stress component value at the spatial location point on the surface of the injection molded part in the adjacent non-shrinkage indentation area, divide the shrinkage indentation area into a set of deep shrinkage indentation areas and a set of shallow shrinkage indentation areas.

[0126] Extract the spatial boundaries of each shrinkage indentation region from the predicted shrinkage indentation distribution data Data_Shrink generated in step S158 after demolding of the injection molded part. For each shrinkage indentation region, calculate the average value σ_avg_inside of the normal stress component σ_nn of all surface points within the region, and the average value σ_avg_outside of the normal stress component σ_nn of the surface points in the adjacent non-shrinkage indentation region. Calculate the stress difference Δσ_shrink = σ_avg_inside - σ_avg_outside. If Δσ_shrink is greater than the deep indentation stress difference threshold Δσ_deep, then the region is classified into the deep shrinkage indentation region set Set_SH_deep; otherwise, the region is classified into the shallow shrinkage indentation region set Set_SH_shallow.

[0127] Step S540: Perform spatial overlay analysis on the first set of residual stress regions, the second set of residual stress regions, the third set of warped deformation regions, the fourth set of warped deformation regions, the deep shrinkage dent region set, and the shallow shrinkage dent region set to generate a spatial distribution map of the surface quality risk regions of the injection molded part. The spatial distribution map of the surface quality risk regions of the injection molded part includes the boundary of the first risk region that belongs to both the first set of residual stress regions and the third set of warped deformation regions, the boundary of the second risk region that belongs to both the second set of residual stress regions and the fourth set of warped deformation regions, and the boundary of the third risk region that belongs to both the deep shrinkage dent region set and the first set of residual stress regions.

[0128] Spatial overlay analysis is performed on Set_R1 and Set_R2 obtained in step S510, Set_W3 and Set_W4 obtained in step S520, and Set_SH_deep and Set_SH_shallow obtained in step S530. The intersection region of Set_R1 and Set_W3 is calculated, and its boundary is marked as the first risk region boundary, Boundary_Risk1. The intersection region of Set_R2 and Set_W4 is calculated, and its boundary is marked as the second risk region boundary, Boundary_Risk2. The intersection region of Set_SH_deep and Set_R1 is calculated, and its boundary is marked as the third risk region boundary, Boundary_Risk3. These three risk region boundaries are integrated into a spatial distribution map of the injection molded part surface quality risk region, Map_Risk.

[0129] Step S550: Based on the first risk area boundary, the second risk area boundary, and the third risk area boundary in the spatial distribution map of the surface quality risk areas of the injection molded part, generate a set of injection mold correction parameters. The set of injection mold correction parameters includes correction suggestions for mold temperature zoning control parameters for the first risk area boundary, correction suggestions for mold pressure segmented loading parameters for the second risk area boundary, and correction suggestions for mold locking timing control parameters for the third risk area boundary.

[0130] For the first risk region boundary (Boundary_Risk1), the following correction suggestions are made for the mold temperature zoning control parameters: increase the temperature loading value of the corresponding temperature control zone, increase the linear temperature rise slope, and shorten the time interval from the start to the end of the temperature loading. For the second risk region boundary (Boundary_Risk2), the following correction suggestions are made for the mold pressure segmented loading parameters: decrease the pressure loading value of the corresponding pressure loading stage, increase the pressure diffusion attenuation index, and decrease the pressure diffusion rate parameter. For the third risk region boundary (Boundary_Risk3), the following correction suggestions are made for the mold locking timing control parameters: decrease the locking force loading value of the corresponding locking force loading stage, increase the duration of the locking force loading stage, and adjust the ratio of the normal to tangential components of the locking force loading direction vector to make it more tangential.

[0131] For example, after step S150, the method further includes: step S610: inputting the predicted data of residual stress distribution after demolding of the injection molded part into the structural strength verification model of the injection molded part for structural strength verification calculation. The structural strength verification model of the injection molded part uses the residual stress tensor component values ​​in the predicted data of residual stress distribution after demolding of the injection molded part as the input load, and uses the yield strength tensor threshold and fracture strength tensor threshold of the injection molded part material as the verification benchmark to output the structural strength safety factor distribution map at each spatial location point on the surface of the injection molded part.

[0132] The residual stress distribution prediction data Data_Residual_Stress generated in step S156 after demolding of the injection molded part is input into the structural strength verification model of the injection molded part. It reads the residual stress tensor σ_residual of each surface point and calculates the equivalent stress σ_vonMises=sqrt(((σ_xx-σ_yy)^2+(σ_yy-σ_zz)^2+(σ_zz-σ_xx)^2+6) ((σ_xy^2 + σ_xz^2 + σ_yz^2)) / 2). Obtain the yield strength tensor threshold σ_yield_tensor and the fracture strength tensor threshold σ_fracture_tensor of the injection-molded part material. For each surface point, calculate the yield safety factor SF_yield = σ_yield_tensor / σ_vonMises, calculate the fracture safety factor SF_fracture = σ_fracture_tensor / σ_vonMises, and take the smaller value of the two as the structural strength safety factor of this point. The safety factors of all points constitute the structural strength safety factor distribution map Map_SF.

[0133] Step S620: Input the predicted data of the warpage deformation distribution after demolding the injection-molded part into the injection-molded part assembly accuracy verification model for assembly accuracy verification calculation. The injection-molded part assembly accuracy verification model takes the warpage deformation driving direction vector in the predicted data of the warpage deformation distribution after demolding the injection-molded part as the input deformation amount, and takes the tolerance zone boundary of the injection-molded part assembly mating surface as the verification reference, and outputs the assembly accuracy out-of-tolerance probability distribution map at each spatial position point on the surface of the injection-molded part.

[0134] Input the predicted data of the warpage deformation distribution after demolding the injection-molded part Data_Warp_Vector generated in step S157 into the injection-molded part assembly accuracy verification model, which reads the warpage deformation driving direction vector V_warp of each surface point and calculates the expected warpage deformation displacement D_warp = |V_warp| T_warp, where T_warp is the warpage deformation duration. Obtain the tolerance zone boundary Boundary_tolerance of the injection-molded part assembly mating surface, and this tolerance zone boundary defines the maximum allowable deformation displacement D_tolerance_max and the minimum deformation displacement D_tolerance_min. For each surface point, calculate the probability that the warpage deformation displacement exceeds the tolerance zone boundary: If D_warp > D_tolerance_max, then P_out_tol = ((D_warp - D_tolerance_max) / D_tolerance_max)^2 K_prob; If D_warp < D_tolerance_min, then P_out_tol = ((D_tolerance_min - D_warp) / D_tolerance_min)^2 K_prob; Otherwise, P_out_tol = 0. The out-of-tolerance probabilities of all points constitute the assembly accuracy out-of-tolerance probability distribution map Map_Prob.

[0135] Step S630: Input the shrinkage dent distribution prediction data after demolding the injection molded part into the appearance quality checking model of the injection molded part for appearance quality checking calculation. The appearance quality checking model of the injection molded part uses the spatial region boundary of the shrinkage dent occurrence area in the shrinkage dent distribution prediction data after demolding the injection molded part as the input defect area, and uses the surface roughness threshold and glossiness threshold of the injection molded part appearance as the checking benchmark, and outputs the boundary map of the appearance quality unqualified area at each spatial position point on the surface of the injection molded part.

