Intelligent optimization method for forming parameters of plastic luggage

Abstract

The application discloses a plastic luggage forming parameter intelligent optimization method, relates to the technical field of processing optimization, fits the response curve of each feature in the material-technological feature matrix to forming defects, calculates the unit structure performance index under different technological configurations based on the response curve, and is used for evaluating the forming quality of the plastic luggage, according to the design partition of the plastic luggage, statistics the unit structure forming parameter and the performance index of each region, constructs the overall performance mapping model through regional weighting, judges whether the plastic luggage forming parameter needs to be optimized according to the performance index of each region and the overall performance mapping model, and generates the corresponding plastic luggage forming parameter optimization strategy according to the judgment result. The optimization method establishes the overall performance mapping model through the performance index statistics and the weighted processing of the regional partition, makes the forming optimization strategy consider the local quality control and the whole performance cooperation, and enhances the comprehensiveness and reliability of the optimization decision.

Description

Technical Field

[0001] This invention relates to the field of processing optimization technology, specifically to a method for intelligent optimization of molding parameters for plastic bags and cases. Background Technology

[0002] Plastic bag molding is an important branch of plastic product processing. It mainly uses processes such as injection molding, blow molding, and hot pressing to process plastic raw materials (such as PP, ABS, PC, etc.) into various bag products, including suitcases, storage boxes, toolboxes, etc. This process involves key parameters such as temperature, pressure, and cooling time, which directly affect the product's strength, surface quality, and dimensional accuracy. Intelligent optimization methods refer to the use of technologies such as artificial intelligence, machine learning, and big data analysis to dynamically model and optimize key parameters (such as temperature, pressure, time, cooling rate, etc.) in processes such as injection molding, blow molding, or hot pressing, in order to improve product quality, reduce energy consumption, and reduce defect rates.

[0003] The existing technology has the following drawbacks:

[0004] In practice, existing optimization methods often rely on subjective adjustments to molding parameters such as temperature, pressure, and cooling time. Furthermore, they fail to adequately consider the complex coupling relationship between material properties, mold structure, and actual molding behavior, making it difficult to achieve precise control and stable quality. At the same time, by ignoring the differences in the sensitivity of molding parameters to the structure of each molding area (such as the box body, corners, and reinforcing ribs), it is impossible to assess the impact of regional performance on the entire plastic box from a holistic perspective, resulting in significant fluctuations in product quality.

[0005] Based on this, the present invention proposes an intelligent optimization method for molding parameters of plastic bags. By statistically analyzing and weighting the performance indicators of regional partitions, an overall performance mapping model is established, which enables the molding optimization strategy to take into account both local quality control and overall performance coordination, thereby enhancing the comprehensiveness and reliability of optimization decisions. Summary of the Invention

[0006] The purpose of this invention is to provide an intelligent optimization method for molding parameters of plastic bags and cases, so as to solve the problems in the prior art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an intelligent optimization method for molding parameters of plastic bags, the optimization method comprising the following steps:

[0008] S1: Obtain the molecular structure parameters of the plastic raw material, load the molecular structure parameters into the molding cavity frame formed by the target bag mold structure, and obtain the plastic-mold composite structure after pretreatment;

[0009] S2: Perform multiphysics coupling simulation on the plastic-mold composite structure, perform dynamic calculations in conjunction with the processing path, and generate a molding response process model after extracting behavioral data;

[0010] S3: Based on the molding response process model, quantitative values ​​are assigned to material parameters and process parameters to form a material-process feature matrix, and the response curves of each feature in the material-process feature matrix to molding defects are fitted.

[0011] S4: Calculate the unit structural performance index under different process configurations based on the response curve to evaluate the molding quality of plastic bags;

[0012] S5: Based on the design zones of plastic bags, statistically analyze the unit structural molding parameters and performance indicators of each zone, and construct an overall performance mapping model through zone weighting;

[0013] S6: Based on the performance indicators of each region and the overall performance mapping model, determine whether the molding parameters of plastic bags need to be optimized, and generate corresponding optimization strategies for the molding parameters of plastic bags based on the judgment results.

[0014] Preferably, in step S3: based on the molding response process model, quantitative values ​​are assigned to the material parameters and process parameters to form a material-process feature matrix, and the response curves of each feature in the material-process feature matrix to molding defects are fitted, including the following steps:

[0015] The material parameters and process parameters are quantified and sampled, and a material-process feature matrix containing multiple molding combinations is constructed using a hierarchical uniform design or Latin hypercube sampling method. Each row of the matrix corresponds to a molding configuration combination, and each column corresponds to a feature variable.

[0016] Extract the defect performance corresponding to each set of formed configuration combinations, and use the defect indicators as response variables;

[0017] For each feature variable and its corresponding defect index, a response curve fitting operation is performed.

[0018] Preferably, the material parameters include melt viscosity, specific heat capacity, coefficient of thermal expansion, and solid-phase transformation temperature range; the process parameters include injection pressure, mold temperature, screw speed, holding time, cooling time, and gate location; and the defects include warpage value, shrinkage degree, and bubble distribution density.

[0019] Preferably, in step S4: calculating the unit structural performance index under different process configurations based on the response curve to evaluate the molding quality of plastic bags includes the following steps:

[0020] The performance indicators include volume shrinkage rate and cooling residual stress distribution factor;

[0021] The volume shrinkage rate and the cooling residual stress distribution factor were respectively subjected to maximum-minimum normalization.

[0022] The normalized volume shrinkage rate is summed with the cooling residual stress distribution factor to obtain the anomaly index, which is used to comprehensively evaluate the molding quality level under the current process configuration.

[0023] The larger the anomaly index, the greater the dimensional deviation or structural stress anomaly in the product under the corresponding process parameters, and the worse the molding quality.

[0024] The smaller the anomaly index, the higher the stability of the molding dimensions and the more uniform the stress distribution, indicating excellent molding quality.

[0025] Preferably, the calculation logic for the volume shrinkage rate is as follows: extract the volume values ​​of the molded part when the mold cavity is filled and at the moment of demolding from the simulation data, and calculate the ratio of the volume difference to the original volume as the shrinkage rate index.

[0026] The calculation logic of the cooling residual stress distribution factor is as follows: select multiple measurement points in the target structural region, obtain the residual stress value of each measurement point, and calculate the standard deviation of the cooling residual stress in the region as the stress distribution factor.

[0027] Preferably, in step S5: based on the design zoning of the plastic bags, the unit structural molding parameters and performance indicators of each region are statistically analyzed, and an overall performance mapping model is constructed through regional weighting, including the following steps:

[0028] The bag is designed with separate sections, including the handle connection area, the corner transition area, the bottom and wheel support area, and the middle area of ​​the wall;

[0029] For each design zone, the unit structure forming parameters and performance indicators of that zone are extracted. The unit structure forming parameters are used to generate the zone weighting coefficients, and the performance indicators are anomaly indices, which are used to reflect the quality performance of that zone under the process configuration.

[0030] The global anomaly index under the current process parameter configuration is obtained by weighting and summing the anomaly indices of all regions with their corresponding normalized weight coefficients.

[0031] The global anomaly index is bound to the current material parameters and process parameters to construct the response relationship between the input parameters and the global anomaly index, thereby generating an overall performance mapping model.

[0032] Preferably, the unit structure forming parameter is the wall thickness change rate, which includes: calculating the wall thickness change rate value of each region, summing the wall thickness change rates of all regions, and using the ratio of the region's wall thickness change rate to the overall wall thickness change rate as the weighting coefficient of that region.

