Apparatus for estimating molded product quality variation, method for estimating molded product quality variation, and injection molding system

Abstract

To estimate quality variation of a molding when a material with an unstable physical property is employed.SOLUTION: A molding quality variation estimation device comprises: a physical property calculation model estimation unit for estimating a physical property calculation model corresponding to a second material on the basis of process data when manufacturing a molding using a first material with known physical properties, a learned regression model generated based on a physical property calculation model coefficient corresponding to the first material, and process data when manufacturing the molding using the second material with unknown physical properties; and a flow analysis unit for performing flow analysis on the basis of the physical property calculation model corresponding to the estimated second material to estimate quality variation of the molding using the second material. The physical property calculation model estimation unit estimates two or more physical property calculation models each corresponding to respective two or more different physical properties. The flow analysis unit performs the flow analysis on the basis of the estimated two or more physical property calculation models.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a molded product quality variation estimation device, a molded product quality variation estimation method, and an injection molding system.

Background Art

[0002] In the field of injection molding, in response to the demand for social resource circulation, the switch from virgin materials such as synthetic resins to recycled materials with low environmental impact is being promoted. However, since recycled materials are made from various plastic products with different thermal histories, even if they are shipped with the same product number from a recycled material supplier, when comparing different lots at the time of shipment, their physical properties are not necessarily the same and are not stable.

[0003] The instability of the physical properties of the material can affect the quality of the molded product. In particular, among the physical properties of the material, the pressure-temperature dependence of viscosity and specific volume (hereinafter referred to as PVT characteristics) is known to cause variations in the dimensions of the molded product.

[0004] When injection molding is performed using newly adopted materials with different physical properties such as viscosity and PVT characteristics, in order to assist in the confirmation of moldability and the adjustment of molding conditions, a flow analysis using CAE (Computer Aided Engineering) is performed. The method of using flow analysis in the design and material selection of plastic parts is common. However, for flow analysis, it is necessary to create a physical property calculation model based on the physical property data of the material, but it is not realistic to create a physical property calculation model corresponding to all lots of recycled materials because the amount of work becomes enormous.

[0005] Regarding fluid analysis, for example, Patent Document 1 describes a "fluid analysis apparatus for performing fluid analysis of molten material in the injection molding process, characterized by comprising a material property data calculation means for calculating viscosity data as material property data based on data obtained from an injection molding machine, and a fluid analysis means for performing fluid analysis based on an analytical shape model, material property data, and molding condition data." [Prior art documents] [Patent Documents]

[0006] [Patent Document 1] Japanese Patent Publication No. 2003-145577 [Overview of the project] [Problems that the invention aims to solve]

[0007] In the case of the flow analysis apparatus described in Patent Document 1, since flow analysis is performed using only viscosity as a material property, the analysis accuracy is lower compared to cases where multiple material properties are used for flow analysis. Therefore, for example, it is not possible to appropriately estimate the quality variation of molded products when multiple materials with the same product number but different lots are used.

[0008] This invention has been made in view of the above points, and aims to improve the accuracy of flow analysis when using materials with unstable physical properties, and to enable the estimation of variations in the quality of molded products. [Means for solving the problem]

[0009] This application includes several means to solve at least some of the above problems, and some examples are as follows.

[0010] To solve the above problems, a molded product quality variation estimation device according to one aspect of the present invention is a molded product quality variation estimation device for estimating the quality variation of a molded product by injection molding, comprising: a physical property calculation model estimation unit that estimates a physical property calculation model corresponding to a second material based on process data when a molded product is manufactured using a first material with known physical properties, a learned regression model generated based on physical property calculation model coefficients corresponding to the first material, and process data when a molded product is manufactured using a second material with unknown physical properties; and a flow analysis unit that estimates the quality variation of a molded product when a molded product is manufactured using the second material by performing a flow analysis based on the estimated physical property calculation model corresponding to the second material, wherein the physical property calculation model estimation unit estimates two or more physical property calculation models corresponding to two or more different physical properties based on two or more learned regression models corresponding to two or more different physical properties, and the flow analysis unit performs the flow analysis based on the estimated two or more physical property calculation models. [Effects of the Invention]

[0011] According to the present invention, the accuracy of flow analysis can be improved when using materials with unstable physical properties, and it becomes possible to estimate the variation in the quality of molded products.

[0012] Other issues, configurations, and effects not mentioned above will be clarified by the following description of the embodiments. [Brief explanation of the drawing]

[0013] [Figure 1] Figure 1 shows examples of the physical properties of multiple recycled materials with the same product number but different lots. Figure 1(A) shows the PVT properties, and Figure 1(B) shows the viscosity. [Figure 2] Figure 2 shows an example of an injection molding system according to an embodiment of the present invention. [Figure 3] Figure 3 shows an example of the configuration of a functional block in a molded product quality variation estimation device. [Figure 4]FIG. 4 is a diagram showing an example of sensor information. [Figure 5] FIG. 5 is a diagram showing an example of a feature quantity DB (database). [Figure 6] FIG. 6 is a diagram showing an example of a quality DB. [Figure 7] FIG. 7 is a diagram showing an example of a selected learning data set. FIG. 7(A) is a diagram showing a learning data set for PVT characteristics, and FIG. 7(B) is a diagram showing a learning data set for viscosity. [Figure 8] FIG. 8 is a diagram showing a configuration example of a PVT characteristic calculation model estimation unit and a viscosity calculation model estimation unit. [Figure 9] FIG. 9 is a diagram showing a configuration example of a general computer. [Figure 10] FIG. 10 is a cross-sectional view showing a configuration example of an injection molding machine. [Figure 11] FIG. 11 is a diagram showing an example of a mold. FIG. 11(A) is a top view of the product part of the mold, FIG. 11(B) is a side view of the product part of the mold, FIG. 11(C) is a top view of the runner part of the mold, and FIG. 11(D) is a side view of the runner part of the mold. [Figure 12] FIG. 12 is a flowchart for explaining an example of a molding process data acquisition process. [Figure 13] FIG. 13 is a diagram for explaining the effectiveness of an estimated physical property calculation model. FIG. 13(A) is a diagram showing an example of an estimated PVT characteristic calculation model, and FIG. 13(B) is a diagram showing an example of an estimated viscosity calculation model. [Figure 14] FIG. 14 is a diagram showing a display example of a UI (User Interface) screen.