[0136] Input the shrinkage dent distribution prediction data Data_Shrink after demolding the injection molded part generated in Step S158 into the appearance quality checking model of the injection molded part. It reads the spatial region boundaries of each shrinkage dent occurrence area. For each shrinkage dent area, calculate the average surface roughness Ra_avg and the average glossiness Gu_avg within this area. Obtain the surface roughness threshold Ra_limit and the glossiness threshold Gu_limit of the injection molded part appearance. If Ra_avg > Ra_limit or Gu_avg < Gu_limit, then mark this shrinkage dent area as an appearance quality unqualified area. The boundaries of all unqualified areas form the boundary map of the appearance quality unqualified area Map_Defect.

[0137] Step S640: Mark the spatial position points with a structural strength safety factor lower than the safety factor threshold in the structural strength safety factor distribution map as structural strength danger points, mark the spatial position points with an assembly accuracy out-of-tolerance probability exceeding the out-of-tolerance probability threshold in the assembly accuracy out-of-tolerance probability distribution map as assembly accuracy danger points, and mark the spatial region boundaries in the appearance quality unqualified area boundary map as appearance quality unqualified areas.

[0138] Traverse each point in the structural strength safety factor distribution map Map_SF generated in Step S610. If the safety factor SF of this point is less than the safety factor threshold SF_limit, then mark this point as a structural strength danger point and include it in the set Set_Danger_Strength. Traverse each point in the assembly accuracy out-of-tolerance probability distribution map Map_Prob generated in Step S620. If the out-of-tolerance probability P_out_tol of this point is greater than the out-of-tolerance probability threshold P_limit, then mark this point as an assembly accuracy danger point and include it in the set Set_Danger_Assembly. Mark each region boundary in the appearance quality unqualified area boundary map Map_Defect generated in Step S630 as an appearance quality unqualified area.

[0139] Step S650: Perform spatial location correlation analysis on the spatial coordinate set of the structural strength hazard points, the spatial coordinate set of the assembly accuracy hazard points, and the spatial region boundary of the appearance quality defective area to identify the first comprehensive risk region boundary that simultaneously contains structural strength hazard points and assembly accuracy hazard points, the second comprehensive risk region boundary that simultaneously contains assembly accuracy hazard points and appearance quality defective area, and the third comprehensive risk region boundary that simultaneously contains structural strength hazard points and appearance quality defective area.

[0140] Perform spatial location correlation analysis on the Set_Danger_Strength, Set_Danger_Assembly, and Map_Defect obtained in step S640. Calculate the intersection region of Set_Danger_Strength and Set_Danger_Assembly, and mark the boundary of this region as the first comprehensive risk region boundary Boundary_Risk_SA. Calculate the intersection region of Set_Danger_Assembly and Map_Defect (i.e., the part where assembly accuracy hazards are located within the appearance quality non-conforming area), and mark the boundary of this region as the second comprehensive risk region boundary Boundary_Risk_AD. Calculate the intersection region of Set_Danger_Strength and Map_Defect (i.e., the part where structural strength hazards are located within the appearance quality non-conforming area), and mark the boundary of this region as the third comprehensive risk region boundary Boundary_Risk_SD.

[0141] Step S660: Generate a set of injection molding process parameter optimization paths based on the boundaries of the first comprehensive risk area, the second comprehensive risk area, and the third comprehensive risk area. The set of injection molding process parameter optimization paths includes an optimization path for mold temperature zoning control parameters for the first comprehensive risk area boundary, an optimization path for mold pressure segmented loading parameters for the second comprehensive risk area boundary, and an optimization path for mold locking timing control parameters for the third comprehensive risk area boundary.

[0142] For the first comprehensive risk area boundary (Boundary_Risk_SA), an optimization path for mold temperature zone control parameters is generated: The temperature loading value of the corresponding temperature control zone is gradually increased by ΔT_step per step, while the linear temperature rise slope is gradually increased by ΔK_step per step. After each optimization step, steps S610 to S640 are re-executed until the structural strength safety factor and assembly accuracy deviation probability of the zone meet the requirements. For the second comprehensive risk area boundary (Boundary_Risk_AD), an optimization path for mold pressure segmented loading parameters is generated: The pressure loading value of the corresponding pressure loading stage is gradually decreased by ΔP_step per step, while the pressure diffusion attenuation index is gradually increased by ΔAlpha_step per step. After each optimization step, steps S620 to S640 are re-executed until the assembly accuracy deviation probability and appearance quality of the zone meet the requirements. For the boundary of the third comprehensive risk area, Boundary_Risk_SD, the optimization path for the mold locking timing control parameters is generated as follows: gradually decrease the locking force loading value of the corresponding locking force loading stage in this area, with a reduction of ΔF_step in each step, while gradually increasing the duration of the locking force loading stage, with an increase of ΔD_step in each step. After each optimization, steps S610, S630, and S640 are re-executed until the structural strength safety factor and appearance quality of this area meet the requirements.

[0143] For example, before step S140, the method further includes: step S710: calling the stress field distribution prediction convolutional network to process the set of stress loading boundary conditions. The stress field distribution prediction convolutional network includes a boundary condition input port and a path constraint input port. The boundary condition input port receives the first set of stress loading boundary conditions and the second set of stress loading boundary conditions. The path constraint input port receives the main path network of stress transmission inside the injection molded part and the branch path network of stress transmission inside the injection molded part.

[0144] Before inputting the set of stress loading boundary conditions into the mold stress field distribution simulation model for iterative calculation in step S140, the stress field distribution prediction convolutional network is first invoked to preprocess the above data. The structure of this convolutional network includes two independent input ports: the boundary condition input port is used to receive the first stress loading boundary condition subset Set_BC1 and the second stress loading boundary condition subset Set_BC2 generated in step S130; the path constraint input port is used to receive the trunk path network Gt generated in step S125 and the branch path network Gb generated in step S126.

[0145] Step S720: The stress field distribution prediction convolutional network converts the first stress loading boundary condition subset into a first stress loading condition tensor, which includes a temperature loading time series channel, a pressure loading time series channel, and a locking force loading time series channel.

[0146] The stress field distribution prediction convolutional network performs tensor transformation operations on Set_BC1. For the temperature time series T(t) corresponding to the boundary of each first projection region in Set_BC1, it is discretized into a time series vector of length Nt, which serves as the feature value of the temperature loading time series channel. Similarly, the pressure time series P(t) is discretized into a pressure loading time series channel, and the locking force time series F(t) is discretized into a locking force loading time series channel. The above three channels are stacked along the channel dimension to form the first stress loading condition tensor Ten_1 with the shape (H, W, Nt, 3), where H and W are the height and width dimensions of the projection region boundary on the spatial grid, Nt is the number of time steps, and 3 is the number of channels.