[0033] Preferably, in step S6: based on the performance indicators of each region and the overall performance mapping model, it is determined whether the molding parameters of the plastic bags need to be optimized, and based on the determination result, a corresponding optimization strategy for the molding parameters of the plastic bags is generated, including the following steps:

[0034] Based on the performance indicators of each region and the overall performance mapping model, it is determined whether the molding parameters of plastic bags need to be optimized, and corresponding molding parameter optimization strategies are generated based on the judgment results.

[0035] Iterate through the anomaly indices of all regions. If the anomaly index of at least one region is greater than the region's anomaly threshold, it is determined that the local performance is unqualified, and a local optimization strategy is triggered.

[0036] The local optimization strategy includes: identifying regions where the anomaly index is greater than the regional anomaly threshold, calculating the partial derivatives of the anomaly-related parameters in the region based on the response curve, and selecting at least two process parameters for local optimization based on the partial derivatives.

[0037] Provided that all regional anomaly indices are less than or equal to the regional anomaly thresholds, if the global anomaly index exceeds the global threshold, it is considered an overall quality risk, triggering a global optimization strategy.

[0038] The global optimization strategy includes: jointly optimizing all process parameters that affect the global anomaly index, and using a multi-objective optimization algorithm to perform constraint optimization on multiple objectives.

[0039] Preferably, in step S1: obtaining the molecular structure parameters of the plastic raw material, loading the molecular structure parameters into the molding cavity frame formed by the target bag mold structure, and obtaining the plastic-mold composite structure after pretreatment, includes the following steps:

[0040] Obtain the molecular structure parameters of plastic raw materials, wherein the plastic raw materials include at least one of polypropylene and acrylonitrile-butadiene-styrene copolymer;

[0041] Molecular structural parameters, including the connection mode of monomer repeating units, chain segment length, side group structure, polar distribution, molecular mass distribution range, and molecular chain conformation, are obtained through molecular modeling tools.

[0042] The molecular structure parameters are loaded into the molding cavity frame formed by the target bag mold structure. The target bag mold structure is input in CAD format and imported to form an assembly state.

[0043] After pre-processing the molding cavity frame, the pre-processing includes geometric fitting, boundary setting, and mesh generation to form a plastic-mold composite structure.

[0044] Preferably, in step S2: multiphysics coupling simulation is performed on the plastic-mold composite structure, dynamic calculation is performed in conjunction with the processing path, and behavioral data is extracted to generate a molding response process model, including the following steps:

[0045] The multiphysics field includes a thermophysics field, a fluid physics field, and a force physics field;

[0046] The thermophysical field is used to simulate the heat exchange process between the plastic melt and the mold;

[0047] The fluid physics field is used to simulate the non-Newtonian flow behavior of melt in the mold cavity;

[0048] The physical field of force is used to analyze the stress and displacement distribution during the molding process;

[0049] Combine the processing paths in injection molding, blow molding, or hot pressing to execute a dynamic simulation process including injection, filling, holding pressure, cooling, and demolding;

[0050] Full-field coupling calculations were performed at multiple time steps, and the filling rate, pressure response during the holding period, temperature gradient distribution, cooling rate, and flow front morphology were extracted.

[0051] The behavioral data is associated with the spatial coordinate system of the molding cavity, and output as a spatial distribution map or time series data stream, and a molding response process model is generated.

[0052] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0053] This invention, based on a molding response process model, quantitatively assigns values ​​to material and process parameters, forming a material-process feature matrix. It then fits the response curves of each feature in the matrix to molding defects, calculating unit structural performance indicators under different process configurations based on these response curves to evaluate the molding quality of plastic bags. According to the design zones of the plastic bags, it statistically analyzes the unit structural molding parameters and performance indicators of each zone. An overall performance mapping model is constructed through regional weighting. Based on the performance indicators of each zone and the overall performance mapping model, it determines whether the molding parameters of the plastic bags need optimization, and generates corresponding optimization strategies based on the results. This optimization method establishes an overall performance mapping model through the statistical and weighted processing of performance indicators by region, enabling the molding optimization strategy to consider both local quality control and overall component performance synergy, thus enhancing the comprehensiveness and reliability of the optimization decision. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0055] Figure 1 This is a flowchart of the optimization method of the present invention. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0057] Example 1: Please refer to Figure 1 As shown, this embodiment provides an intelligent optimization method for molding parameters of plastic bags. The optimization method includes the following steps:

[0058] S1: Obtain the molecular structure parameters of plastic raw materials (such as PP, ABS, etc.), load the molecular structure parameters into the molding cavity frame formed by the target bag mold structure, and perform preprocessing. The preprocessing includes geometric fitting, boundary setting, mesh generation and initial energy minimization in sequence to obtain the plastic-mold composite structure.

[0059] S2: Perform multi-physics coupling simulation on plastic-mold composite structures. The multi-physics fields include thermophysics, fluid physics, and force physics. Combined with the processing path, perform dynamic calculations. The processing processes include injection molding / blow molding / hot pressing. After extracting behavioral data, generate a molding response process model. The behavioral data includes filling speed, holding time, temperature gradient, cooling rate, etc.

[0060] S3: Based on the molding response process model, quantitative values ​​are assigned to material parameters (such as melt viscosity, specific heat capacity, coefficient of thermal expansion) and process parameters (injection pressure, mold temperature, screw speed, etc.) to form a material-process feature matrix, and the response curves of each feature in the material-process feature matrix to molding defects (such as warpage, shrinkage, bubbles, etc.) are fitted.

[0061] S4: Calculate the unit structural performance index under different process configurations based on the response curve, including volume shrinkage rate and cooling residual stress distribution factor (the volume shrinkage rate and cooling residual stress distribution factor are normalized to the maximum and minimum values ​​so that the value range of volume shrinkage rate and cooling residual stress distribution factor is mapped to [0,1], and the normalized volume shrinkage rate and cooling residual stress distribution factor are summed to obtain the anomaly index). The anomaly index is used to evaluate the molding quality of plastic bags.

[0062] S5: Based on the design zones of plastic bags (such as handles, corners, bottom, etc.), the unit structural forming parameters (used to generate weighting coefficients) and performance indicators (abnormality index) of each zone are statistically analyzed. An overall performance mapping model is constructed through regional weighting (abnormality index weighting of all zones) to provide basic data support for the overall bag forming optimization.

[0063] S6: Based on the performance indicators of each region and the overall performance mapping model, determine whether the molding parameters of plastic bags need to be optimized, and generate corresponding optimization strategies for the molding parameters of plastic bags based on the judgment results. (If the optimization of the molding parameters of plastic bags is determined based on the performance indicators of the region (e.g., the abnormal index of at least one partition is greater than the abnormal threshold) (mainly reflected in the substandard processing quality), then one optimization strategy for the molding parameters of plastic bags is generated. If the optimization of the molding parameters of plastic bags is determined based on the overall performance mapping model (i.e., the abnormal index of each partition is less than or equal to the abnormal threshold, but after comprehensive weighted analysis of the abnormal index of all partitions, it is determined that the overall performance is substandard (mainly reflected in the reduced service life of plastic bags in the later stage)), then another optimization strategy for the molding parameters of plastic bags is generated.)

[0064] This application, based on a molding response process model, quantitatively assigns values ​​to material and process parameters to form a material-process feature matrix. It then fits the response curves of each feature in the matrix to molding defects. Based on these response curves, it calculates unit structural performance indicators under different process configurations to evaluate the molding quality of plastic bags. According to the design zones of the plastic bags, it statistically analyzes the unit structural molding parameters and performance indicators of each zone. An overall performance mapping model is constructed through regional weighting. Based on the performance indicators of each zone and the overall performance mapping model, it determines whether the molding parameters of the plastic bags need optimization and generates corresponding optimization strategies based on the results. This optimization method establishes an overall performance mapping model through the statistical and weighted processing of performance indicators by regional zones. This allows the molding optimization strategy to consider both local quality control and overall component performance synergy, enhancing the comprehensiveness and reliability of the optimization decision.