Embodiments for Carrying Out the Invention

[0014] Hereinafter, embodiments of the present invention will be described with reference to the drawings. The embodiments are examples for explaining the present invention, and for the sake of clarity of explanation, appropriate omissions and simplifications have been made. The present invention can be implemented in various other forms. Unless otherwise particularly limited, each component may be singular or plural. In the drawings, the positions, sizes, shapes, ranges, etc. of each component shown may not represent the actual positions, sizes, shapes, ranges, etc. in order to facilitate understanding of the invention. Therefore, the present invention is not necessarily limited to the positions, sizes, shapes, ranges, etc. disclosed in the drawings. As examples of various information, it may be described in expressions such as "table", "list", "queue", etc., but various information may be represented by other data structures. For example, various information such as "XX table", "XX list", "XX queue" may be referred to as "XX information". When explaining identification information, expressions such as "identification information", "identifier", "name", "ID", "number", etc. are used, but these are mutually replaceable. In all the drawings for explaining the embodiments, the same members are generally given the same reference numerals, and repeated explanations thereof are omitted. Also, in the following embodiments, the components (including element steps, etc.) are not necessarily essential unless specifically stated or considered to be clearly essential in principle. Also, when saying "consisting of A", "comprising A", "having A", "including A", other elements are not excluded unless it is specifically stated that only that element is meant. Similarly, in the following embodiments, when referring to the shapes, positional relationships, etc. of components, etc., unless specifically stated or considered not to be so in principle, those substantially approximating or similar to the shapes, etc. are included.

[0015] <Regarding the physical properties of recycled materials> Figure 1 shows examples of the physical properties of a plurality of recycled materials that are materials for molded products. The plurality of recycled materials are recycled materials A, B, and C with the same product number but different shipment lots (for example, different delivery times).

[0016] Figure (A) shows the PVT properties of recycled materials A, B, and C, with the horizontal axis representing the material temperature [°C] and the vertical axis representing the material specific volume [cm³]. 3 The value is [ / g]. The material pressure is assumed to be a constant value of 50 [MPa]. The point plot shown in Figure (A) represents measured values ​​obtained using a dedicated measuring device. The line plot shows the results of fitting a known calculation model, called the 2-domain Tait PVT model, shown in the following equation (1), to the measured values.

[0017]

number

[0018] Figure (B) shows the viscosity of recycled materials A, B, and C, with the horizontal axis representing the shear rate of the material [1 / sec] and the vertical axis representing the viscosity of the material [Pa·s]. The temperature of the materials is assumed to be a constant 200 [°C]. The point plots shown in Figure (B) are measured values ​​obtained using a dedicated measuring device. The line plots show the results of fitting a known computational model called the Cross-WLF model, shown in equation (2), to the measured values.

[0019]

number

[0020] As is clear from Figure 1, in the case of recycled materials, even if the product number is the same, if the lot is different, it can be confirmed that the physical properties (in this case, PVT properties and viscosity) are not stable.

[0021] <Example of the configuration of the injection molding system 10 according to an embodiment of the present invention> Next, Figure 2 shows an example of the configuration of an injection molding system 10 according to an embodiment of the present invention. The injection molding system 10 estimates a PVT property calculation model and a viscosity calculation model corresponding to a material with unstable physical properties, i.e., a material with unknown physical properties, and quantifies the variation in the quality of molded products when a material with unstable physical properties is used by performing a flow analysis using the estimated PVT property calculation model and viscosity calculation model.

[0022] The injection molding system 10 includes a manufacturing control device 20, factory equipment 30, and a molded product quality variation estimation device 40.

[0023] The manufacturing control device 20 instructs the injection molding machine 31, which is included in the factory equipment 30, to manufacture products (molded products) by injection molding. The manufacturing control device 20 includes a manufacturing condition determination unit 21 and a production instruction unit 22.

[0024] The manufacturing condition determination unit 21 determines the manufacturing conditions based on the standard injection molding conditions (hereinafter referred to as standard molding conditions) set in advance by the user, and outputs them to the production instruction unit 22. The standard molding conditions include parameters corresponding to each step of the injection molding process (details described later).

[0025] For example, the standard molding conditions include parameters for the metering and plasticizing process, such as metering position, suck-back, back pressure, back pressure speed, and rotation speed. The standard molding conditions also include parameters for the injection and holding pressure processes, such as material pressure, material temperature (temperature of the heater that heats the material), mold temperature (temperature and flow rate of the coolant that cools the mold), holding pressure time, and shear rate. Furthermore, the standard molding conditions include parameters for the injection and holding pressure processes, such as the screw position for switching between injection and pressure (VP switching position) and the clamping force of the mold 509. Finally, the standard molding conditions include parameters for the cooling process, such as the cooling time after holding pressure.

[0026] The manufacturing conditions include standard molding conditions, information specifying the injection molding machine 31, information specifying the mold, mold capacity, configuration of the mold runner, information specifying the material (product number, lot number), quantity of molded products to be produced, production period, required quality, etc.

[0027] The production instruction unit 22 instructs the injection molding machine 31 to manufacture products (molded parts) by outputting control information based on the manufacturing conditions. The manufacturing conditions are also supplied to the molded part quality variation estimation device 40 via the injection molding machine 31. The manufacturing conditions supplied to the molded part quality variation estimation device 40 are referenced when estimating the PVT characteristic calculation model and viscosity calculation model for each product number and lot number of the material.

[0028] The factory equipment 30 includes an injection molding machine 31 and a quality inspection device 33.

[0029] The injection molding machine 31 executes the injection molding process according to control information from the production instruction unit 22 and outputs the manufactured molded product to the quality inspection device 33. The injection molding machine 31 has an in-mold sensor 32. The in-mold sensor 32 consists of a plurality of various sensors. The in-mold sensor 32 includes, for example, a temperature sensor for measuring material temperature, a pressure sensor for measuring material pressure, and a temperature sensor for measuring mold temperature. The in-mold sensor 32 measures physical quantities (temperature, pressure, etc.) in the injection molding process as in-mold sensor values ​​and outputs the time-series data of the measurement results as sensor information to the molded product quality variation estimation device 40. The in-mold sensor 32 may also measure, for example, the shear rate of the material, the physical properties of the material, the mold opening amount (mold opening amount), etc. Physical properties of the material refer to, for example, the density of the material, the viscosity of the material, the distribution of fiber length of the material (if the material contains reinforcing fibers), etc. Of these, the physical quantity that most correlates with the fluidity of a material is its viscosity. However, other quantities that correlate with fluidity, such as pressure, temperature, and velocity, can also be used as physical quantities that correlate with fluidity.

[0030] The quality inspection device 33 inspects the quality of the molded product. For example, the quality inspection device 33 measures the weight and dimensions of the molded product as quality parameters. In addition, the inspection items by the quality inspection device 33 may include, for example, shape characteristics (thickness, sink marks, burrs, warping, etc.), surface characteristics such as appearance defects (welds, silvering, burning, whitening, scratches, bubbles, delamination, flow marks, jetting, color, gloss, etc.), and mechanical and optical properties (tensile strength, impact resistance, transmittance, etc.). Alternatively, instead of or in addition to the quality inspection device 33, an inspector (manually) may perform the quality inspection of the molded product. The quality inspection device 33 links the manufacturing conditions during the production of the molded product with the quality inspection results of the molded product and outputs them to the molded product quality variation estimation device 40.

[0031] The molded product quality variation estimation device 40 comprises a physical property calculation model generation unit 41, a flow analysis unit 42, an input unit 43, a display unit 44, and a communication unit 45.