[0147] Step S730: The stress field distribution prediction convolutional network converts the second stress loading boundary condition subset into a second stress loading condition tensor, which includes a temperature loading time series channel, a pressure loading time series channel, and a locking force loading time series channel.

[0148] Perform the same tensor transformation operation as in step S720 on Set_BC2. For the temperature time series T_b(t), pressure time series P_b(t), and locking force time series F_b(t) corresponding to each second projection region boundary in Set_BC2, discretize them into time series vectors, stack them along the channel dimension to form a second stress loading condition tensor Ten_2 with shape (H_b, W_b, Nt, 3), where H_b and W_b are the height and width dimensions of the second projection region boundary on the spatial grid.

[0149] Step S740: The stress field distribution prediction convolutional network converts the stress transmission backbone path network inside the injection molded part into a backbone path constraint mask tensor. The backbone path constraint mask tensor takes an activation value at the spatial position corresponding to the first stress transmission channel set, and a suppression value at the spatial position not corresponding to the first stress transmission channel set.

[0150] The stress field distribution prediction convolutional network performs a mask tensor transformation operation on the backbone path network Gt. An initial all-zero tensor Mask_t of shape (H_mesh, W_mesh, 1) is created. The first stress transfer channel corresponding to each edge in Gt is traversed, and the spatial point sequence of that channel is obtained. This spatial point sequence is projected onto the spatial mesh of the mold cavity surface, and the value of the mesh cell corresponding to the projection position in Mask_t is set to the activation value of 1. All mesh cells corresponding to non-first stress transfer channels are kept with a suppression value of 0.

[0151] Step S750: The stress field distribution prediction convolutional network converts the stress transmission branch path network inside the injection molded part into a branch path constraint mask tensor. The branch path constraint mask tensor takes an activation value at the spatial position corresponding to the second stress transmission channel set, and a suppression value at the spatial position not corresponding to the second stress transmission channel set.

[0152] The stress field distribution prediction convolutional network performs a mask tensor transformation operation on the branch path network Gb. An initial all-zero tensor Mask_b with shape (H_mesh, W_mesh, 1) is created. The second stress transfer channel corresponding to each edge in Gb is traversed, and the spatial point sequence of that channel is obtained. This spatial point sequence is projected onto the spatial mesh of the mold cavity surface, and the value of the mesh cell corresponding to the projection position in Mask_b is set to the activation value of 1. All mesh cells corresponding to non-second stress transfer channels are kept with a suppression value of 0.

[0153] Step S760: The stress field distribution prediction convolutional network processes the first stress loading condition tensor, the second stress loading condition tensor, the trunk path constraint mask tensor, and the branch path constraint mask tensor through a multi-layer convolutional coding structure. Each convolutional layer in the multi-layer convolutional coding structure is followed by a batch normalization layer and an activation layer. The multi-layer convolutional coding structure outputs the stress field latent feature tensor.

[0154] The Ten_1 obtained in step S720, Ten_2 obtained in step S730, Mask_t obtained in step S740, and Mask_b obtained in step S750 are input into a multi-layer convolutional coding structure. This coding structure contains 5 convolutional modules, each of which contains a convolutional layer, a batch normalization layer, and a ReLU activation layer. The first convolutional module has 3 input channels, 64 output channels, a 3x3 kernel size, a stride of 1, and padding of 1. Ten_1 and Ten_2 are processed by the first convolutional module and then multiplied element-wise with Mask_t and Mask_b, respectively. The second convolutional module has 64 input channels, 128 output channels, a 3x3 kernel size, a stride of 2, and padding of 1. The third convolutional module has 128 input channels, 256 output channels, a 3x3 kernel size, a stride of 2, and padding of 1. The fourth convolutional module has 256 input channels, 512 output channels, a 3x3 kernel size, a stride of 2, and padding of 1. The fifth convolutional module has 512 input channels, 512 output channels, a 3x3 kernel size, a stride of 1, and padding of 1. After processing by the five convolutional modules, the output is a stress field latent feature tensor Z_s with shape (H / 8, W / 8, 512).

[0155] Step S770: The stress field distribution prediction convolutional network processes the stress field latent feature tensor through a deconvolution decoding structure. Each deconvolution layer in the deconvolution decoding structure is followed by a batch normalization layer and an activation layer. The deconvolution decoding structure outputs spatial distribution prediction data of the stress field on the mold cavity surface and temporal evolution prediction data of the stress field on the mold cavity surface.

[0156] The Z_s output from step S760 is input into the deconvolution decoding structure. This decoding structure contains four deconvolution modules, each of which sequentially contains a deconvolution layer, a batch normalization layer, and a ReLU activation layer. The first deconvolution module has 512 input channels, 256 output channels, a 3x3 kernel size, a stride of 2, padding of 1, and an output shape of (H / 4, W / 4, 256). The second deconvolution module has 256 input channels, 128 output channels, a 3x3 kernel size, a stride of 2, padding of 1, and an output shape of (H / 2, W / 2, 128). The third deconvolution module has 128 input channels, 64 output channels, a 3x3 kernel size, a stride of 2, padding of 1, and an output shape of (H, W, 64). The fourth deconvolution module has 64 input channels, 2 output channels, a 3x3 kernel size, a stride of 1, a padding of 1, and an output shape of (H, W, 2). The first output channel is the spatial distribution prediction data D_ss of the stress field on the mold cavity surface, and the second output channel is the temporal evolution prediction data D_st of the stress field on the mold cavity surface.

[0157] Step S780: The spatial distribution prediction data of the stress field on the mold cavity surface and the temporal evolution prediction data of the stress field on the mold cavity surface are used as the initial stress field data input of the mold stress field distribution simulation model. The mold stress field distribution simulation model starts the stress field distribution iterative calculation process based on the initial stress field data input.

[0158] The Data_Stress_Spatial and Data_Stress_Temporal output from step S770 are used as initial conditions and input into the mold stress field distribution simulation model in step S140. Data_Stress_Spatial is used as the initial spatial stress distribution state for iterative calculation, and Data_Stress_Temporal is used as the initial reference curve for the stress evolution over time at each spatial location. The model then executes the stress wave propagation iterative calculations described in steps S144 to S146 from the aforementioned initial state, thereby accelerating the convergence process and obtaining more accurate stress field distribution results.

[0159] For example, after step S150, the method further includes: step S810: calling the post-demolding performance backpropagation network to process the post-demolding residual stress distribution prediction data, the post-demolding warpage deformation distribution prediction data, and the post-demolding shrinkage indentation distribution prediction data of the injection molded part. The post-demolding performance backpropagation network includes a residual stress input port, a warpage deformation input port, and a shrinkage indentation input port.

[0160] After obtaining the predicted data of residual stress distribution (Data_Residual_Stress), warpage distribution (Data_Warp_Vector), and shrinkage indentation distribution (Data_Shrink) after demolding of the injection molded part in step S150, the post-demolding performance backpropagation network is called to perform backpropagation analysis on the above data. It contains three independent input ports: the residual stress input port is used to receive Data_Residual_Stress, the warpage input port is used to receive Data_Warp_Vector, and the shrinkage indentation input port is used to receive Data_Shrink.