[0065] This invention proposes an intelligent optimization method for molding parameters in plastic bags. By constructing a composite material-mold structure and introducing multiphysics coupling simulation, the system extracts behavioral data during the molding process, effectively revealing the influence of material properties and process paths on molding defects. Utilizing the coupled analysis of the molding response process model and the material-process feature matrix, a quantitative mapping between parameters and performance is achieved, significantly improving the scientific rigor and precision of parameter control. Simultaneously, this method establishes an overall performance mapping model through regional partitioning performance index statistics and weighted processing, enabling the molding optimization strategy to consider both local quality control and overall part performance synergy, enhancing the comprehensiveness and reliability of optimization decisions. Finally, the system automatically determines whether optimization is needed based on performance evaluation results and outputs targeted process parameter optimization strategies, realizing a closed-loop molding optimization process from data-driven modeling and behavioral response analysis to decision output. This significantly improves the molding consistency, quality stability, and manufacturing intelligence level of plastic bag products.

[0066] Example 2: S1: Obtain the molecular structure parameters of plastic raw materials (such as PP, ABS, etc.), load the molecular structure parameters into the molding cavity frame formed by the target bag mold structure, and perform preprocessing. The preprocessing includes geometric fitting, boundary setting, mesh generation and initial energy minimization in sequence to obtain the plastic-mold composite structure.

[0067] The molecular structure parameters of plastic raw materials (such as PP, ABS, etc.) are obtained and loaded into the molding cavity frame formed by the target bag mold structure. After preprocessing, a plastic-mold composite structure is obtained. This step serves as the basis for the intelligent optimization method of molding parameters, aiming to construct a high-precision coupling system of plastic materials and mold structure that can be used for subsequent simulations, and to ensure that the input data meets the prerequisites for multiphysics simulation calculations.

[0068] First, typical molecular structure parameters of the target plastic raw material need to be obtained through databases or molecular modeling tools (such as Materials-Studio, Gaussian, etc.). These parameters include, but are not limited to, the connection mode of monomer repeating units, chain segment length, side group structure, polarity distribution, molecular weight distribution range, and molecular chain conformation. Taking polypropylene (PP) as an example, data such as methyl side groups, isotactic / random arrangement, and chain segment flexibility should be included; taking ABS as an example, the proportional relationship of the acrylonitrile-butadiene-styrene ternary copolymer structure and chain segment polarity information are required.

[0069] Next, based on the target bag mold structure data obtained in CAD format (e.g., STEP, IGES format), the structure is imported and mapped with material molecular parameters to form a preliminary "molding cavity framework." Based on this, the entire structure needs to undergo preprocessing, specifically including the following four key steps:

[0070] Geometric Fitting: The loaded molecular structure is spatially fused with the three-dimensional structure of the mold to ensure that the molecular filling area is consistent with the geometry of the molding cavity. A fitting method based on point cloud matching and boundary envelope algorithm is adopted, and the arrangement of molecular clouds in the molding cavity is adjusted by the minimum error method to ensure that they have maximum contact with the mold wall but do not overlap.

[0071] Boundary setting: Boundary conditions are set at the edges and interface areas of the composite structure, mainly including thermal boundaries (such as temperature constraints at the mold cooling channels), flow boundaries (such as flow velocity and pressure at the inlet), and displacement boundaries (motion constraints in the mold closing direction). The logic of boundary setting follows the principle of physical process realizability, that is, ensuring that the plastic melt has a complete flow channel during injection / compression and that the boundary conditions do not produce numerical singularities.

[0072] Meshing: The composite structure is discretized in three dimensions using finite element software (such as Ansys-Workbench, Moldflow, or COMSOL), typically employing tetrahedral or hexahedral mesh structures. Adaptive refinement strategies can be used for locally complex regions. The meshing process must adhere to the following logic: denser mesh elements are used in stress concentration areas and regions with large temperature gradients (such as thin-walled areas or corners); while sparser meshes are used in regions with smaller deformation and weaker thermal changes. The goal is to balance simulation accuracy with computational resource consumption.

[0073] Initial energy minimization: After mesh generation, the initial state energy of the system needs to be relaxed to avoid simulation convergence failure due to molecular overlap or unreasonable layout. Energy minimization adopts a molecular force field optimization strategy based on gradient descent or conjugate gradient methods. The commonly used energy functions are Lennard-Jones or COMPASS force field energy functions. The processing logic is as follows: In each iteration step, the interaction energy and potential energy change trend between all molecular units are calculated, and their spatial positions are updated until the total potential energy of the system decreases below the stability threshold or the set number of iterations is reached.

[0074] After the above processing, a "plastic-mold composite structure" with complete physical properties, good simulation conditions, and high structural matching degree is formed. This structure serves as the computational domain basis for subsequent multiphysics coupled simulation analysis, and its construction quality will directly affect the accuracy of behavioral data and the reliability of the optimization model.

[0075] If the copolymerization ratio is unclear when processing ABS raw materials, the glass transition temperature can be deduced by combining DSC experimental data, and then the copolymerization ratio relationship can be fitted by an empirical database for modeling to ensure that the parameter input has engineering usability and verifiability.

[0076] S2: Perform multi-physics coupling simulation on plastic-mold composite structures. The multi-physics fields include thermophysics, fluid physics, and force physics. Combined with the processing path, perform dynamic calculations. The processing processes include injection molding / blow molding / hot pressing. After extracting behavioral data, generate a molding response process model. The behavioral data includes filling speed, holding time, temperature gradient, cooling rate, etc.

[0077] Multiphysics coupled simulations are performed on the plastic-mold composite structure, including thermophysics, fluid physics, and mechanical physics. Dynamic calculations are performed in conjunction with the specific processing path, and key behavioral data are extracted to generate a molding response process model. The core objective of this step is to perform coupled simulations of multiple physical processes involved in actual processing, such as heat conduction, melt flow, and material mechanical response, based on a high-precision plastic-mold composite structure, thereby predicting various molding behavior characteristics and providing quantitative support for subsequent optimization.

[0078] In practical implementation, the first step is to clearly define the processing technology path used, including injection molding, blow molding, or hot pressing. Different process paths will result in significant differences in process timing, thermal boundary conditions, load types, and flow paths. Taking injection molding as an example, the simulation needs to cover the entire lifecycle, including injection, filling, holding pressure, cooling, and demolding; while in hot pressing, the focus is on the influence of the pressure-temperature synergy on the viscoelastic changes of the material. Subsequently, three types of coupled physical fields are applied to the simulation domain:

[0079] Thermophysical Field: Primarily describes the heat exchange between the molten plastic and the mold. The system employs a combined steady-state and transient heat transfer simulation. The processing logic is as follows: within each time step, the heat transfer value between units is calculated based on thermal conductivity, specific heat capacity, and temperature gradient, while also considering interfacial thermal resistance (especially at the plastic-mold contact surface). The heat source includes the high-temperature molten plastic input at the injection screw outlet, and the cold source is the circulating cooling of the mold's water channels.

[0080] Flow Field: This primarily simulates the flow behavior of molten plastic within the molding cavity, requiring consideration of non-Newtonian fluid properties. The Navier-Stokes equations are numerically solved using the control volume method (FVM) or the finite element method (FEM). Viscosity functions are typically calculated using the Cross-WLF or Carreau model. The calculation logic involves dynamically updating the melt viscosity based on local shear rates and temperature at each time step, thereby solving for the velocity and pressure field distributions. For the filling stage, special attention must be paid to the free interface tracking problem at the melt front; commonly used methods include the VOF (Volume of Fluid) method or the Level-Set method.