[0032] The physical property calculation model generation unit 41 generates a physical property calculation model (specifically, a PVT characteristic calculation model and a viscosity calculation model) corresponding to a material whose physical properties are unknown, based on the manufacturing conditions from the manufacturing control device 20, sensor information input from the production plant equipment 30, and quality inspection results. The physical property calculation model generation unit 41 may also generate physical property calculation models for other physical properties (for example, degree of crystallinity, crystallization rate, etc.).

[0033] Generally, a computational model refers to a mathematical model that mathematically abstracts a phenomenon. A PVT characteristic computational model refers to a mathematical model that represents the PVT characteristics of a material. A viscosity computational model refers to a mathematical model that represents the viscosity of a material. In this embodiment, the PVT characteristic computational model and the viscosity computational model refer to computational models in which the parameters (coefficients) of the mathematical model are determined for each predetermined unit of material (for example, each lot).

[0034] The fluid analysis unit 42 uses the PVT property calculation model and viscosity calculation model estimated by the physical property calculation model generation unit 41 for materials with unknown physical properties to perform a fluid analysis on materials with unknown physical properties, and estimates and outputs the variation in the quality of molded products when using materials with unknown physical properties.

[0035] The input unit 43 accepts various inputs from the user. The display unit 44 displays the UI screen 1000 (Figure 14). The communication unit 45 connects the factory equipment 30 via a network (not shown), such as the Internet, and communicates various information.

[0036] Next, Figure 3 shows an example of the configuration of the functional blocks of the physical property calculation model generation unit 41 and the fluid analysis unit 42.

[0037] The physical property calculation model generation unit 41 includes a sensor information holding unit 51, a feature extraction unit 52, a feature DB 53, a quality DB 54, a PVT characteristic calculation model coefficient DB 55, a viscosity calculation model coefficient DB 56, a linking processing unit 57, a learning DB 58, a learning data selection unit 59, a PVT characteristic calculation model estimation unit 60, and a viscosity calculation model estimation unit 61.

[0038] The sensor information holding unit 51 temporarily holds the sensor information. The feature extraction unit 52 reads the sensor information temporarily held by the sensor information holding unit 51 and extracts the feature quantities of the physical quantities contained in the sensor information.

[0039] Figure 4 shows an example of sensor information. The upper part of the figure shows the time-series change in material pressure during the injection molding process, and the lower part shows the time-series change in material temperature during the injection molding process. In the figure, lots A, B, and C refer to materials A, B, and C, which have the same product number but are from different lots. The same applies to subsequent figures.

[0040] For example, the feature extraction unit 52 calculates the integral value of the material pressure during the injection process (from the injection start timing to the timing when the material pressure reaches its peak value) as a feature. The feature extraction unit 52 also detects the peak value of the material temperature during the injection process as a feature. In addition, the feature may be, for example, the peak value of a physical quantity included in the sensor information, or the integral value of a physical quantity from the injection process to the cooling process, or the maximum derivative value of a physical quantity. Alternatively, a physical quantity may be used directly as a feature. The feature extraction unit 52 records the extracted feature in the feature DB 53. The feature extracted by the feature extraction unit 52 and the quality inspection results of the molded product by the quality inspection device 33 correspond to the process data of the present invention.

[0041] Figure 5 shows an example of a feature vector DB53. The feature vector DB53 records sensor information features, combinations of the injection molding machine 31 and molds, and predetermined material units (lots) linked together. The data set recorded in the feature vector DB53, which links predetermined material units, combinations of the injection molding machine 31 and molds, and extracted features, is called a feature vector dataset.

[0042] Here, with the combination of the injection molding machine 31 and the mold fixed, and the standard molding conditions in the injection molding process fixed, if only a predetermined material unit (for example, only the lot) of the material is changed, the feature quantities extracted from the physical quantities obtained from the in-mold sensor 32 are strongly influenced by the changes in material information (for example, fluidity and physical properties) between material units. Therefore, it becomes possible to record material information unique to each material unit as the difference between feature quantities.

[0043] Since the feature quantities extracted from the physical quantities obtained from the in-mold sensor 32 are affected by machine differences between the injection molding machine 31 and the mold, the feature quantity dataset stores unique material information for each combination of injection molding machine 31 and mold, and for each material unit, as differences between feature quantities. In this embodiment, it is assumed that all feature quantity datasets are based on a fixed combination of injection molding machine 31 and mold.

[0044] Return to Figure 3. The quality DB 54 records the quality inspection results input from the quality inspection device 33.

[0045] Figure 6 shows an example of a quality database (DB54). The quality database (DB54) records the quality inspection results (weight and dimensions in this figure) input from the quality inspection device (33), the combination of the injection molding machine (31) and the mold, and a predetermined material unit (lot).

[0046] Returning to Figure 3, the PVT property calculation model coefficients DB55 store each coefficient of the PVT property calculation model, which is constructed based on the PVT properties of a predetermined material unit whose physical properties are known, and is linked to the information of the predetermined material unit. For the PVT property calculation model, calculation models such as the 2-domain-tait model, Spencer-Glimore model, Modified-Cell model, and Simha-Somcynsky model can be used.

[0047] The viscosity calculation model coefficients DB56 store each coefficient of a viscosity calculation model constructed based on the viscosity of a predetermined material unit whose physical properties are known, and these coefficients are linked to information of the predetermined material unit. In this embodiment, a calculation model such as the Cross-WLF model can be used as the viscosity calculation model.

[0048] The concatenation processing unit 57 associates the feature dataset linked to the material information obtained from the feature DB 53, the molded product quality and manufacturing conditions obtained from the quality DB 54, the coefficients of the PVT characteristic calculation model obtained from the PVT characteristic calculation model coefficient DB 55, and the coefficients of the viscosity calculation model obtained from the viscosity calculation model coefficient DB 56, using the material information as the joining key. The concatenation processing unit 57 then records the associated feature dataset, molded product quality, the coefficients of the PVT characteristic calculation model, the coefficients of the viscosity calculation model, and the materials used in manufacturing as a training dataset in the training DB 58.

[0049] The training data selection unit 59 refers to the training DB 58 and selects one or more training datasets corresponding to materials designated by the user as the target of analysis (materials with unknown physical properties) and materials with the same product number but different lots (materials with known physical properties). The more training datasets selected here, the better the training accuracy of the regression model described later can be. The training data selection unit 59 then extracts values ​​with a high correlation to PVT properties (details described later) from the selected training datasets as PVT property training datasets and stores them in the training DB 601 (Figure 8) of the PVT property calculation model estimation unit 60. The training data selection unit 59 also extracts values ​​with a high correlation to viscosity (details described later) from the selected training datasets as viscosity training datasets and stores them in the training DB 611 (Figure 8) of the viscosity calculation model estimation unit 61. The training data selection unit 59 corresponds to the selection unit of the present invention.

[0050] Figure 7 shows an example of a training dataset extracted by the training data selection unit 59. Figure (A) shows an example of a PVT characteristic training dataset, and Figure (B) shows an example of a viscosity training dataset.

[0051] As shown in Figure (A), the PVT characteristic calculation model uses characteristic values ​​of material pressure, characteristic quantities of material temperature, and weight and dimensions as molded product quality as explanatory variables, all of which are highly correlated with PVT characteristics. As shown in Figure (B), the viscosity calculation model uses characteristic values ​​of material pressure, which are highly correlated with viscosity, as explanatory variables.