[0161] Step S820: The residual stress input port converts the predicted data of residual stress distribution after demolding of the injection molded part into a residual stress distribution feature map, which contains independent channels for each component of the residual stress tensor.

[0162] After demolding, the residual stress input port of the performance backpropagation network performs a feature map transformation operation on Data_Residual_Stress. Each surface point in Data_Residual_Stress contains six residual stress tensor components: σ_r_xx, σ_r_yy, σ_r_zz, σ_r_xy, σ_r_xz, and σ_r_yz. Each component is mapped to an independent channel, forming a residual stress distribution feature map F_s with shape (H_part, W_part, 6).

[0163] Step S830: The warpage deformation input port converts the warpage deformation distribution prediction data of the injection molded part after demolding into a warpage deformation distribution feature map, which includes the directional component channels of the warpage deformation driving direction vector.

[0164] After demolding, the warp deformation input port of the performance backpropagation network performs feature map transformation on the Data_Warp_Vector. Each surface point in the Data_Warp_Vector contains three directional components of the warp deformation driving direction vector V_warp: V_warp_x, V_warp_y, and V_warp_z. Each directional component is mapped to an independent channel, forming a warp deformation distribution feature map F_w with shape (H_part, W_part, 3).

[0165] Step S840: The shrinkage dent input port converts the predicted data of shrinkage dent distribution after demolding of the injection molded part into a binary feature map of shrinkage dent distribution. The binary feature map of shrinkage dent distribution takes a first value in the shrinkage dent occurrence area and a second value outside the shrinkage dent occurrence area.

[0166] After demolding, the shrinkage indentation input port of the backpropagation network performs a binary feature map transformation operation on Data_Shrink. An initial all-zero tensor F_h of shape (H_part, W_part, 1) is created. For each shrinkage indentation region in Data_Shrink, all mesh cells within the boundary of that region are set to 1 in F_h. Mesh cells outside the shrinkage indentation region remain at 0.

[0167] Step S850: The post-demolding performance backpropagation network fuses the residual stress distribution feature map, the warping deformation distribution feature map, and the shrinkage indentation distribution binary feature map through the channel splicing module to generate a multimodal performance fusion feature tensor.

[0168] After demodulation, the channel stitching module of the performance backpropagation network stitches together F_s obtained in step S820, F_w obtained in step S830, and F_h obtained in step S840 along the channel dimension. The shape of F_s is (H_part, W_part, 6), the shape of F_w is (H_part, W_part, 3), and the shape of F_h is (H_part, W_part, 1). After stitching, a multimodal performance fusion feature tensor F_fu with the shape (H_part, W_part, 10) is obtained.

[0169] Step S860: The post-demolding performance backpropagation network processes the multimodal performance fusion feature tensor through an attention module. The attention module includes a channel attention submodule and a spatial attention submodule. The channel attention submodule generates a channel attention weight vector and performs a channel-wise weighted multiplication with the multimodal performance fusion feature tensor to generate a channel-enhanced fusion feature tensor. The spatial attention submodule generates a spatial attention weight map and performs a position-wise weighted multiplication with the channel-enhanced fusion feature tensor to generate an attention-enhanced fusion feature tensor.

[0170] The input F_fu is fed into the attention module. The channel attention submodule first performs global average pooling and global max pooling on F_fu, obtaining two pooling feature vectors of shape (1, 1, 10). These two pooling feature vectors are then fed into a shared two-layer fully connected network. The first fully connected layer compresses the number of channels from 10 to 2, and the second fully connected layer restores the number of channels from 2 to 10. The outputs of the two fully connected networks are then summed element-wise and passed through a sigmoid activation function to obtain a channel attention weight vector A_c of shape (1, 1, 10). A_c is then multiplied by F_fu channel-wise with weights: F_ca = F_fu The spatial attention submodule A_c calculates the average and maximum values ​​of F_ca along the channel dimension, resulting in two spatial feature maps of shape (H_part, W_part, 1). These two spatial feature maps are concatenated along the channel dimension and passed through a 7x7 convolutional layer and a sigmoid activation function to obtain a spatial attention weight map A_s of shape (H_part, W_part, 1). A_s is then multiplied positionally by F_ca with weighted averages: F_att = F_ca A_s.

[0171] Step S870: The post-demolding performance backpropagation network processes the attention-enhanced fusion feature tensor through a fully convolutional inversion module. The fully convolutional inversion module contains multiple stride convolutional layers, each followed by a batch normalization layer and an activation layer. The fully convolutional inversion module outputs an injection molding process parameter inversion vector. Each component of the injection molding process parameter inversion vector corresponds to the correction coefficient of the mold temperature zone control parameter, the correction coefficient of the mold pressure segmented loading parameter, and the correction coefficient of the mold locking timing control parameter.

[0172] The F_att output from step S860 is input into the fully convolutional inversion module. This fully convolutional inversion module contains four stride convolutional layers. The first stride convolutional layer has 10 input channels, 64 output channels, a 3x3 kernel size, a stride of 2, and padding of 1, followed by a batch normalization layer and a ReLU activation layer. The second stride convolutional layer has 64 input channels, 128 output channels, a 3x3 kernel size, a stride of 2, and padding of 1, followed by a batch normalization layer and a ReLU activation layer. The third stride convolutional layer has 128 input channels, 256 output channels, a 3x3 kernel size, a stride of 2, and padding of 1, followed by a batch normalization layer and a ReLU activation layer. The fourth stride convolutional layer has 256 input channels, 512 output channels, a 3x3 kernel size, a stride of 2, and padding of 1, followed by a batch normalization layer and a ReLU activation layer. The feature map output from the fourth layer is subjected to global average pooling to obtain a global feature vector of shape (1, 1, 512). This global feature vector is flattened and input into a fully connected layer. The fully connected layer has an input dimension of 512 and an output dimension of 3, outputting an injection molding process parameter inversion vector V_inv of shape (3,). The first component corresponds to the correction coefficient C_t of the mold temperature zone control parameter, the second component corresponds to the correction coefficient C_p of the mold pressure segment loading parameter, and the third component corresponds to the correction coefficient C_l of the mold locking timing control parameter.

[0173] Step S880: The post-demolding performance backpropagation network processes the injection molding process parameter inversion vector through the output mapping module. The output mapping module maps the correction coefficients of the mold temperature zone control parameters, the mold pressure segment loading parameters, and the mold locking timing control parameters into a set of mold structure correction parameters.

[0174] The output mapping module receives V_inv from step S870 and maps its components to specific mold structure correction parameters. For the correction coefficient C_t of the temperature zone control parameters, the corrected temperature loading value T_c = C_t is calculated. T_o; For the correction factor C_p of the pressure segment loading parameter, calculate the corrected pressure loading value P_c=C_p. P_o; For the correction coefficient C_l of the locking timing control parameter, calculate the corrected locking force loading value F_c=C_l. F_o. Integrate the above-mentioned modified parameters into a mold structure modification parameter set Set_SC={T_c, P_c, F_c}.