[0081] Stress / Displacement Field: Used to analyze stress-strain evolution during the molding process, including residual stress distribution caused by cooling shrinkage, pressure transmission within the mold cavity, and product warping tendency. This field is based on the stress balance equations and solved using the constitutive relations of the material (such as linear elasticity, elastoplasticity, or viscoelasticity). The processing logic is as follows: in each iteration, the internal force response is derived from the local deformation state of the material, and the displacement and stress fields are iteratively updated in conjunction with boundary conditions to ultimately determine the degree of structural deformation.

[0082] After completing the three-field coordinated loading, dynamic simulation calculations need to be performed according to the preset process path. The system progresses step by step according to the processing timeline, performing a full-field coupled update at each time step (e.g., 0.01 seconds) and outputting key behavioral data. Common output items include:

[0083] Filling speed: This indicates the rate at which the melt front advances within the mold cavity, reflecting the flow efficiency during the injection phase;

[0084] Pressure response during the holding period: used to evaluate the pressure maintenance and shrinkage compensation of the melt after it has filled the mold cavity;

[0085] Temperature gradient distribution: describes the thermal equilibrium characteristics of different regions during injection molding and cooling processes, and determines stress distribution and degree of crystallization;

[0086] Cooling rate: defined as the rate at which the temperature of the product decreases per unit time, directly affecting internal stress and the risk of demolding deformation;

[0087] Flow front morphology and temperature bandwidth: assess potential defects such as short shot, encapsulation, and bubbles.

[0088] The aforementioned behavioral data is recorded as a data stream during the simulation and linked in real time to the molding cavity spatial coordinate system, outputting as a spatial distribution map or time series, and finally summarized to form a molding response process model. This model is essentially a multi-dimensional behavioral data mapping platform that couples process parameters with molding process characteristics in a logically traceable and structured manner, providing accurate data support for subsequent parameter assignment, response curve fitting, and optimization strategy generation.

[0089] For example, during the injection molding of a thick-bottomed tie rod box, simulations revealed that the cooling rate in the bottom reinforcing rib area was significantly lower than that in the box wall area due to the large wall thickness, leading to stress concentration at the end of the holding pressure period. By adjusting the gate position and mold temperature, the cooling rate can be balanced and deformation reduced. This data-driven process modeling approach is significantly superior to traditional "post-hoc correction" parameter adjustments.

[0090] S3: Based on the molding response process model, quantitative values ​​are assigned to material parameters (such as melt viscosity, specific heat capacity, coefficient of thermal expansion) and process parameters (injection pressure, mold temperature, screw speed, etc.) to form a material-process feature matrix, and the response curves of each feature in the material-process feature matrix to molding defects (such as warpage, shrinkage, bubbles, etc.) are fitted.

[0091] Based on the molding response process model, material parameters (such as melt viscosity, specific heat capacity, and coefficient of thermal expansion) and process parameters (such as injection pressure, mold temperature, and screw speed) are quantitatively assigned to form a material-process feature matrix, and the response curves of each feature to molding defects (such as warpage, shrinkage, and bubbles) are fitted.

[0092] This step is the core data processing step for molding parameter optimization. Its purpose is to establish a structured relationship framework that can be used for prediction and optimization by systematically encoding the original physical properties and actual process operation variables and quantifying their impact on molding defects.

[0093] First, based on the molding response process model generated in step S2, key material and process parameters that are known to affect the response results are extracted. These parameters are divided into two main categories:

[0094] Material parameters (static properties):

[0095] Melt viscosity: describes the flow resistance of a material at a specific shear rate and temperature;

[0096] Specific heat capacity: measures the temperature change caused by the absorption of heat energy per unit mass of material;

[0097] Coefficient of thermal expansion: used to predict internal stress caused by volume contraction during cooling;

[0098] The solid-phase transformation temperature range and thermal decomposition threshold can also be selected.

[0099] Process parameters (dynamic control variables):

[0100] Injection pressure: determines the magnitude of the driving force during the filling process;

[0101] Mold temperature: affects the heat transfer rate at the interface between the melt and the mold;

[0102] Screw speed: Affects the shear rate of the melt in the barrel and the uniformity of preheating;

[0103] Holding time, cooling time, gate location, etc. can also be selected into the parameter set.

[0104] These parameters, after quantization and sampling, are integrated into a material-process feature matrix, the structure of which is similar to a multidimensional factorial experimental table. Each row of the matrix represents a molding configuration combination (obtained through simulation or experiment), and each column represents an independent feature. To ensure sample representativeness, statistical methods such as hierarchical uniform design or Latin hypercube sampling (LHS) are used to traverse and sample the parameter space.

[0105] Based on this, the defect characteristics under each parameter combination are obtained from simulation results or actual molding experiments, including: warpage deformation value: described by the maximum edge displacement of the molded part or the overall deformation rate; shrinkage degree: usually reflected by wall thickness shrinkage rate or volume shrinkage rate; bubble distribution density: statistically analyzing the number of voids per unit area and their maximum diameter. Next, response curve fitting is performed for each feature to each defect. The fitting process can use the following two methods:

[0106] Univariate mapping fitting (suitable for obvious single control relationships):

[0107] The processing logic is as follows: while keeping other parameters fixed, a certain variable is gradually adjusted, the changes in the defect index are recorded, and spline interpolation, polynomial fitting or exponential fitting function is used to perform regression analysis on the results, and finally a one-dimensional response function is formed.

[0108] Multivariate Response Surface Analysis (RSM) (suitable for scenarios with significant interactive effects):

[0109] The processing logic is as follows: a second-order response surface model is constructed based on the results of multi-factor experiments, the coefficients are solved by the least squares method, and a prediction function containing main effect terms, interaction terms and second-order nonlinear terms is established.

[0110] For example, warpage = f1(melt viscosity, mold temperature, holding time), and its processing steps include: normalizing variables → constructing a coefficient matrix → performing regression fitting → extracting regression residuals → verifying the fit (such as R² or sum of squared errors).

[0111] For example, suppose in 10 sets of experiments, it was found that as the mold temperature increased from 50℃ to 70℃, the shrinkage rate gradually decreased from 3.1% to 2.0%, exhibiting a non-linear trend. Within this range, a second-order polynomial fitting function can be used to obtain the mold temperature response curve to shrinkage. Then, combining the interaction term of injection pressure and cooling time, a ternary response surface expression is constructed to provide a more refined evaluation of molding quality. The final output is: each defect term corresponds to a set of response curves or response functions, reflecting the changing trend of each parameter (or parameter combination)'s sensitivity to the defect. These response relationships will be used in the subsequent optimization decision-making stage to select the optimal parameter range combination, achieving a multi-objective trade-off between defect minimization and performance optimization.

[0112] S4: Calculate the unit structural performance index under different process configurations based on the response curve, including volume shrinkage rate and cooling residual stress distribution factor (the volume shrinkage rate and cooling residual stress distribution factor are normalized to the maximum and minimum values ​​so that the value range of volume shrinkage rate and cooling residual stress distribution factor is mapped to [0,1], and the normalized volume shrinkage rate and cooling residual stress distribution factor are summed to obtain the anomaly index). The anomaly index is used to evaluate the molding quality of plastic bags.

[0113] Based on the response curve, the unit structural performance index under different process configurations is calculated, including volume shrinkage rate and cooling residual stress distribution factor. The maximum-minimum value normalization is performed to map the values ​​to the range [0,1]. The normalization results are then summed to obtain the "abnormality index", which is used to comprehensively evaluate the molding quality of plastic bags.