[0052] Return to Figure 3. The PVT characteristic calculation model estimation unit 60 estimates a PVT characteristic calculation model corresponding to a material with unknown physical properties. The viscosity calculation model estimation unit 61 estimates a viscosity calculation model corresponding to a material with unknown physical properties.

[0053] Next, Figure 8 shows an example configuration of the PVT characteristic calculation model estimation unit 60 and the viscosity calculation model estimation unit 61.

[0054] The PVT characteristic calculation model estimation unit 60 includes a learning DB 601, an estimation DB 602, a regression model learning unit 603, a trained regression model DB 604, a physical property calculation model estimation unit 605, and a PVT characteristic calculation model DB 606.

[0055] The training DB601 stores the PVT characteristic training dataset used to generate the PVT characteristic calculation model selected by the training data selection unit 59. The estimation DB602 stores the estimation dataset (feature dataset, molded product quality) corresponding to materials with unknown physical properties.

[0056] The regression model learning unit 603 reads the PVT characteristic learning dataset from the learning DB 601 and specifies the feature dataset and molded product quality included in the PVT characteristic learning dataset as explanatory variables for the regression model. The regression model learning unit 603 also specifies the PVT characteristic calculation model coefficients as the target variable for the regression model, performs machine learning on a regression model that predicts the target variable from the explanatory variables, generates a trained regression model, and stores it in the trained regression model DB 604.

[0057] Generally, a regression model refers to a model that predicts the target variable y from an explanatory variable x (y=f(x)). The parameters within a regression model are determined by the training data. In this embodiment, "regression model" refers to regression models in general, and "trained regression model" refers to a regression model whose parameters have been determined by the training dataset.

[0058] In this embodiment, the regression model can be, for example, linear regression, ridge regression, support vector machine, neural network, random forest regression, or a combination thereof. Furthermore, if a regression model that can have multiple objective variables, such as a neural network, is adopted, multiple coefficients may be selected from the coefficients of the PVT characteristic calculation model as the objective variables.

[0059] The physical property calculation model estimation unit 605 estimates a PVT property calculation model for materials whose PVT properties are unknown. Specifically, the physical property calculation model estimation unit 605 inputs the estimation dataset read from the estimation DB 602 into a trained regression model read from the trained regression model DB 604 (for example, a trained regression model corresponding to a material with the same product number as a predetermined material unit whose PVT properties are unknown, but which is different in a predetermined material unit (e.g., lot)), and estimates a PVT property calculation model for materials whose PVT properties are unknown by obtaining the coefficients of the PVT property calculation model output from the trained regression model. The physical property calculation model estimation unit 605 also stores the estimated PVT property calculation model (hereinafter referred to as the estimated PVT property calculation model) in the PVT property calculation model DB 606.

[0060] The PVT characteristic calculation model DB606 stores known PVT characteristic calculation models stored in the PVT characteristic calculation model coefficient DB55 and estimated PVT characteristic calculation models estimated by the material property calculation model estimation unit 605 as data linked to the corresponding material units and information indicating whether the PVT characteristic calculation model is known or estimated.

[0061] Similarly, the viscosity calculation model estimation unit 61 includes a learning DB 611, an estimation DB 612, a regression model learning unit 613, a learned regression model DB 614, a physical property calculation model estimation unit 615, and a viscosity calculation model DB 616.

[0062] The training DB611 stores viscosity training datasets used to generate viscosity calculation models selected by the training data selection unit 59. The estimation DB612 stores estimation datasets (feature datasets, molded product quality) corresponding to materials with unknown physical properties.

[0063] The regression model learning unit 613 reads the viscosity learning dataset from the learning DB 611 and specifies the feature dataset included in the viscosity learning dataset as explanatory variables for the regression model. The regression model learning unit 613 also specifies the viscosity calculation model coefficients as the target variable for the regression model, performs machine learning on a regression model that predicts the target variable from the explanatory variables, generates a trained regression model, and stores it in the trained regression model DB 614.

[0064] The physical property calculation model estimation unit 615 estimates a viscosity calculation model for a material whose viscosity is unknown. Specifically, the physical property calculation model estimation unit 615 inputs the estimation dataset read from the estimation DB 612 into a trained regression model read from the trained regression model DB 614 (for example, a trained regression model corresponding to a material of a predetermined material unit whose viscosity is unknown and which has the same product number but a different predetermined material unit (e.g., lot)), and estimates a viscosity calculation model for a material of a predetermined material unit whose viscosity is unknown by obtaining the coefficients of the viscosity calculation model output from the trained regression model. The physical property calculation model estimation unit 615 also stores the estimated viscosity calculation model (hereinafter referred to as the estimated viscosity calculation model) in the viscosity calculation model DB 616.

[0065] The viscosity calculation model DB616 stores known viscosity calculation models stored in the viscosity calculation model coefficient DB56 and estimated viscosity calculation models estimated by the material property calculation model estimation unit 615 as data linked to the corresponding material units and information indicating whether the viscosity calculation model is known or estimated.

[0066] Returning to Figure 3, the fluid analysis unit 42 comprises an analytical property calculation model DB 71, an analytical property model readout unit 72, an analysis unit 73, and an analysis result output unit 74.

[0067] The analytical property calculation model DB71 stores the estimated PVT property calculation model and the estimated viscosity calculation model generated by the property calculation model generation unit 41. The analytical property model reading unit 72 reads the coefficients of the estimated PVT property calculation model and the estimated viscosity calculation model from the analytical property calculation model DB71 and outputs them to the analysis unit 73. The analysis unit 73 performs a flow analysis on a material with unknown properties using the coefficients of the estimated PVT property calculation model and the estimated viscosity calculation model, according to the analysis shape and analysis conditions specified by the user, and outputs the estimated quality (weight, dimensions, etc.) of the molded product as an analysis result to the analysis result output unit 74.

[0068] Here, the analysis shape specified by the user refers to the shape of the mold. This analysis shape (mold shape) may be different from the mold shape used when the regression model was machine-trained. Also, the analysis conditions specified by the user may be different from the standard molding conditions used when the regression model was machine-trained.

[0069] The analysis result output unit 74 aggregates the analysis results (quality of molded products) from the analysis unit 73 for each lot and calculates the quality variation as the difference between the maximum and minimum values ​​of a quality-representing value (weight, dimensions, etc.). Alternatively, instead of or in addition to the difference between the maximum and minimum values ​​of the quality-representing value, the variance, standard deviation, etc., of the quality-representing value may be calculated. The analysis result output unit 74 also determines whether the calculated quality variation meets the control conditions set in advance by the user. Furthermore, the analysis result output unit 74 displays the calculated quality variation and the determination result of whether or not the control conditions are met on the UI screen 1000 (Figure 14) and notifies the user.