[0175] Figure 2 This application illustrates a mold stress simulation testing system 100 based on the structural performance of injection molded parts, comprising a processor 1001 and a memory 1003. The processor 1001 and the memory 1003 are connected, for example, via a bus 1002. Optionally, the mold stress simulation testing system 100 may further include a transceiver 1004, which can be used for data interaction between this mold stress simulation testing system and other mold stress simulation testing systems based on the structural performance of injection molded parts, such as sending and / or receiving data. It should be noted that in actual scheduling, the transceiver 1004 is not limited to one, and the structure of this mold stress simulation testing system 100 does not constitute a limitation on the embodiments of this application.

[0176] The memory 1003 is used to store program code for executing the embodiments of this application, and its execution is controlled by the processor 1001. The processor 1001 is used to execute the program code stored in the memory 1003 to implement the steps shown in the foregoing method embodiments.

[0177] The above description is only an optional implementation method for some implementation scenarios of this application. It should be noted that for those skilled in the art, other similar implementation methods based on the technical concept of this application, without departing from the technical concept of this application, also fall within the protection scope of the embodiments of this application.

Claims

1. A mold stress simulation testing method based on the structural performance of injection molded parts, characterized in that, The method includes: Obtain a set of structural performance parameters for injection molded parts and a set of mold operating parameters. The set of structural performance parameters for injection molded parts includes geometric configuration distribution parameters, rib layout parameters, and gate location parameters. The set of mold operating parameters includes mold temperature zone control parameters, mold pressure segmented loading parameters, and mold locking timing control parameters. The set of structural performance parameters of the injection molded part is subjected to stress transmission path analysis processing to generate a main stress transmission path network and a branch stress transmission path network inside the injection molded part. The main stress transmission path network inside the injection molded part includes a first set of stress transmission channels extending along the main direction of the geometric configuration of the injection molded part. The branch stress transmission path network inside the injection molded part includes a second set of stress transmission channels extending from the branch path network to the local feature area of ​​the injection molded part. Based on the main path network of stress transmission inside the injection molded part and the branch path network of stress transmission inside the injection molded part, the set of mold working condition parameters is subjected to stress loading boundary condition mapping to obtain the set of stress loading boundary conditions for each spatial region on the surface of the mold cavity. The set of stress loading boundary conditions includes a first set of stress loading boundary conditions corresponding to the end of the first set of stress transmission channels and a second set of stress loading boundary conditions corresponding to the end of the second set of stress transmission channels. The set of stress loading boundary conditions is input into the mold stress field distribution simulation model for iterative calculation of stress field distribution, and the results of the iterative calculation of stress field distribution are output. The mold stress field distribution simulation model uses the set of stress loading boundary conditions as driving boundary conditions and the main path network and branch path network of stress transmission inside the injection molded part as stress propagation constraints. The results of the iterative calculation of stress field distribution independently include stress field distribution data on the surface of the mold cavity, pressure field distribution data inside the mold cavity, and locking force field distribution data on the mold closed contact surface. The stress field distribution iterative calculation results are subjected to inverse mapping processing of injection molded part structural performance to obtain the inverse mapping results of injection molded part structural performance. The inverse mapping results of injection molded part structural performance include the predicted data of residual stress distribution after demolding, the predicted data of warpage deformation distribution after demolding, and the predicted data of shrinkage indentation distribution after demolding.

2. The mold stress simulation test method based on the structural performance of injection molded parts according to claim 1, characterized in that, The process of performing stress transmission path analysis on the set of structural performance parameters of the injection molded part to generate the main stress transmission path network and the branch stress transmission path network inside the injection molded part includes: The main direction extension line and the secondary direction extension line of the injection molded part's geometric configuration distribution parameters are analyzed. Based on the main direction extension line of the injection molded part's geometric configuration, an initial set of main transmission paths extending along the main direction of the geometric configuration inside the injection molded part is generated. The path direction of each initial main transmission path in the initial set of main transmission paths is spatially parallel to the direction of the main direction extension line of the injection molded part's geometric configuration. Based on the rib extension direction line and the connection boundary contour of the rib and injection molded part base area in the rib layout parameters, target trunk transmission paths that have a spatial intersection relationship with the rib extension direction line are selected from the initial trunk transmission path set, and the target trunk transmission paths are marked as rib associated trunk transmission paths. Extract the spatial intersection coordinates of the main stress transmission path associated with the stiffener and the extension direction line of the stiffener. Using the spatial intersection coordinates as the starting point of the stress transmission path bifurcation, extend along the extension direction line of the stiffener away from the main stress transmission path associated with the stiffener to generate an initial set of branch transmission paths connected to the main stress transmission path associated with the stiffener. The end of each initial branch transmission path in the initial branch transmission path set is extended to the coordinates of the end point of the rib in the injection molded part rib layout parameters. At the coordinates of the end point of the rib, the end direction of the initial branch transmission path is adjusted according to the boundary of the gate adjacent area in the injection molded part gate position parameters, so that the end direction of the initial branch transmission path points to the local feature area of ​​the injection molded part within the boundary of the gate adjacent area, thereby generating a final branch transmission path set connected to the rib associated main transmission path. The main stress transmission path associated with the rib plate and other main stress transmission paths in the initial set of main stress transmission paths other than the main stress transmission path associated with the rib plate are connected by a path network topology to form a main stress transmission path network for the internal stress transmission of the injection molded part, in which multiple main stress transmission paths are interconnected through path intersections. Each main stress transmission path in the main stress transmission path network for the internal stress transmission of the injection molded part corresponds to a first stress transmission channel in the first set of stress transmission channels. Each final branch transmission path in the set of final branch transmission paths is processed by path network topology connection to form an internal stress transmission branch path network of the injection molded part, in which multiple branch transmission paths are connected to the main stress transmission path network of the injection molded part through the coordinates of the spatial intersection point. Each branch transmission path in the internal stress transmission branch path network of the injection molded part corresponds to a second stress transmission channel in the set of second stress transmission channels.

