JP2022030606A - Resin molding analysis method, program, and recording medium - Google Patents

Abstract

To provide a resin molding analysis method which allows for easily predicting internal defects (shrinkage, cavities), surface defects, and overall defects (warp) of molded products.SOLUTION: A resin molding analysis method disclosed herein comprises computing a flow trajectory of a resin flowing to a position of interest of a resin mold during resin molding at the position of interest, and predicting resin defects at the position of interest on the basis of the state of the resin flow trajectory.SELECTED DRAWING: Figure 3

Description

The present invention relates to resin molding analysis methods, programs and recording media.

Conventionally, a resin molding analysis method is known (see, for example, Patent Document 1).

In Patent Document 1, the flow analysis of the resin during resin molding is performed to calculate the flow ratio distribution of the resin, and the filling of the molded product is insufficient based on the calculated flow ratio distribution and a predetermined threshold value. A method for supporting the design of a molded product (resin molding analysis method) for predicting molding defects due to the above is disclosed.

Japanese Unexamined Patent Publication No. 2004-318863

In the design support method (resin molding analysis method) of Patent Document 1, based on the calculated flow ratio distribution and a predetermined threshold value, molding defects due to insufficient filling of the molded product are predicted. Therefore, in the design support method of Patent Document 1, it is possible to predict molding defects due to insufficient filling of the molded product by using a threshold value. However, in order to predict other defects such as internal defects (sinks, nests) in the molded product, surface defects in the molded product, and defects (warp) in the entire molded product, multiple analysis results are judged from multiple perspectives. There is a need to. Therefore, a resin molding analysis method capable of easily predicting internal defects (sinks, nests) of the molded product, surface defects of the molded product, and defects (warp) of the entire molded product is desired.

The present invention has been made to solve the above-mentioned problems, and one object of the present invention is internal defects (sinks, nests) of the molded product, defects on the surface of the molded product, and the entire molded product. It is an object of the present invention to provide a resin molding analysis method, a program and a recording medium capable of easily predicting defects (warpage).

In order to achieve the above object, the resin molding analysis method according to the first aspect of the present invention calculates the flow locus until the resin reaches the attention position at the time of resin molding at the attention position of the resin molded product, and calculates the flow locus of the resin. Based on the state in the flow locus, the molding defect at the position of interest is predicted.

In the resin molding analysis method according to the first aspect of the present invention, by configuring as described above, it is possible to predict molding defects in the region of interest based on the flow trajectory until the resin reaches the position of interest. It is not necessary to judge the occurrence of molding defects by judging multiple analysis results in the region of interest from multiple aspects. Thereby, it is possible to easily predict the internal defect (sink, nest) of the molded product, the surface defect of the molded product, and the defect (warp) of the entire molded product.

In the resin molding analysis method according to the first aspect, preferably, the state in the flow locus of the resin is the wall thickness, the temperature, the flow velocity of the resin, the elastic ratio, the pressure, the flow front speed, the resin pressure, and the air in contact with the flow front. A molding defect at the position of interest is predicted based on at least one of temperature, gate passage speed, runner and gate cross-sectional area, flow velocity vector, shrinkage strain, solidification timing, and anisotropic physical characteristics. With this configuration, the wall thickness of the place where the resin passed, the temperature state when the resin flows, the state of the flow velocity of the resin, the elastic modulus, the pressure, the flow front speed, the resin pressure, the air temperature in contact with the flow front, The state of the resin in the region of interest, which is the final arrival position, can be determined based on the gate passage speed, runner and gate cross-sectional area, flow velocity vector, shrinkage strain, solidification timing, or anisotropic physical properties. It is possible to easily predict the molding defect of the resin at the attention position, which is the final arrival position, by judging the state of the resin.

In this case, preferably, the states in the flow locus of the resin are the wall thickness, the temperature, the flow velocity of the resin, the elastic modulus, the pressure, the flow front speed, the resin pressure, the air temperature in contact with the flow front, the gate passage speed, the runner and the like. Prediction of molding defects at the position of interest based on changes in at least one of the gate cross-sectional area, flow velocity vector, shrinkage strain, solidification timing and anisotropic physical properties. With this configuration, the wall thickness during flow, temperature, sudden change in resin flow velocity, elastic modulus, pressure, flow front speed, resin pressure, air temperature in contact with flow front, gate passage speed, runner and gate Based on the cross-sectional area, the flow velocity vector, the shrinkage strain, the solidification timing, and the anisotropic physical characteristics, it is possible to easily predict the molding defect of the resin that occurs at the downstream attention position.

In the resin molding analysis method according to the first aspect, preferably, the resin in which the molding defect occurs is based on the flow locus of the resin at the time of resin molding at the position where the molding defect has occurred and the state in the flow locus of the resin. The state in the flow locus of the resin is machine-learned, and the molding defect at the position of interest is predicted based on the result of the machine learning and the state in the flow locus of the resin. With this configuration, the relationship between the state in the flow locus of the resin and the state of the attention position can be accumulated by machine learning, so that molding defects in the region of interest are based on the state in the flow locus of the resin. Can be predicted easily and accurately.