[0114] The core of this step is to transform the functional relationship between each feature and the defect response established in the previous step (S3) into a structured index system that can be used for performance evaluation, and to standardize and integrate these indicators to form a unified quality criterion "abnormal index", which facilitates subsequent horizontal comparisons between different regions and schemes.

[0115] First, based on the simulation results or experimental data of the process parameter combinations, obtain the key performance indicators for the target region or structural unit:

[0116] Volume shrinkage rate: used to measure the dimensional deviation of a molded plastic part caused by volume change during the cooling and solidification process. The calculation logic is as follows: extract the volume data of the product at the initial injection (when the mold cavity is full) and at the moment of demolding from the simulation data; for each simulation sample, calculate its volume shrinkage ratio, that is, the ratio of the difference between the volume before shrinkage and the volume after shrinkage to the original volume, as the volume shrinkage rate index.

[0117] Cooling residual stress distribution factor: This index is used to characterize the internal stress gradient caused by non-uniform cooling, thereby reflecting the structural stability and warpage risk of the product. Its processing logic is as follows: extract the stress distribution data of the residual stress field in the target structural region after simulation; use the regional standard deviation method as the cooling residual stress distribution factor (i.e., select multiple measurement points in the region, obtain the cooling residual stress at each measurement point, and calculate the cooling residual stress standard deviation as the cooling residual stress distribution factor).

[0118] Based on the response curve, the unit structural performance indicators (volume shrinkage rate, cooling residual stress distribution factor) corresponding to each combination of process parameters are quantified, and the anomaly index (E) is formed by summing after normalization, which is used to comprehensively evaluate the molding quality of plastic bags.

[0119] Volume shrinkage rate (S): , among which, among which, The volume when the mold filling is complete. This represents the volume after cooling and demolding.

[0120] Cooling residual stress distribution factor : ,in, Let be the stress value at the i-th stress measurement point within a certain structural region. The number of stress measurement points, The average stress

[0121] After extracting the above indicators, in order to ensure comparability between different process configurations, these two indicators need to be uniformly normalized to map them to a fixed interval ([0,1]). The specific processing logic is as follows:

[0122] For a specific indicator (such as volume shrinkage rate), determine the maximum (max) and minimum (min) values ​​among all samples. For each sample's current indicator value X, perform linear normalization: normalized value = (X - min) / (max - min). Perform the same operation on the cooling residual stress distribution factor to ensure that both indicators are between 0 and 1. At this point, the normalized volume shrinkage rate and normalized stress distribution factor corresponding to each process configuration have been obtained.

[0123] Next, the two normalized values ​​are summed and defined as the "anomaly index," which serves as a unified quality evaluation indicator. The processing logic is as follows: Anomaly Index = Normalized Volumetric Shrinkage Rate + Normalized Residual Stress Distribution Factor; the larger the anomaly index, the more severe the deformation or stress anomaly exists under this process configuration, and the worse the overall molding quality; the smaller the anomaly index, the better the dimensional stability of the product and the more uniform the structural stress distribution, indicating a high-quality molding parameter combination.

[0124] This evaluation method can not only unify two performance indicators with different physical mechanisms but related results into the same evaluation system, but also be used for rapid horizontal comparison between different structural regions and different parameter configurations to identify performance bottleneck areas or process paths with great optimization potential.

[0125] For example, in a simulation, the normalized values ​​of volume shrinkage rate and stress factor for three configuration samples were (0.8, 0.6), (0.5, 0.4), and (0.2, 0.1), respectively, with corresponding anomaly indices of 1.4, 0.9, and 0.3. Clearly, the third group represents the optimal molding configuration, exhibiting uniform stress distribution and minimal volume change, demonstrating high manufacturing stability. The final output is a set of structured data, where each combination of molding parameters corresponds to an anomaly index. This data can be used for sorting, filtering, and grading, or as a constraint input into subsequent multi-objective optimization algorithms to achieve quantitative decision support.

[0126] S5: Based on the design zones of plastic bags (such as handles, corners, bottom, etc.), the unit structural forming parameters (used to generate weighting coefficients) and performance indicators (abnormality index) of each zone are statistically analyzed. An overall performance mapping model is constructed through regional weighting (abnormality index weighting of all zones) to provide basic data support for the overall bag forming optimization.

[0127] Based on the design zones of plastic bags (such as handles, corners, bottom, etc.), the unit structural forming parameters (used to generate weighting coefficients) and performance indicators (abnormality index) of each zone are statistically analyzed. An overall performance mapping model is constructed through a zone weighting method to provide basic data support for the overall bag forming optimization.

[0128] The core of this step lies in extending the single-point or local structural performance indicators (especially the anomaly index) established in the previous step to the entire product dimension. This ensures that each parameter configuration not only corresponds to the quality performance of a specific area but also reflects the overall molding performance of the entire plastic bag. This regional weighting strategy is key to establishing a "local-whole" mapping relationship, and is particularly suitable for bag products with complex structures and significantly different performance requirements.

[0129] First, the plastic bags need to be functionally divided into zones. Common zones include, but are not limited to, the following typical areas: handle connection area (bearing tensile force, often a stress concentration point).

[0130] The following areas are considered: corner transition zones (where wall thickness changes drastically and warping is likely); bottom and wheel support areas (where the stress area is large and the stress duration is long); and the middle area of ​​the wall (a typical stress area where cooling uniformity is important). For each area, two key variables need to be extracted during simulation or actual testing:

[0131] Unit Configuration Parameter (UCP): This parameter represents the forming sensitivity of a region in terms of geometry, dimensions, wall thickness, and structural complexity, and is used to generate region weighting coefficients. The specific statistical method is as follows: Calculate the wall thickness change rate for each region; sum the wall thickness change rates of all regions to obtain the overall wall thickness change rate of the plastic bag; divide the wall thickness change rate of a region by the overall wall thickness change rate of the plastic bag to obtain the weight coefficient for that region (this step normalizes the UCP values ​​of all regions to ensure that the sum of all region weights is 1).

[0132] Regional performance index (anomaly index E): Obtained in step S4, reflecting the structural quality performance of this region under the current process configuration. A higher anomaly index indicates worse quality. Next, a global performance mapping model (GPM) is constructed, with the following processing logic:

[0133] Define a "global anomaly index" for the entire plastic case under the current process parameter configuration. Its processing logic is: Global_E = Σ(wi × Ei), where Ei is the anomaly index of region i, wi is the corresponding weight coefficient of region i, and Global_E represents the global anomaly index. This calculation is equivalent to weighted integration of the quality performance of all local structures, forming a scalar function that comprehensively reflects the overall quality. Bind the above global anomaly index to the current input parameter configuration (i.e., material-process combination); construct a "input parameter → global anomaly index" response function to form the overall performance mapping model GPM.

[0134] For example, suppose in a simulation, the handle area w1=0.3, the bottom area w2=0.4, the corner area w3=0.2, and the wall area w4=0.1; the corresponding anomaly indices are 0.6, 0.8, 0.3, and 0.2 respectively. Then, after weighted calculation, the global anomaly index = 0.3×0.6 + 0.4×0.8 + 0.2×0.3 + 0.1×0.2 = 0.62. This value can be regarded as the global anomaly index of the entire bag under this parameter configuration.

[0135] The overall performance mapping model established by the above methods can not only perform global quality ranking for different combinations of process parameters, but also analyze which regions have the greatest impact on the overall quality of the part. This allows for the focus on controlling the parameters of key regions in subsequent optimization, thereby improving the accuracy and efficiency of molding control.