3. The mold stress simulation test method based on the structural performance of injection molded parts according to claim 1, characterized in that, The stress loading boundary condition mapping process is performed on the mold working condition parameter set based on the main stress transmission path network and the branch stress transmission path network inside the injection molded part, to obtain the stress loading boundary condition set for each spatial region on the mold cavity surface, including: Extract the position coordinates of the end point of each first stress transmission channel in the stress transmission backbone path network inside the injection molded part, and perform a coordinate projection mapping operation in the spatial coordinate system of the mold cavity surface according to the position coordinates of the end point of the channel to obtain the boundary of the first projection area of ​​the mold cavity surface corresponding to the position coordinates of the end point of each first stress transmission channel. Extract the channel end point coordinates of each second stress transmission channel in the internal stress transmission branch path network of the injection molded part, and perform a coordinate projection mapping operation in the space coordinate system of the mold cavity surface according to the channel end point coordinates to obtain the boundary of the second projection area of ​​the mold cavity surface corresponding to the channel end point coordinates of each second stress transmission channel. The mold temperature zoning control parameters in the mold working condition parameter set are analyzed to obtain the spatial boundaries of each temperature control zone and the temperature loading value corresponding to each temperature control zone. The spatial boundary of each temperature control zone is subjected to spatial overlap detection processing with the boundary of the first projection area of ​​the mold cavity surface and the boundary of the second projection area of ​​the mold cavity surface, respectively, to generate a first temperature loading boundary condition subset corresponding to the boundary of the first projection area of ​​the mold cavity surface and a second temperature loading boundary condition subset corresponding to the boundary of the second projection area of ​​the mold cavity surface. The mold pressure segment loading parameters in the mold working condition parameter set are analyzed to obtain the time interval of each pressure loading stage and the pressure loading value corresponding to each pressure loading stage. The time interval of each pressure loading stage is subjected to time overlap interval detection processing with the time axis of the injection molding process to generate a first pressure loading boundary condition subset corresponding to the boundary of the first projection area of ​​the mold cavity surface and a second pressure loading boundary condition subset corresponding to the boundary of the second projection area of ​​the mold cavity surface. The mold locking timing control parameters in the mold working condition parameter set are analyzed to obtain the start time and duration of each locking force loading stage and the corresponding locking force loading value of each locking force loading stage. A locking force loading time window is generated based on the start time and duration. The locking force loading time window is then subjected to time overlap interval detection processing with the time axis of the injection molding process to generate a first locking force loading boundary condition subset corresponding to the boundary of the first projection area of ​​the mold cavity surface and a second locking force loading boundary condition subset corresponding to the boundary of the second projection area of ​​the mold cavity surface. The first temperature loading boundary condition subset, the first pressure loading boundary condition subset, and the first locking force loading boundary condition subset are combined and processed to generate a first stress loading boundary condition subset corresponding to the end of the first stress transmission channel set. The first stress loading boundary condition subset includes temperature loading boundary conditions, pressure loading boundary conditions, and locking force loading boundary conditions that are synchronously defined on the time axis. The second temperature loading boundary condition subset, the second pressure loading boundary condition subset, and the second locking force loading boundary condition subset are combined and processed to generate a second stress loading boundary condition subset corresponding to the end of the second stress transmission channel set. The second stress loading boundary condition subset includes temperature loading boundary conditions, pressure loading boundary conditions, and locking force loading boundary conditions that are synchronously defined on the time axis. The first set of stress loading boundary conditions and the second set of stress loading boundary conditions are combined into a set of stress loading boundary conditions for each spatial region of the mold cavity surface.

4. The mold stress simulation test method based on the structural performance of injection molded parts according to claim 1, characterized in that, The step of inputting the set of stress loading boundary conditions into the mold stress field distribution simulation model for iterative calculation of stress field distribution and outputting the iterative calculation results of stress field distribution includes: The first set of stress loading boundary conditions in the set of stress loading boundary conditions is loaded into the input boundary condition interface of the mold stress field distribution simulation model. The mold stress field distribution simulation model applies the first driving boundary condition on the spatial region corresponding to the end of the first set of stress transmission channels on the surface of the mold cavity according to the temperature loading boundary condition, pressure loading boundary condition and locking force loading boundary condition synchronously defined in the first set of stress loading boundary conditions. The second stress loading boundary condition subset in the stress loading boundary condition set is loaded into the input boundary condition interface of the mold stress field distribution simulation model. According to the coordinated loading sequence of temperature loading value, pressure loading value and locking force loading value contained in each second stress loading boundary condition in the second stress loading boundary condition subset on the time axis, a second driving boundary condition is applied on the spatial region corresponding to the end of the second stress transmission channel set on the mold cavity surface. The main stress transmission path network inside the injection molded part is used as the main propagation channel constraint for stress wave propagation inside the mold cavity, and the branch stress transmission path network inside the injection molded part is used as the branch propagation channel constraint for stress wave propagation inside the mold cavity. The main propagation channel constraint limits the propagation speed priority and propagation amplitude attenuation rate of stress wave in the spatial direction corresponding to the first set of stress transmission channels, and the branch propagation channel constraint limits the propagation speed priority and propagation amplitude attenuation rate of stress wave in the spatial direction corresponding to the second set of stress transmission channels. Using the first driving boundary condition and the second driving boundary condition as stress wave excitation sources, and the main propagation channel constraint and the branch propagation channel constraint as stress wave propagation path restriction conditions, an iterative calculation of the stress wave propagation process is performed in the internal space of the mold cavity. In each iteration calculation, the stress wave starts from the spatial position where the first driving boundary condition and the second driving boundary condition are located, and propagates layer by layer into the internal space of the mold cavity along the spatial direction defined by the main propagation channel constraint and the branch propagation channel constraint. During each iteration of the calculation, the arrival time, propagation direction, and characteristic parameters of the stress wave on each spatial grid node inside the mold cavity are recorded. Based on the arrival time, propagation direction, and characteristic parameters of the stress wave, the stress tensor component values ​​on each spatial grid node are calculated and updated using the stress wave propagation model. The iterative calculation of the stress wave propagation process is stopped when the preset iteration termination condition is reached. The iteration termination condition is that the stress wave propagates to all spatial grid nodes inside the mold cavity and the change in the stress tensor component value on each spatial grid node is lower than the preset change threshold. After stopping the iterative calculation, the set of stress tensor component values ​​on each spatial grid node on the surface of the mold cavity is output as the stress field distribution data of the mold cavity surface. The set of pressure scalar values ​​on each spatial grid node inside the mold cavity is output as the pressure field distribution data inside the mold cavity. The set of locking force vector component values ​​on each spatial grid node on the closed contact surface of the mold is output as the locking force field distribution data of the closed contact surface of the mold.