[0070] For determining whether a control condition is met, for example, the system may compare the quality variation for each lot with the control range specified by the user and notify whether the quality variation falls within the control range. Alternatively, the system may notify whether the quality of the molded product falls within the control range when the shape conditions and standard molding conditions are changed while keeping the material fixed. Furthermore, the system may notify whether the quality of the molded product falls within the control range when the material is changed while keeping the shape conditions and standard molding conditions fixed.

[0071] Next, Figure 9 shows an example configuration of a general computer 100, such as a personal computer or server computer, that implements the molded product quality variation estimation device 40.

[0072] The computer 100 includes a processor 101 such as a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit), memory 102 such as DRAM (Dynamic Random Access Memory), storage 103 such as an HDD (Hard Disk Drive) and an SSD (Solid State Drive), input devices 104 such as a keyboard, mouse, and touch panel, output devices 105 such as a display, and a communication module 106 such as a NIC (Network Interface Card).

[0073] Computer 100 executes a program using a processor 101 and performs processing defined by the program using memory resources (memory 102 and storage 103) and a communication module 106, etc. Therefore, the processor 101 may be the main entity performing the processing by executing the program. Similarly, the main entity performing the processing by executing the program may be a controller, device, system, computer, or node having a processor. The main entity performing the processing by executing the program may be an arithmetic unit, and may include a dedicated circuit that performs a specific processing. Here, a dedicated circuit is, for example, an FPGA (Field Programmable Gate Array), an ASIC (Application Specific Integrated Circuit), or a CPLD (Complex Programmable Logic Device).

[0074] The program may be installed on the computer 100 from the program source. The program source may be, for example, a program distribution server or a storage medium readable by the computer 100. If the program source is a program distribution server, the program distribution server includes a processor and storage resources for storing the program to be distributed, and the processor of the program distribution server may distribute the program to other computers. In addition, in the embodiment, two or more programs may be implemented as one program, or one program may be implemented as two or more programs.

[0075] The functional blocks in the physical property calculation model generation unit 41, namely the sensor information holding unit 51, the feature extraction unit 52, the concatenation processing unit 57, the training data selection unit 59, the PVT characteristic calculation model estimation unit 60 (excluding various internal databases), and the viscosity calculation model estimation unit 61 (excluding various internal databases), are all implemented by the processor 101 of the computer 100. Similarly, the functional blocks in the fluid analysis unit 42, namely the analysis physical property model reading unit 72, the analysis unit 73, and the analysis result output unit 74, are all implemented by the processor 101 of the computer 100.

[0076] These functional blocks are implemented by the processor 101 executing a predetermined program loaded into memory. However, some or all of these functional blocks may be implemented as hardware using integrated circuits or the like. Furthermore, these functional blocks may be implemented by one computer 100 or by multiple computers 100. If implemented by multiple computers 100, the multiple computers 100 only need to be connected via a network and may be distributed and located in remote locations.

[0077] The feature quantities DB53, quality DB54, PVT characteristic calculation model coefficients DB55, viscosity calculation model coefficients DB56, learning DB58 in the physical property calculation model generation unit 41 of the molded product quality variation estimation device 40, various DBs within the PVT characteristic calculation model estimation unit 60, and various DBs within the viscosity calculation model estimation unit 61 are realized by the memory 102 and storage 103 of the computer 100. Similarly, the fluid analysis unit 42's analysis physical property calculation model DB71 is realized by the memory 102 and storage 103 of the computer 100.

[0078] The input unit 43 is implemented by an input device 104 provided by the computer. The display unit 44 is implemented by an output device 105 provided by the computer. The communication unit 45 is implemented by a communication module 106 provided by the computer.

[0079] Furthermore, if the molded product quality variation estimation device 40 is located on a cloud server, for example, the user will connect to the molded product quality variation estimation device 40 via the network using a terminal device such as a PC and operate the molded product quality variation estimation device 40.

[0080] <Injection molding process using injection molding machine 31> Next, the injection molding process using the injection molding machine 31 will be described. Figure 10 is a cross-sectional view showing an example of the configuration of the injection molding machine 31.

[0081] As described above, the injection molding process is broadly divided into the metering and plasticizing process, the injection process, the holding pressure process, the cooling process, and the removal process.

[0082] In the weighing and plasticizing process, the injection molding machine 31 uses a plasticizing motor 501 to drive the screw 502 back, supplying resin pellets 504, the material, from the hopper 503 into the cylinder 505. Then, heating the cylinder 505 by the heater 506 and the rotation of the screw 502 plasticize the material to a uniform molten state. The back pressure and rotation speed of the screw 502 change the density of the molten material and the degree of fracture of the reinforcing fibers depending on their settings, and these changes can affect the quality of the molded product.

[0083] In the injection and holding pressure processes, the injection molding machine 31 uses the injection motor 507 as a driving force to advance the screw 502 and inject the molten material into the mold 509 via the nozzle 508. At this time, the pressure measured by the load cell 510 is controlled to approach the holding pressure value included in the injection molding conditions. The molten material injected into the mold 509 is subjected to cooling from the walls of the mold 509 and shear heating due to flow in parallel. That is, the molten material flows through the mold 509 while undergoing both cooling and heating. If the clamping force, which is the force that keeps the mold 509 closed, is insufficient, tiny gaps may form in the mold 509 after the molten material solidifies, which can affect the quality of the molded product.

[0084] In the cooling process, the injection molding machine 31 cools the mold 509 to below its solidification temperature to solidify the molten material. Residual stress generated during the cooling process can affect the quality of the molded product. Residual stress is generated due to the anisotropy of material properties caused by flow within the mold 509, the density distribution due to holding pressure, and the unevenness of the molding shrinkage rate.

[0085] In the removal process, the injection molding machine 31 opens the mold 509 by driving the clamping mechanism 512 with the motor 511 as the driving force. Then, the ejector mechanism 514 is driven with the ejection motor 513 as the driving force to remove the molded product from the mold 509, which is solidified from the molten material. At this time, if sufficient ejection force from the ejector mechanism 514 is not applied evenly to the molded product, residual stress may remain in the molded product, which may affect the quality of the molded product.

[0086] In the injection molding machine 31, the pressure value of the load cell 510 is controlled to approach the pressure value included in the standard molding conditions. The temperature of the cylinder 505 is controlled by multiple heaters 506. However, because there are differences between injection molding machines 31 in the shape of the screw 502, the cylinder 505, and the nozzle 508, different pressure losses may occur for each injection molding machine 31. As a result, the pressure of the material at the material inlet of the mold 509 will be lower than the pressure specified as the standard molding conditions. In addition, the temperature of the material at the material inlet of the mold 509 may differ from the temperature specified as the standard molding conditions due to the arrangement of the heaters 506 and the shear heating of the material at the nozzle 508. These differences in pressure and temperature can affect the quality of the molded product.

[0087] In this embodiment, the weight and dimensions of the molded product are evaluated as quality inspections of the molded product. The weight of the molded product is thought to change depending on the specific volume of the material relative to the volume of the mold 509, provided that there are no defects such as short shots or overpacking in the shape of the molded product. Therefore, the weight of the molded product is a physical quantity that has a high correlation with the PVT characteristics. The dimensions of the molded product are thought to change depending on the shrinkage rate of the material, provided that there are no defects such as short shots or overpacking in the shape of the molded product. Therefore, the dimensions of the molded product are also a physical quantity that has a high correlation with the PVT characteristics.