[0136] S6: Based on the performance indicators of each region and the overall performance mapping model, determine whether the molding parameters of plastic bags need to be optimized, and generate corresponding optimization strategies for the molding parameters of plastic bags based on the judgment results. (If the optimization of the molding parameters of plastic bags is determined based on the performance indicators of the region (e.g., the abnormal index of at least one partition is greater than the abnormal threshold) (mainly reflected in the substandard processing quality), then one optimization strategy for the molding parameters of plastic bags is generated. If the optimization of the molding parameters of plastic bags is determined based on the overall performance mapping model (i.e., the abnormal index of each partition is less than or equal to the abnormal threshold, but after comprehensive weighted analysis of the abnormal index of all partitions, it is determined that the overall performance is substandard (mainly reflected in the reduced service life of plastic bags in the later stage)), then another optimization strategy for the molding parameters of plastic bags is generated.)

[0137] The goal of this step is to determine whether the current molding process parameters meet production and performance requirements based on the regional and global anomaly indices established in previous steps. If quality or structural defects exist, a differentiated optimization strategy generation mechanism is triggered through different judgment paths to achieve targeted improvement in molding performance. The entire process mainly includes two entry points: regional anomaly identification and judgment, and overall performance evaluation and judgment. Based on the judgment results, corresponding parameter optimization strategies are constructed. The specific process is as follows:

[0138] Identify localized quality defects caused by severe flaws in individual structural zones (such as warping, air bubbles, shrinkage, etc.), typically manifested as functional or aesthetic failures to meet factory standards. Set a preset regional anomaly threshold, which can be determined based on historical simulation statistics or engineering experience (e.g., 0.6 or 0.7). For all zones i, check whether their anomaly index meets the following conditions:

[0139] If the anomaly index of at least one region exceeds the region anomaly threshold, it is determined to be a local performance failure. This region is marked as a critical fault region, and local optimization of process parameters is prioritized. The optimization target focuses on the rapid suppression of explicit defects. A local reinforcement adjustment mechanism is enabled, that is, only a few core process parameters that have a significant impact on the anomaly of this region are adjusted (such as adjusting mold temperature, injection pressure, local cooling channel layout, etc.). The optimization algorithm recommends using a sensitivity-driven parameter adjustment method. First, the partial derivatives of each process parameter with respect to the target defect are calculated based on the existing response curves, and the two variables with the greatest impact are selected for fine-tuning.

[0140] Although each area appears "qualified" on the surface, the system identifies a systemic risk that may lead to a reduced service life in the later stages of the bag due to problems such as uneven distribution of the overall molding process, latent residual stress, and poor structural coordination. The judgment condition is: the global anomaly index is greater than the global threshold. The global threshold is usually set at the upper limit of the average quality standard of the entire product (such as 0.5 or 0.6), indicating that the overall manufacturing stability and mechanical coordination have reached the lower limit. At this time, the system judges it as a global quality risk. Even if the current finished product has no appearance or structural defects, there may still be hidden dangers such as fatigue cracks, stress concentration, and creep failure during long-term use.

[0141] The optimization objectives focus on "potential quality improvements" such as structural distribution, cooling equilibrium, and stress coordination. A global fine-tuning strategy mechanism is activated to comprehensively re-optimize all influencing parameters. A multi-objective optimization method (such as NSGA-II) is adopted, and the optimization objectives include: making the global anomaly index less than or equal to the global threshold, limiting the anomaly index of any region to not exceed the regional anomaly threshold, and controlling the parameter change range to not exceed 5% of the historical configuration (to prevent large jumps from causing new problems). The overall performance mapping model is used as the objective function evaluator to select the global optimal solution or Pareto front solution from multiple sets of candidate parameter configurations.

[0142] Once the optimization strategy is generated, the system will output a "recommended parameter combination" and record the corresponding predicted regional anomaly index and global anomaly index. If the recommended result still does not meet the set criteria (i.e., after iterative optimization, the global anomaly index is still greater than the global threshold or the regional anomaly index is greater than the regional anomaly threshold), the next round of optimization will be automatically triggered. The system can set the maximum number of iterations or the minimum improvement rate as the stopping condition. Finally, the optimization history is retained for manufacturing engineers to compare, analyze, and make decisions.

[0143] In the parameter optimization process of plastic bag molding, to improve overall manufacturing quality and structural reliability, the system needs to determine whether the current process configuration meets the standards based on the performance indicators of each region and the overall performance mapping model. If local or global quality risks are detected, the corresponding optimization strategy will be automatically triggered. The following uses a set of specific numerical scenarios to illustrate the execution process of the entire optimization strategy.

[0144] First, the system's optimization objectives include three constraints: First, the global anomaly index (Global_E) must be less than or equal to 0.52; second, the regional anomaly indices (E1, E2, E3, etc.) for all design partitions must not exceed 0.65; and third, the adjustment range of process parameters must not exceed ±5% of the historical process configuration to prevent new manufacturing risks caused by drastic parameter fluctuations. Simultaneously, the system employs a multi-objective optimization algorithm (such as NSGA-II) to perform parameter search and uses the overall performance mapping model as the objective function evaluator to calculate the quality performance under each candidate parameter combination.

[0145] In the initial state, the process parameters for a certain bag are configured as follows: injection pressure 90MPa, mold temperature 65℃, holding time 6.0 seconds, and cooling time 12.0 seconds. After system evaluation, the regional anomaly indices for each design zone are obtained: handle area 0.58, corner area 0.54, bottom area 0.61, and wall area 0.50. Combining the regional weight coefficients of 0.25, 0.25, 0.30, and 0.20, the global anomaly index under the current process is calculated as follows:

[0146] Global_E = 0.25 × 0.58 + 0.25 × 0.54 + 0.30 × 0.61 + 0.20 × 0.50 = 0.5685. Since 0.5685 is greater than the set global threshold of 0.52, the system determines that this process configuration has a global quality risk and a global optimization strategy needs to be triggered.

[0147] During the optimization process, the system generates multiple candidate process combinations. For example, a new configuration might be: injection pressure 87 MPa, mold temperature 68℃, holding time 6.3 seconds, and cooling time 12.4 seconds. After using the performance mapping model for prediction, the anomaly indices for each region are obtained: handle region 0.50, corner region 0.48, bottom region 0.52, and wall region 0.45. The global anomaly index is then recalculated.

[0148] Global_E = 0.25 × 0.50 + 0.25 × 0.48 + 0.30 × 0.52 + 0.20 × 0.45 = 0.492. This result indicates that the global anomaly index has decreased to 0.492, below the 0.52 threshold, and the anomaly indices for all regions do not exceed 0.65. Furthermore, the adjustment range of all parameters is controlled within ±5% of historical values. Therefore, the system confirms that this process configuration is the recommended combination that meets the quality objectives and outputs this combination and the corresponding predicted quality indicators.

[0149] If, in a candidate solution, Global_E has decreased to 0.545, but the anomaly index in the bottom region reaches 0.67, exceeding the region threshold of 0.65, then the current solution is considered unqualified, and the system will automatically proceed to the next round of optimization iteration. This iterative process will continue until all set constraints are met or the system-defined termination conditions are reached, such as the maximum number of iterations or a performance improvement rate of less than 1% for two consecutive rounds.

[0150] The logic for obtaining the regional anomaly threshold is as follows: collect a large number of historical simulation samples and actual product test results, and extract the anomaly index of each design area (such as handles, corners, etc.) under different process configurations; plot the distribution of all anomaly indices of the area (such as histograms or cumulative distribution functions); mark the samples that correspond to "unqualified" or "rework" in actual testing; use the anomaly index range of these unqualified samples as a reference to determine the anomaly judgment threshold; usually, the maximum anomaly index corresponding to more than 95% of qualified samples is selected as the regional anomaly threshold.