5. The mold stress simulation test method based on the structural performance of injection molded parts according to claim 1, characterized in that, The step of performing a reverse mapping process on the stress field distribution iterative calculation results to obtain the reverse mapping results of the injection molded part's structural performance includes: The stress field distribution data of the mold cavity surface is back-projected from the spatial coordinate system of the mold cavity surface to the geometric spatial coordinate system of the injection molded part to generate surface stress field distribution data before demolding corresponding to each spatial position point on the surface of the injection molded part. The surface stress field distribution data before demolding includes the surface stress tensor component values ​​at each spatial position point on the surface of the injection molded part. The pressure field distribution data inside the mold cavity is back-projected from the internal spatial coordinate system of the mold cavity to the geometric spatial coordinate system of the injection molded part to generate internal pressure field distribution data before demolding corresponding to each spatial location point inside the injection molded part. The internal pressure field distribution data before demolding includes the internal pressure scalar value at each spatial location point inside the injection molded part. The locking force field distribution data of the mold closed contact surface is back-projected from the mold closed contact surface spatial coordinate system to the injection part parting surface region corresponding to the mold closed contact surface in the injection part geometric spatial coordinate system, generating the pre-demolding parting surface locking force distribution data corresponding to each spatial position point in the injection part parting surface region. The pre-demolding parting surface locking force distribution data includes the locking force vector component values ​​at each spatial position point in the injection part parting surface region. Based on the surface stress field distribution data before demolding, the internal pressure field distribution data before demolding, and the parting surface locking force distribution data before demolding, a coupled distribution field is constructed for the injection molded part before demolding. Each spatial location point in the coupled distribution field simultaneously contains the surface stress tensor component value, the internal pressure scalar value, and the parting surface locking force vector component value. The coupling distribution field is subjected to demolding boundary release processing, the locking force vector component values ​​at the spatial position points of the parting surface area of ​​the injection molded part are removed, and a transient redistribution sequence of the three-dimensional stress-pressure field of the injection molded part during demolding is generated. The transient redistribution sequence of the three-dimensional stress-pressure field of the injection molded part during demolding includes the stress-pressure field distribution state at the beginning of demolding, the stress-pressure field distribution state at the middle of demolding, and the stress-pressure field distribution state at the end of demolding. Based on the surface stress tensor component values ​​in the stress-pressure field distribution state at the end of the demolding, the residual stress tensor component values ​​that have not been released at each spatial location point on the surface of the injection molded part are extracted, and the residual stress tensor component values ​​at each spatial location point on the surface of the injection molded part are processed by spatial distribution set processing to generate the residual stress distribution prediction data after the demolding of the injection molded part. Based on the surface stress tensor component values ​​and internal pressure scalar values ​​in the stress-pressure field distribution state at the end of the demolding, the stress gradient vector direction and pressure gradient vector direction at each spatial location point on the surface of the injection molded part are calculated. The vector difference between the stress gradient vector direction and the pressure gradient vector direction is taken as the warping deformation driving direction vector at each spatial location point on the surface of the injection molded part. The warping deformation driving direction vectors at each spatial location point on the surface of the injection molded part are spatially distributed and processed to generate the warping deformation distribution prediction data of the injection molded part after demolding. Based on the normal stress component value in the surface stress tensor component value and the pressure gradient change rate in the internal pressure scalar value in the stress-pressure field distribution state at the end of the demolding, target spatial location points where the normal stress component value exceeds the normal stress threshold and target spatial region boundaries where the pressure gradient change rate exceeds the pressure gradient change rate threshold are identified at various spatial location points on the surface of the injection molded part. The intersection area of ​​the target spatial location point and the target spatial region boundary is marked as the shrinkage indentation area, and the shrinkage indentation distribution prediction data after the demolding of the injection molded part is generated.

6. The mold stress simulation test method based on the structural performance of injection molded parts according to claim 1, characterized in that, After obtaining the set of structural performance parameters of the injection molded part and the set of mold operating condition parameters, the method further includes: The boundaries of the wall thickness distribution region and the wall thickness gradient transition region of the injection molded part in the geometric configuration distribution parameters of the injection molded part are analyzed to generate a spatial distribution topology map of the wall thickness of the injection molded part. The spatial distribution topology map of the wall thickness of the injection molded part includes the boundaries of the thin-walled region, the thick-walled region, and the wall thickness gradient transition region of the injection molded part. The rib extension direction line, rib height distribution curve, and rib root radius distribution curve in the rib layout parameters of the injection molded part are analyzed to generate a spatial layout topology diagram of the injection molded part rib. The spatial layout topology diagram of the injection molded part rib includes the spatial trajectory of the rib extension direction line in the geometric coordinate system of the injection molded part, the height change trajectory of the rib height distribution curve along the rib extension direction line, and the radius change trajectory of the rib root radius distribution curve along the rib extension direction line. The gate location parameters of the injection molded part are analyzed, including the gate opening position coordinates, gate opening shape contour, and gate opening direction vector on the injection molded part geometry. A spatial distribution topology map of the injection molded part gate is generated. The spatial distribution topology map of the injection molded part gate includes a local coordinate system of the gate with the opening position coordinates as the origin, a set of contour point coordinates of the gate opening shape contour in the local coordinate system, and the direction vector component of the gate opening direction vector in the local coordinate system. The topology diagrams of the wall thickness spatial distribution, the rib spatial layout, and the gate spatial distribution of the injection molded part are overlaid and fused to generate a spatial topology fusion diagram of the structural performance parameters of the injection molded part. The spatial topology fusion diagram of the structural performance parameters of the injection molded part contains a wall thickness category identifier, a rib layout association identifier, and a gate distance field value at each spatial location point in the geometric coordinate system of the injection molded part. Based on the wall thickness category identifier and rib layout association identifier in the spatial topology fusion diagram of the injection molded part structural performance parameters, the target spatial location point with the wall thickness category identifier being the boundary of the thin-walled region and the rib layout association identifier being connected to the distribution curve of the radius of the fillet at the root of the rib is identified, and the target spatial location point is marked as the set of stress transmission path starting points. Based on the gate distance field values ​​in the spatial topology fusion diagram of the injection molded part's structural performance parameters, a subset of near-gate stress transmission starting points with gate distance field values ​​less than the distance field threshold and a subset of far-gate stress transmission starting points with gate distance field values ​​greater than or equal to the distance field threshold are selected from the set of stress transmission path starting points. The near-gate stress transmission starting point subset is used to preferentially construct the main transmission path in the internal stress transmission backbone path network of the injection molded part, and the far-gate stress transmission starting point subset is used to secondarily construct the backbone transmission path in the internal stress transmission backbone path network of the injection molded part.

7. The mold stress simulation test method based on the structural performance of injection molded parts according to claim 1, characterized in that, After obtaining the set of structural performance parameters of the injection molded part and the set of mold operating condition parameters, the method further includes: The spatial boundaries of each temperature control zone in the mold temperature zone control parameters and the temperature loading time curve corresponding to each temperature control zone are analyzed. The temperature loading time curve includes the temperature loading start time point, the temperature loading end time point, and the temperature linear rise slope between the temperature loading start time point and the temperature loading end time point. The time intervals of each pressure loading stage in the segmented loading parameters of the mold pressure and the spatial distribution pattern of pressure loading corresponding to each pressure loading stage are analyzed. The spatial distribution pattern of pressure loading includes the coordinates of the pressure loading center point, the diffusion rate parameter of pressure loading from the center point to the edge, and the diffusion attenuation index of pressure loading from the center point to the edge. The starting time and duration of each locking force loading stage in the mold locking timing control parameters are analyzed, as well as the locking force loading direction vector corresponding to each locking force loading stage. The locking force loading direction vector includes the normal component of the locking force in the normal direction of the mold closed contact surface and the tangential component of the locking force in the tangential direction of the mold closed contact surface. The temperature loading time curve, the pressure loading spatial distribution pattern, and the locking force loading direction vector are aligned on the time axis to generate a joint distribution field. The joint distribution field includes, at each time point, the correspondence between the spatial boundary of the temperature control area and the temperature loading value, the correspondence between the coordinates of the pressure loading center point and the pressure diffusion velocity parameter and the pressure diffusion attenuation index, and the values ​​of the normal and tangential components of the locking force loading direction vector. The joint distribution field is input into the mold response characteristic prediction model to perform mold cavity surface response characteristic prediction calculation. The mold response characteristic prediction model takes the joint distribution field as input excitation and the thermal conductivity coefficient matrix, elastic modulus tensor and Poisson's ratio distribution field of the mold material as mold intrinsic property constraints. It outputs the thermal expansion displacement, elastic deformation displacement and viscoelastic creep at each spatial position point on the mold cavity surface at different time points. Based on the thermal expansion displacement, elastic deformation displacement, and viscoelastic creep of each spatial location point on the surface of the mold cavity at different time points, calculate the total displacement of each spatial location point on the surface of the mold cavity at different time points, and mark the target spatial location point at the target time point that exceeds the elastic limit displacement threshold of the mold material as the mold yield initiation point. The set of coordinates of the mold yield initiation point in the spatial coordinate system of the mold cavity surface is used as the spatial distribution map of the mold yield initiation point. The spatial distribution map of the mold yield initiation point is spatially superimposed with the stress transmission backbone path network inside the injection molded part to identify the first mold yield initiation point subset located in the spatial region corresponding to the end of the first stress transmission channel set and the second mold yield initiation point subset located in the spatial region corresponding to the end of the second stress transmission channel set. Based on the distribution density of the first mold yield initiation point subset in the corresponding spatial region at the end of the first stress transmission channel set, adjust the loading order priority of the temperature loading value, pressure loading value and locking force loading value of the first stress loading boundary condition in the first stress loading boundary condition subset on the time axis. Based on the distribution density of the second mold yield initiation point subset in the corresponding spatial region at the end of the second stress transmission channel set, the loading order priority of the temperature loading value, pressure loading value, and locking force loading value of the second stress loading boundary condition in the second stress loading boundary condition subset on the time axis is adjusted.