[0088] The shape characteristics (weight, dimensions, etc.) of molded products are strongly correlated with the pressure history, temperature history, and clamping force during the injection molding, holding pressure, and cooling processes.

[0089] Next, Figure 11 shows an example of a mold 509, where Figure (A) is a top view of the product section 5091 of the mold 509, Figure (B) is a side view of the product section 5091, Figure (C) is a top view of the runner section 5092 of the mold 509, and Figure (D) is a side view of the runner section 5092.

[0090] In the mold 509, molten material flows into the product section 5091 from five gate sections 5093 of the runner section 5092. The product section 5091 is equipped with a material temperature sensor 321 for measuring the material temperature inside the mold 509, a material pressure sensor 322 for measuring the material pressure, and a mold temperature sensor 323 for measuring the temperature of the mold 509. The runner section 5092 is equipped with the material temperature sensor 321 and the material pressure sensor 322. The material temperature sensor 321, the material pressure sensor 322, and the mold temperature sensor 323 correspond to the in-mold sensor 32 (Figure 1).

[0091] Furthermore, regarding the arrangement of each sensor (material temperature sensor 321, material pressure sensor 322, and mold temperature sensor 323), it is preferable that the sprue portion 5094 or runner portion 5092 extends at least from the material inlet in the mold 509 to the product portion (cavity) 5091. In this specification, "preferred" simply means that some advantageous effect can be expected, and does not mean that the configuration is essential.

[0092] Furthermore, the placement of each sensor may be in locations where characteristic flow can be observed, such as directly below the gate, the resin confluence (weld section), or the flow end (none of which are shown) within the product section 5091. In this case, physical quantities correlated with the material's inherent PVT characteristics can be determined with higher accuracy from the physical quantities obtained from multiple sensors with different placements. In addition, by using the integral value up to the peak obtained from the pressure sensor directly below the gate, which correlates with the material's inherent viscosity, predictions can be made with high accuracy.

[0093] For example, in molding under standard molding conditions, the pressure and temperature of the molten material during molding vary depending on its position within the mold 509. Therefore, by using multiple sensors positioned differently, it is possible to measure the physical quantities of the material under different pressure and temperature conditions within a single molding condition. This allows for more accurate determination of physical quantities correlated with the material's inherent PVT properties.

[0094] The appropriate placement of each sensor will vary depending on the mold structure and the physical quantity being measured. For physical quantities other than the mold opening amount, it is preferable to place the sensor in the sprue section 5094, if possible, regardless of the mold structure.

[0095] If the gate section 5093 is a side gate, jump gate, submarine gate, or banana gate, the sensors are placed in the runner section 5092 directly below the sprue section 5094, or in the runner section 5092 immediately before the gate section 5093, etc.

[0096] If the gate section 5093 is a pin gate, it will have a three-plate structure, requiring careful consideration of the placement of each sensor. One possible solution is to place each sensor in the runner section 5092 directly below the sprue section 5094. Alternatively, if the gate section 5093 is a pin gate, a dummy runner section 5092 that is not connected to the product section 5091 may be provided for measurement purposes, and each sensor may be placed in this dummy runner section 5092. Providing a runner section 5092 specifically for measurement increases the flexibility of the mold design.

[0097] If the gate section 5093 is a film gate or a fan gate, the sensors are installed in the runner section 5092 before the material flows into the gate section 5093.

[0098] <Processing for estimating molded product quality variations by injection molding system 10> Next, Figure 12 is a flowchart illustrating an example of the molding quality variation estimation process performed by the injection molding system 10.

[0099] As a prerequisite for the molding product quality variation estimation process, the learning DB 58 of the physical property calculation model generation unit 41 is assumed to already contain learning datasets corresponding to materials that may be specified by the user as the material to be analyzed (material with unknown physical properties) and materials that have the same product number but are from different lots (material with known physical properties). Furthermore, the estimation DBs 602 and 612 of the physical property calculation model generation unit 41 are assumed to already contain estimation datasets corresponding to materials that may be specified by the user as the material to be analyzed (material with unknown physical properties). However, if an estimation dataset corresponding to a material to be analyzed by the user (material with unknown physical properties) is not stored in the estimation DBs 602 and 612, the injection molding process using that material can be executed in the injection molding machine 31 to generate an estimation dataset corresponding to that material and store it in the estimation DBs 602 and 612.

[0100] The molding product quality variation estimation process is initiated, for example, when the user specifies the product number and lot number of the material to be analyzed, the analysis shape (product number of the mold), the analysis properties, the analysis conditions, and the quality control range of the molded product on the UI screen 1000 (Figure 14), and then operates the "Execute Analysis" button 1006.

[0101] First, the learning data selection unit 59 of the physical property calculation model generation unit 41 refers to the learning DB 58 and selects one or more learning datasets corresponding to materials designated by the user as the target of analysis (materials with unknown physical properties) and materials with the same product number but different lots (materials with known physical properties). Then, the learning data selection unit 59 extracts the PVT property learning dataset to be used for generating the PVT property calculation model from the selected learning datasets and stores it in the learning DB 601 of the PVT property calculation model estimation unit 60. In addition, the learning data selection unit 59 extracts the viscosity learning dataset to be used for generating the viscosity calculation model from the selected learning datasets and stores it in the learning DB 611 of the viscosity calculation model estimation unit 61 (step S1).

[0102] Next, the regression model learning unit 603 generates a trained regression model that outputs PVT characteristic calculation model coefficients and stores it in the trained regression model DB 604. Simultaneously, the regression model learning unit 613 generates a trained regression model that outputs viscosity calculation model coefficients and stores it in the trained regression model DB 614 (step S2).

[0103] Specifically, the regression model learning unit 603 reads the PVT characteristic learning dataset from the learning DB 601 and specifies the feature dataset and molded product quality included in the PVT characteristic learning dataset as explanatory variables for the regression model. The regression model learning unit 603 also specifies the PVT characteristic calculation model coefficients as the target variable for the regression model and generates a trained regression model by learning a regression model that predicts the target variable from the explanatory variables. Simultaneously, the regression model learning unit 613 reads the viscosity learning dataset from the learning DB 611 and specifies the feature dataset included in the viscosity learning dataset as explanatory variables for the regression model. The regression model learning unit 613 also specifies the viscosity calculation model coefficients as the target variable for the regression model and generates a trained regression model by learning a regression model that predicts the target variable from the explanatory variables.

[0104] Next, the physical property calculation model estimation unit 605 generates an estimated PVT property calculation model corresponding to a material whose PVT properties are unknown and stores it in the PVT property calculation model DB606. Simultaneously, the physical property calculation model estimation unit 615 generates an estimated viscosity calculation model for a material whose viscosity is unknown and stores it in the viscosity calculation model DB616 (step S3).