[0151] The global anomaly threshold is obtained as follows: the global anomaly index of historical samples is calculated by weighting the anomaly index of all design partitions, the global anomaly index of all process configurations is normalized and distributed, a confidence interval is set according to the bag pass rate (e.g., 95% pass rate), and the corresponding global anomaly value is taken as the threshold. If there are 200 sets of historical data, and the weighted global anomaly index is concentrated in the range of 0.30 to 0.66, and the global anomaly index of 95% of qualified products does not exceed 0.52, then the global anomaly threshold is set to 0.52.

[0152] Ultimately, the system will not only output recommended parameter combinations, but also record parameter changes and corresponding predicted anomaly indicators for each round of optimization, providing manufacturing engineers with data for source analysis, solution comparison, and technical decision-making. This optimization strategy deeply couples process parameter control with structural performance requirements, improving the intelligence and precision of the molding process.

[0153] In the initial simulation of a certain bag, the E values ​​for each region are 0.55 for the handle, 0.60 for the corner, 0.58 for the bottom, and 0.52 for the wall, with a Global_E of 0.61. Although each individual E value does not exceed the local threshold of 0.65, the overall anomaly index is higher than the set upper limit of 0.6. Therefore, this sample will trigger Strategy B (global optimization strategy). The system will recommend adjusting the holding time and cooling time to improve cooling balance and reduce overall residual stress. This judgment and strategy system not only achieves dual control over "short-term production defects" and "long-term structural risks," but also makes the optimization logic hierarchical and traceable, providing a stable and reliable decision support mechanism for the intelligent manufacturing of high-quality plastic bags.

[0154] Example 3: Obtaining Example Material Parameters (Taking Polypropylene (PP) as an Example):

[0155] Obtain the molecular structure model of PP segments using materials modeling databases (such as Materials-Project) or quantum chemistry software (such as Materials-Studio), with parameters including:

[0156] Repeating unit: –CH2–CH(CH3)–;

[0157] Side group structure: methyl (–CH3);

[0158] Conformation state: Assumed to be isotactic polypropylene (iPP);

[0159] Molecular weight distribution: Mn=120,000, Mw=190,000;

[0160] Specific heat capacity (reference or DSC test): c = 1.9 J / g·K;

[0161] Melt viscosity: η = η0 / [1 + (λγ̇)²] n Where η0 is the zero-shear viscosity, λ is the relaxation time, and γ̇ is the shear rate. The viscosity function is modeled using the Cross-WLF model for subsequent flow behavior.

[0162] The mold structure is imported into the simulation platform (such as ANSYS or Moldflow) in .STEP format. The PP molecular configuration is mapped to the mold cavity region in three dimensions. A point cloud fitting algorithm is used for spatial fusion. The calculation logic is as follows:

[0163] Input: A set of molecular particles {P1, P2, ..., P} n} and the set of cavity surfaces {Q1,Q2,...,Q n};

[0164] Execution: Minimize the objective function E=Σ||Pi−Qi||², that is, optimize the molecular arrangement through least squares fitting;

[0165] Output: The location of the fitted molecular filling region.

[0166] Define the following boundaries based on simulation requirements:

[0167] Thermal boundary: The mold cooling circuit is set to a temperature of T1=40℃, and the feed inlet is set to T2=220℃;

[0168] Flow boundary: The inlet is set to a constant pressure P_in=80MPa, and the outlet is allowed to vent freely;

[0169] Displacement boundary: The mold mating surface is fixed in the z-axis direction to avoid overall rigid body drift.

[0170] Moldflow was used to generate a tetrahedral mesh for the mold cavity, with the following parameter settings:

[0171] Thin-walled region (corner reinforcement): mesh density h1 = 0.2 mm;

[0172] Central region (planar): Mesh density h2 = 0.5 mm;

[0173] Local refinement strategy: When the temperature gradient ∇T > 30℃ / mm, the local refinement is h1.

[0174] Ensure the total number of grid cells is controlled at 2×10. 6 Within a single unit, balance accuracy with computational burden.

[0175] In the example above, the constructed "plastic-mold composite structure" can serve as the input domain for high-confidence simulations. In subsequent simulations, the system can further predict the filling rate, residual stress, and cooling efficiency, thereby deriving the optimal molding parameters.

[0176] Example 4: A material-process feature matrix is ​​constructed by sampling, and the response relationship of each variable to molding defects (such as warpage, shrinkage, bubbles, etc.) is fitted for subsequent parameter optimization and molding control. The parameters to be studied are shown in Table 1:

[0177] Table 1: Research Parameters

[0178]

[0179] We used Latin hypercube sampling (LHS) to generate 10 sample combinations, resulting in the feature matrix shown in Table 2:

[0180] Table 2: Feature Matrix

[0181]

[0182] Using Moldflow or actual injection molding experiments, the molding defect values ​​corresponding to each sample were obtained. Typical defect indicators include: Shrinkage (shrinkage rate, unit %), Warp (warpage value, unit mm), and VoidDensity (bubble density, unit count / cm²).

[0183] Example 5: The following is a detailed description and example of the process for step S2, including the three-field coupling simulation content, data extraction logic and related calculation output format. Finally, the spatial and temporal evolution results of the simulation behavior data are shown in the example table.

[0184] Thermophysical field:

[0185] Objective: To describe the heat transfer behavior between molten plastic and mold;

[0186] Input parameters: material thermal conductivity, mold wall temperature, feed temperature;

[0187] Processing logic: Within each time step Δt, the amount of heat migration is calculated using the thermal diffusion model, and the temperature gradient is output.

[0188] Flow-Field:

[0189] Objective: To simulate the non-Newtonian flow state of melt within the mold cavity;

[0190] Input parameters: viscosity model (e.g., Carreau-Yasuda), shear rate, inlet pressure;

[0191] Processing logic: The finite volume method is used to solve the continuity equation and the momentum conservation equation, and the shape of the frontal interface is tracked.

[0192] Stress / Displacement Field:

[0193] Objective: To predict the residual stress and displacement distribution of the product during the molding process;

[0194] Input parameters: material elastic modulus, coefficient of thermal expansion, cooling rate;

[0195] Processing logic: Iteratively calculate the evolution of internal stress based on the stress balance equation and constitutive relation.

[0196] Simulation path steps:

[0197] Processing technology path: Injection molding;

[0198] Timing stages: Injection → Filling → Holding pressure → Cooling → Demolding;

[0199] Time step control: Perform a three-field fully coupled calculation every 0.01s;

[0200] Solver: The convergence process is performed using an iterative coupled thermal-fluid-mechanical solver.

[0201] Behavioral data extraction items:

[0202] Filling speed: the rate of advance of the flow front, in mm / s;

[0203] Pressure holding response pressure: Maintaining the average pressure within the stage cavity;

[0204] Temperature gradient: The range of temperature change at a specified cross-section;

[0205] Cooling rate: The rate at which different areas of a product cool down per unit time;

[0206] Flow front morphology: The shape parameters of the fluid front at key cross-sections (such as closure, asymmetry, etc.).

[0207] Key simulation behavioral data are shown in Table 3 (partial):

[0208] Table 3: Simulation Key Behavior Data

[0209]

[0210] All the above behavioral data are bound to the coordinate grid of the mold cavity; the spatial distribution map shows the physical fields such as temperature, stress, and flow rate through structural cross-section extraction (XY or XZ cross-section); the time series data constitutes the core input of the molding response process model and can be used for subsequent feature matrix assignment and optimization objective function evaluation.