8. The mold stress simulation test method based on the structural performance of injection molded parts according to claim 1, characterized in that, After inputting the set of stress loading boundary conditions into the mold stress field distribution simulation model for iterative calculation of stress field distribution, the method further includes: Extract the stress wave propagation path trajectory on each spatial grid node inside the mold cavity recorded in the stress field distribution iterative calculation process of the mold stress field distribution simulation model. The stress wave propagation path trajectory includes the node number sequence of each spatial grid node that the stress wave passes through in sequence from the excitation source position. Based on the node number sequence in the stress wave propagation path trajectory, target spatial grid nodes in which the node number sequence in the stress wave propagation path trajectory appears to repeat cyclically are identified, and the target spatial grid nodes are marked as stress wave standing wave nodes. Extract the coordinate set of all target spatial mesh nodes marked as stress wave standing wave nodes in the spatial coordinate system inside the mold cavity to generate a spatial distribution map of stress wave standing wave nodes inside the mold cavity; The spatial distribution map of stress wave standing wave nodes inside the mold cavity is spatially superimposed with the stress transmission trunk path network inside the injection molded part to identify stress wave standing wave nodes located on the trunk transmission path in the stress transmission trunk path network inside the injection molded part. The stress wave standing wave nodes located on the trunk transmission path are marked as stress wave propagation blocking nodes. The spatial distribution map of stress wave standing wave nodes inside the mold cavity is spatially superimposed with the stress transmission branch path network inside the injection molded part to identify stress wave standing wave nodes located on the branch transmission path in the stress transmission branch path network inside the injection molded part. The stress wave standing wave nodes located on the branch transmission path are marked as stress wave energy dissipation nodes. Based on the spatial distribution of the stress wave propagation blocking node in the stress transmission backbone path network inside the injection molded part, the stress wave propagation speed priority parameter of the backbone propagation channel constraint in the mold stress field distribution simulation model is modified to reduce the stress wave propagation speed priority on the backbone transmission path where the stress wave propagation blocking node is located, and increase the stress wave propagation speed priority on the detour backbone transmission path that bypasses the stress wave propagation blocking node. Based on the spatial distribution of the stress wave energy dissipation nodes in the stress transmission branch path network inside the injection molded part, the stress wave propagation amplitude attenuation rate parameter of the branch propagation channel constraint in the mold stress field distribution simulation model is modified to increase the stress wave propagation amplitude attenuation rate on the branch transmission path where the stress wave energy dissipation nodes are located, and decrease the stress wave propagation amplitude attenuation rate on the branch transmission path where non-stress wave energy dissipation nodes are located.

9. The mold stress simulation test method based on the structural performance of injection molded parts according to claim 1, characterized in that, After performing the reverse mapping processing of the injection molded part structural performance on the iterative calculation results of the stress field distribution, the method further includes: The residual stress tensor component values ​​at various spatial locations on the surface of the injection molded part in the predicted residual stress distribution data after demolding are classified. Based on the spatial angle between the direction of the maximum principal stress in the residual stress tensor component values ​​and the extension line of the main direction of the geometric configuration of the injection molded part, the spatial locations on the surface of the injection molded part are divided into a first set of residual stress regions where the residual stress is distributed along the main direction of the geometric configuration and a second set of residual stress regions where the residual stress is distributed along the secondary direction of the geometric configuration. The warping deformation driving direction vectors at various spatial locations on the surface of the injection molded part in the warping deformation distribution prediction data after demolding are clustered. Based on the size of the vector angle between the warping deformation driving direction vectors, the spatial locations on the surface of the injection molded part are divided into a third set of warping deformation regions where the warping deformation driving direction tends to be consistent and a fourth set of warping deformation regions where the warping deformation driving direction is divergent. The spatial boundaries of the shrinkage indentation area in the predicted data of shrinkage indentation distribution after demolding of the injection molded part are processed by depth classification of the shrinkage indentation area. Based on the difference between the normal stress component value at the spatial location point on the surface of the injection molded part within the shrinkage indentation area and the normal stress component value at the spatial location point on the surface of the injection molded part in the adjacent non-shrinkage indentation area, the shrinkage indentation area is divided into a set of deep shrinkage indentation areas and a set of shallow shrinkage indentation areas. Spatial overlay analysis of the first set of residual stress regions, the second set of residual stress regions, the third set of warped deformation regions, the fourth set of warped deformation regions, the deep shrinkage indentation region set, and the shallow shrinkage indentation region set is performed to generate a spatial distribution map of the surface quality risk regions of the injection molded part. The spatial distribution map of the surface quality risk regions of the injection molded part includes the boundary of the first risk region that belongs to both the first set of residual stress regions and the third set of warped deformation regions, the boundary of the second risk region that belongs to both the second set of residual stress regions and the fourth set of warped deformation regions, and the boundary of the third risk region that belongs to both the deep shrinkage indentation region set and the first set of residual stress regions. Based on the first risk area boundary, the second risk area boundary, and the third risk area boundary in the spatial distribution map of the surface quality risk areas of the injection molded part, a set of injection mold correction parameter suggestions is generated. The set of injection mold correction parameter suggestions includes correction suggestions for mold temperature zoning control parameters for the first risk area boundary, correction suggestions for mold pressure segmented loading parameters for the second risk area boundary, and correction suggestions for mold locking timing control parameters for the third risk area boundary.

10. A mold stress simulation and testing system based on the structural performance of injection molded parts, characterized in that, The method includes a processor and a computer-readable storage medium storing machine-executable instructions that, when executed by the processor, implement the mold stress simulation test method based on the structural performance of injection molded parts as described in any one of claims 1-9.

Images