[0105] Specifically, the physical property calculation model estimation unit 605 reads an estimation dataset corresponding to the material under analysis from the estimation DB 602, inputs it into the trained regression model for the PVT property calculation model generated in step S2, and obtains the coefficients of the PVT property calculation model, thereby generating an estimated PVT property calculation model corresponding to the material under analysis. Simultaneously, the physical property calculation model estimation unit 605 inputs an estimation dataset corresponding to the material under analysis from the estimation DB 612 into the trained regression model for the viscosity calculation model generated in step S2, and obtains the coefficients of the viscosity calculation model, thereby estimating an estimated viscosity calculation model corresponding to the material under analysis.

[0106] Furthermore, the estimated PVT characteristic calculation model and estimated viscosity calculation model generated in step S3 are also stored in the analytical physical property calculation model DB71 of the fluid analysis unit 42.

[0107] Next, the analysis property model readout unit 72 of the flow analysis unit 42 reads the coefficients of the estimated PVT property calculation model and the estimated viscosity calculation model corresponding to the material under analysis from the analysis property calculation model DB 71 and outputs them to the analysis unit 73. Then, the analysis unit 73 performs a flow analysis on the material under analysis using the coefficients of the estimated PVT property calculation model and the estimated viscosity calculation model, according to the specified analysis shape and analysis conditions specified by the user, and outputs the estimated quality (weight, dimensions, etc.) of the molded product as an analysis result to the analysis result output unit 74 (step S4).

[0108] Next, the analysis result output unit 74 aggregates the analysis results (quality of molded products) from the analysis unit 73 for each lot and calculates the difference between the maximum and minimum values ​​of quality-representing values ​​(weight, dimensions, etc.) as quality variation. The analysis result output unit 74 also determines whether the calculated quality variation meets the control conditions set in advance by the user. Finally, the analysis result output unit 74 displays the quality variation and the pass / fail judgment results against the control conditions on the UI screen 1000 (Figure 14) (step S5). The above is an example of the molded product quality variation estimation process by the injection molding system 10.

[0109] The molded product quality variation estimation process can generate an estimated PVT property calculation model and an estimated viscosity calculation model corresponding to the material being analyzed. Furthermore, based on the generated estimated PVT property calculation model and estimated viscosity calculation model, the process can estimate the PVT properties and viscosity corresponding to the material being analyzed. Then, based on the estimated PVT properties and viscosity corresponding to the material being analyzed, the process can estimate the quality and variation of molded products when using the material being analyzed. This allows users to obtain the quality variation without mass-producing products (molded products) using the material being analyzed, enabling them to appropriately select materials suitable for mass production.

[0110] <Validity of the estimated PVT physical property calculation model and estimated viscosity calculation model> Next, Figure 13 is a diagram illustrating the validity of the estimated PVT physical property calculation model and the estimated viscosity calculation model generated by the molded product quality variation estimation process.

[0111] In this validation evaluation, materials A, B, C, and D, which had the same product number but were from different lots, were used, and the physical properties of materials A through D were measured using existing methods. Subsequently, a trained regression model was created using training datasets corresponding to materials A, B, and C. Then, by inputting an estimation training dataset corresponding to material D into the trained regression model based on materials A, B, and C, an estimated PVT physical property calculation model and an estimated viscosity calculation model for material D were generated. Furthermore, the estimated PVT physical property calculation model and estimated viscosity calculation model were compared and evaluated with the measured physical properties of material D.

[0112] Figure (A) shows a comparison between the estimated PVT property calculation model (line plot) corresponding to the generated material D and the measured values (point plot) of the PVT properties of material D when the temperature is changed in 5°C increments from 30°C to 250°C with the pressure level fixed at 50 [MPa], 100 [MPa], or 150 [MPa]. The horizontal axis in Figure (A) is temperature and the vertical axis is specific volume. In Figure (A), since the estimated PVT property calculation model almost coincides with the measured values, it can be seen that the generated estimated PVT physical property calculation model is highly valid.

[0113] Figure (B) shows a comparison between the estimated viscosity calculation model (line plot) corresponding to the generated material D and the measured values (point plot) of the viscosity of material D when the shear rate is changed in the range of 6 to 12000 [1 / sec] with the temperature fixed at 200°C. Figure (B) is a double logarithmic graph, where the horizontal axis is the shear rate and the vertical axis is the viscosity. In Figure (B), since the estimated viscosity calculation model almost coincides with the measured values, it can be seen that the generated estimated viscosity calculation model is highly valid.

[0114] <Example display of UI screen 1000> Next, Figure 14 shows an example display of UI screen 1000. UI screen 1000 is provided with an input field 1001 for specifying the material to be analyzed, an input field 1002 for specifying the mold as the analysis shape, an input field 1003 for selecting and specifying the analysis conditions, an input field 1004 for selecting and specifying the analysis physical property model, an input field 1005 for specifying the control width, an "Analysis Execution" button 1006 for instructing the execution of fluid analysis under the specified various conditions, and a display field 1007 for displaying the analysis results.

[0115] Input field 1001 allows you to enter, for example, one product number and multiple lot numbers for the material to be analyzed. Input field 1002 allows you to enter, for example, the product number of the mold. Note that the product number of the mold entered here may be different from the product number of the mold corresponding to the training dataset used to train the regression model. Input field 1003 allows you to select and specify items displayed as analysis conditions (material temperature, mold temperature, injection speed, etc.) by checking the boxes. Input field 1004 allows you to select and specify the analysis property model by checking the boxes displayed for physical properties (viscosity, PVT properties). Input field 1005 allows you to enter the type of quality (weight, dimensions), its lower limit, and its upper limit as the control range.

[0116] Display field 1007 displays the user-specified control range along with the quantitative quality variation that may occur when using materials D, E, and F, which have the same product number but are from different lots, as specified in input field 1001. Furthermore, display field 1007 also displays the pass / fail result for the quality variation relative to the control range.

[0117] According to UI screen 1000, the user can understand the potential quality variations that may occur when using the material they have designated for analysis, and can visually confirm whether or not those variations fall within a manageable range, thereby enabling them to select the appropriate material.

[0118] The present invention is not limited to the embodiments described above, and various modifications are possible. For example, the embodiments described above are described in detail to make the present invention easier to understand, and are not necessarily limited to those having all the configurations described. Furthermore, it is possible to replace or add to the configurations of one embodiment with those of another embodiment.