[0211] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0212] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. An intelligent optimization method for forming parameters of plastic luggage, characterized in that: The optimization method comprises the following steps: S1: Obtain the molecular structure parameters of the plastic raw material, load the molecular structure parameters into the molding cavity framework formed by the target luggage mold structure, and obtain a plastic-mold composite structure after pretreatment; S2: Perform multi-physical field coupling simulation on the plastic-mold composite structure, perform dynamic calculation combined with the processing path, and generate a molding response process model after extracting the behavior data; S3: According to the molding response process model, quantitatively assign material parameters and process parameters to form a material-process feature matrix, and fit the response curves of each feature in the material-process feature matrix to the molding defects; S4: Calculate the unit structure performance index under different process configurations based on the response curve, which is used to evaluate the molding quality of the plastic luggage; S5: According to the design partition of the plastic luggage, the unit structure molding parameters and performance indicators of each region are counted, and the overall performance mapping model is constructed through regional weighting; S6: According to the performance indicators of each region and the overall performance mapping model, it is judged whether the plastic luggage molding parameters need to be optimized, and the corresponding plastic luggage molding parameter optimization strategy is generated according to the judgment result; In step S5: According to the design partition of the plastic luggage, the unit structure molding parameters and performance indicators of each region are counted, and the overall performance mapping model is constructed through regional weighting, comprising the following steps: Design partition of the luggage, including handle connection area, corner transition area, bottom and wheel seat support area, wall middle area; For each design partition, the unit structure molding parameters and performance indicators of the region are extracted, respectively, the unit structure molding parameters are used to generate regional weight coefficients, and the performance indicators are abnormal indexes used to reflect the quality performance of the region under the process configuration; Weighted sum of the abnormal indexes of all regions and the corresponding normalized weight coefficients to obtain the global abnormal index under the current process parameter configuration; Bind the global abnormal index to the current material parameter and process parameter combination to construct the response relationship between the input parameters and the global abnormal index, and generate the overall performance mapping model; In step S6: According to the performance indicators of each region and the overall performance mapping model, it is judged whether the plastic luggage molding parameters need to be optimized, and the corresponding plastic luggage molding parameter optimization strategy is generated according to the judgment result, comprising the following steps: Based on the performance indicators of each region and the overall performance mapping model, it is judged whether the molding parameters of the plastic luggage need to be optimized, and the corresponding molding parameter optimization strategy is generated according to the judgment result; If there is at least one region whose abnormal index is greater than the regional abnormal threshold, it is determined that the local performance is unqualified, and the local optimization strategy is triggered; The local optimization strategy includes: identifying the region whose abnormal index is greater than the regional abnormal threshold, calculating the partial derivative of the abnormal related parameters of the region based on the response curve, and selecting at least two process parameters for local optimization according to the partial derivative; Under the premise that all regional abnormal indexes are less than or equal to the regional abnormal threshold, if the global abnormal index exceeds the global threshold, it is determined that the overall quality risk exists, and the global optimization strategy is triggered; The global optimization strategy includes: jointly optimizing all process parameters that affect the global anomaly index, using a multi-objective optimization algorithm to constrain and optimize multiple objectives. 2.The intelligent optimization method for forming parameters of plastic luggage according to claim 1, characterized in that: In step S3: according to the forming response process model, the material parameters and process parameters are quantitatively assigned to form a material-process characteristic matrix, and the response curves of each characteristic in the material-process characteristic matrix to the forming defects are fitted, including the following steps: Quantitative sampling of the material parameters and process parameters, using hierarchical uniform design or Latin hypercube sampling to construct a material-process characteristic matrix containing multiple forming combinations, wherein each row of the matrix corresponds to a forming configuration combination, and each column corresponds to a characteristic variable; Extract the defect performance corresponding to each forming configuration combination, and take the defect indicators in the defect performance as response variables; For each characteristic variable and corresponding defect indicator, perform response curve fitting operation. 3.The intelligent optimization method for forming parameters of plastic luggage according to claim 2, characterized in that: The material parameters include melt viscosity, specific heat capacity, thermal expansion coefficient, solid phase transition temperature range, the process parameters include injection pressure, mold temperature, screw speed, holding time, cooling time and gate location, and the defects include warpage deformation value, shrinkage degree and bubble distribution density.

4. The intelligent optimization method for molding parameters of plastic luggage according to claim 1, characterized in that: In step S4: based on the response curve, the unit structure performance index under different process configurations is calculated to evaluate the forming quality of the plastic luggage, including the following steps: The performance indicators include volume shrinkage rate and cooling residual stress distribution factor; The volume shrinkage rate and cooling residual stress distribution factor are normalized by maximum and minimum value respectively; Sum the normalized volume shrinkage rate and cooling residual stress distribution factor to get the anomaly index, which is used to evaluate the forming quality level under the current process configuration; The larger the anomaly index, the greater the size deviation or structural stress anomaly of the product under the corresponding process parameters, and the worse the forming quality; The smaller the anomaly index, the higher the forming dimensional stability and the uniform stress distribution, and the better the forming quality.

5. The intelligent optimization method for forming parameters of plastic luggage according to claim 4, characterized in that: The calculation logic of the volume shrinkage rate is: extract the volume values of the molded part at the completion of the mold cavity filling and at the demolding moment in the simulation data, calculate the ratio of the volume difference value to the original volume as the shrinkage rate indicator; The calculation logic of the cooling residual stress distribution factor is: select multiple measurement points in the target structure area, obtain the residual stress value of each measurement point, and calculate the standard deviation of the cooling residual stress in the area as the stress distribution factor.

6. The intelligent optimization method for forming parameters of plastic luggage according to claim 5, characterized in that: The unit structure forming parameter is the wall thickness variation rate, including calculating the wall thickness variation rate value of each region, summing the wall thickness variation rates of all regions, and taking the ratio of the regional wall thickness variation rate to the overall wall thickness variation rate as the weight coefficient of the region.

7. The intelligent optimization method for molding parameters of plastic luggage according to claim 1, characterized in that: In step S1: obtain the molecular structure parameters of the plastic raw material, load the molecular structure parameters into the forming cavity frame formed by the target luggage mold structure, and obtain the plastic-mold composite structure after preprocessing, including the following steps: Obtain the molecular structure parameters of the plastic raw material, the plastic raw material includes at least one of polypropylene, acrylonitrile-butadiene-styrene copolymer; Molecular structure parameters including connection mode of monomer repeating units, segment length, side group structure, polarity distribution, molecular mass distribution range and molecular chain conformation are obtained by a molecular modeling tool; The molecular structure parameters are loaded into a molding cavity framework formed by a target luggage mold structure, the target luggage mold structure is input in a CAD format and formed into an assembled state through structure import; After preprocessing the molding cavity framework, the preprocessing includes geometric fitting, boundary setting and mesh division, forming a plastic-mold composite structure. 8.The intelligent optimization method for forming parameters of plastic luggage according to claim 1, characterized in that: In step S2, a multi-physical field coupling simulation is performed on the plastic-mold composite structure, a dynamic calculation is performed in combination with a processing technology path, and a molding response process model is generated after behavior data is extracted, including the following steps: The multi-physical field includes thermal physical field, flow physical field and force physical field; The thermal physical field is used to simulate the heat exchange process between the plastic melt and the mold; The flow physical field is used to simulate the non-Newtonian flow behavior of the melt in the mold cavity; The force physical field is used to analyze the stress and displacement distribution in the molding process; In combination with the processing technology path in injection molding, blow molding or hot pressing, a dynamic simulation process including injection, filling, pressure maintaining, cooling and demolding is executed; Full-field coupling operation is performed at multiple time steps, and the filling velocity, pressure response in the pressure maintaining period, temperature gradient distribution, cooling rate and flow front morphology are extracted; The behavior data is associated with the molding cavity spatial coordinate system, and output is a spatial distribution map or a time series data stream, and a molding response process model is generated.

Images