[0119] Furthermore, each of the above configurations, functions, processing units, and processing means may be implemented in hardware, either partially or entirely, by designing them as integrated circuits, for example. Alternatively, each of the above configurations and functions may be implemented in software by a processor interpreting and executing programs that implement each function. Information such as programs, tables, and files that implement each function can be stored in memory, recording devices such as hard disks and SSDs, or recording media such as IC cards, SD cards, and DVDs. Also, control lines and information lines are shown only if deemed necessary for explanation, and not all control lines and information lines are necessarily shown in the actual product. In practice, it can be assumed that almost all configurations are interconnected. [Explanation of symbols]

[0120] 10...Injection molding system, 20...Manufacturing control device, 21...Manufacturing condition determination unit, 22...Production instruction unit, 31...Injection molding machine, 32...In-mold sensor, 321...Material temperature sensor, 322...Material pressure sensor, 323...Mold temperature sensor, 33...Quality inspection device, 40...Estimation device, 41...Material property calculation model generation unit, 42...Flow analysis unit, 43 ...Input unit, 44...Display unit, 45...Communication unit, 51...Sensor information holding unit, 52...Feature extraction unit, 53...Feature DB, 54...Quality DB, 55...PVT characteristic calculation model coefficient DB, 56...Viscosity calculation model coefficient DB, 57...Linking processing unit, 58...Learning DB, 59...Learning data selection unit, 60...PVT characteristic calculation model estimation unit, 6 01...Training DB, 602...Estimation DB, 603...Regression model learning unit, 604...Trained regression model DB, 605...Material property calculation model estimation unit, 606...PVT property calculation model DB, 61...Viscosity calculation model estimation unit, 611...Training DB, 612...Estimation DB, 613...Regression model learning unit, 614...Trained regression model DB, 615 ...Material properties calculation model estimation unit, 606...Viscosity calculation model DB, 71...Analysis material properties calculation model DB, 72...Analysis material properties calculation model readout unit, 73...Analysis unit, 74...Analysis result output unit, 100...Computer, 509...Mold, 5091...Product unit, 5092...Runner unit, 5093...Gate unit, 5094...Sprull unit, 1000...UI screen

Claims

1. A molded product quality variation estimation device for estimating the quality variation of molded products produced by injection molding, A physical property calculation model estimation unit estimates a physical property calculation model corresponding to a second material based on process data when a molded product is manufactured using a first material with known physical properties, a trained regression model generated based on physical property calculation model coefficients corresponding to the first material, and process data when a molded product is manufactured using a second material with unknown physical properties. The system includes a flow analysis unit that estimates the quality variation of molded products when molded products are manufactured using the second material by performing a flow analysis based on the material property calculation model corresponding to the estimated second material, The material property calculation model estimation unit estimates two or more material property calculation models corresponding to two or more different material properties, based on two or more of the learned regression models corresponding to two or more different material properties. The fluid analysis unit performs the fluid analysis based on the two or more estimated physical property calculation models. A device for estimating variations in molded product quality.

2. A molded product quality variation estimation device according to claim 1, The first material and the second material are recycled materials. The first material and the second material have the same product number but are from different lots. A device for estimating variations in molded product quality.

3. A molded product quality variation estimation device according to claim 1, The system includes a regression model learning unit that generates the trained regression model by learning a regression model in which the process data obtained when a molded product is manufactured using the first material is used as the explanatory variable and the material property calculation model coefficients are used as the objective variable. A device for estimating variations in molded product quality.

4. A molded product quality variation estimation device according to claim 3, The system includes a selection unit for selecting the explanatory variables to be used in training the regression model. A device for estimating variations in molded product quality.

5. A molded product quality variation estimation device according to claim 4, The selection unit selects process data that has a high correlation with the physical properties corresponding to the physical property calculation model coefficients, which are the objective variables, as explanatory variables. A device for estimating variations in molded product quality.

6. A molded product quality variation estimation device according to claim 1, The two or more different physical properties mentioned above include at least one of the following: PVT properties, viscosity, degree of crystallinity, and crystallization rate. A device for estimating variations in molded product quality.

7. A molded product quality variation estimation device according to claim 5, The process data includes characteristic values ​​of the in-mold sensor and the quality of the molded product. A device for estimating variations in molded product quality.

8. A molded product quality variation estimation device according to claim 7, The in-mold sensor values ​​include at least one of pressure and temperature. A device for estimating variations in molded product quality.

9. A molded product quality variation estimation device according to claim 8, The aforementioned feature quantity includes at least one of the following: the integral value of the in-mold sensor value during the injection process, the peak value of the in-mold sensor value, the integral value of the in-mold sensor value from the injection process to the cooling process, and the maximum differential value of the in-mold sensor value. A device for estimating variations in molded product quality.

10. A molded product quality variation estimation device according to claim 9, The quality of the molded product includes at least one of weight and dimensions. A device for estimating variations in molded product quality.

11. A molded product quality variation estimation device according to claim 10, The aforementioned selection unit is As the explanatory variable of the regression model in which viscosity calculation model coefficients are the dependent variable, the integral value of the pressure as the in-mold sensor value during the injection process is selected. As explanatory variables for the regression model in which the PVT characteristic calculation model coefficient is the dependent variable, the following are selected: a feature quantity based on pressure as the in-mold sensor value, a feature quantity based on temperature as the in-mold sensor value, and weight and dimensions as the quality. A device for estimating variations in molded product quality.

12. A molded product quality variation estimation device according to claim 1, The flow analysis unit estimates the minimum and maximum values ​​of at least one of the weight and dimensions of the molded product as the quality variation of the molded product when the molded product is manufactured using the second material. A device for estimating variations in molded product quality.

13. A molded product quality variation estimation device according to claim 12, The fluid analysis unit compares the estimated quality variation with the control range, and if the quality variation falls outside the control range, it notifies the user accordingly. A device for estimating variations in molded product quality.

14. A method for estimating the quality variation of molded products by injection molding using a molded product quality variation estimation device, A property calculation model estimation step in which a property calculation model corresponding to a second material is estimated based on process data when a molded product is manufactured using a first material with known physical properties, a trained regression model generated based on the property calculation model coefficients for the first material, and process data when a molded product is manufactured using a second material with unknown physical properties, The process includes a flow analysis step of estimating the quality variation of a molded product when manufacturing a molded product using the second material by performing a flow analysis based on the material property calculation model corresponding to the estimated second material, The aforementioned material property calculation model estimation step estimates two or more material property calculation models corresponding to two or more different material properties, based on two or more of the previously trained regression models corresponding to two or more different material properties, The aforementioned flow analysis step involves performing the flow analysis based on the two or more estimated physical property calculation models. Method for estimating variations in molded product quality.

15. Factory equipment including injection molding machines and quality inspection devices, An injection molding system comprising: a molded product quality variation estimation device for estimating the quality variation of molded products manufactured by the injection molding machine, The injection molding machine outputs the values ​​from the in-mold sensor during injection molding. The quality inspection device outputs the quality inspection results of the molded product manufactured by the injection molding machine. The aforementioned molded product quality variation estimation device is A physical property calculation model estimation unit estimates a physical property calculation model corresponding to a second material based on process data including feature quantities based on in-mold sensor values ​​and quality inspection results when a molded product is manufactured using a first material with known physical properties, a trained regression model generated based on the physical property calculation model coefficients for the first material, and process data when a molded product is manufactured using a second material with unknown physical properties. The system includes a flow analysis unit that estimates the quality variation of molded products when molded products are manufactured using the second material by performing a flow analysis based on the material property calculation model corresponding to the estimated second material, The material property calculation model estimation unit estimates two or more material property calculation models corresponding to two or more different material properties, based on two or more of the learned regression models corresponding to two or more different material properties. The fluid analysis unit performs the fluid analysis based on the two or more estimated physical property calculation models. Injection molding system.

Images