In the resin molding analysis method according to the first aspect, preferably, the flow characteristics of the resin obtained by simulating the resin molding with a predetermined mold and the resin obtained by simulating the resin molding with an arbitrary mold are performed. Based on the flow characteristics of, the flow characteristics of the resin with respect to the shape of the mold are machine-learned, and based on the result of the machine learning, the flow characteristics of the resin up to the position of interest are predicted, and the predicted flow characteristics and the resin flow characteristics Based on the state in the flow locus, the molding defect at the position of interest is predicted. With this configuration, even if there is an error between the flow characteristics of a predetermined mold and the flow characteristics of any mold, the position of interest in any mold can be reached based on the results of machine learning. The flow characteristics of the resin can be predicted accurately.

The program according to the second aspect of the present invention causes a computer to execute the resin molding analysis method according to the first aspect.

In the program according to the second aspect of the present invention, by causing the computer to execute the resin molding analysis method according to the first aspect, the internal defects (sinks, nests) of the molded product, the surface defects of the molded product, and the molded product are used. The overall defect (warp) can be easily predicted.

The storage medium according to the third aspect of the present invention is readable by a computer on which the program according to the second aspect is recorded.

In the storage medium according to the third aspect of the present invention, by recording the program according to the second aspect, the internal defects (sinks, nests) of the molded product, the surface defects of the molded product, and the defects of the entire molded product (defects of the entire molded product) are recorded. It is possible to provide a computer-readable recording medium capable of easily predicting warpage).

According to the present invention, as described above, it is possible to easily predict the internal defects (sinks, nests) of the molded product, the surface defects of the molded product, and the defects (warp) of the entire molded product.

It is a block diagram which showed the structural example for carrying out the resin molding analysis method by one Embodiment. It is a figure which showed the molding model by one Embodiment. It is a figure which showed an example of the relationship between the flow locus and the wall thickness by one Embodiment. It is a figure which showed an example of the relationship between the flow locus and temperature by one Embodiment. It is a figure which showed an example of the relationship between the flow locus and the flow velocity by one Embodiment. It is a figure which showed the example of the case which predicts the molding defect based on the flow locus by one Embodiment. It is a figure for demonstrating machine learning for predicting a molding defect by one Embodiment. It is a figure for demonstrating the machine learning for predicting the flow length by one Embodiment. It is a figure which showed an example of the database for the prediction of the flow length by one Embodiment. It is a flowchart for demonstrating the learning phase for predicting the flow length by one Embodiment. It is a flowchart for demonstrating the operation phase for predicting the flow length by one Embodiment.

Hereinafter, embodiments embodying the present invention will be described with reference to the drawings.

A resin molding analysis method according to an embodiment will be described with reference to FIGS. 1 to 8.

The resin molding analysis method according to the present embodiment is a method for predicting molding defects at the position of interest of the molded product. This resin molding analysis method is a method of analyzing the characteristics of a molded product including at least one of internal defects (sinks, nests) of the molded product, defects on the surface of the molded product, and defects (warp) of the entire molded product. be.

(Device configuration example)
The resin molding analysis method according to the present embodiment can be carried out by causing the computer 1 to execute the program 3a. The resin molding analysis method can be carried out, for example, by the apparatus configuration as shown in FIG. The computer 1 is configured to be able to execute the program 3a. The resin molding analysis apparatus 100 is configured by causing the computer 1 to execute the program 3a. A part or all of the processing performed by causing the computer 1 to execute the program 3a may be performed by hardware such as a dedicated arithmetic circuit.

In the configuration example of FIG. 1, the computer 1 is a storage unit including one or a plurality of processors 2 including a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), and a storage device. 3 and. The storage device is, for example, a hard disk drive or a semiconductor storage device.

The computer 1 can perform the resin molding analysis by causing the processor 2 to execute the program 3a stored in the storage unit 3. In addition to being read from the recording medium 7, the program 3a may be provided from an external server or the like via a network such as the Internet or a transmission path 8 such as a LAN (Local Area Network). The recording medium 7 is a computer-readable recording medium such as an optical disk, a magnetic disk, or a non-volatile semiconductor memory, and a program 3a is recorded on the recording medium 7.

In addition to the program 3a, various analysis data 3b used for performing resin molding analysis are stored in the storage unit 3. The analysis data 3b includes resin molding condition information, resin molding information including characteristics of the resin molded product, parameters for simulating the state of the resin at the time of molding and details of the characteristics of the molded product, product category information, and shape characteristics. The amount, the group of optimum parameters, the error rate, the numerical data used for the analysis, the data of the analysis conditions, etc. are stored.

Further, the computer 1 includes a display unit 4 such as a liquid crystal display device, an input unit 5 including an input device such as a keyboard and a mouse, and a reading unit 6 for reading a program 3a and various data from a recording medium 7. The reading unit 6 is a reader device or the like according to the type of the recording medium 7. The analysis condition data can be input by the user using the input unit 5. The analysis data 3b may be read from a recording medium created by the user, or may be created by the user on an external server or the like and acquired from the external server via the transmission path 8.

(Resin molding analysis method)
Next, the resin molding analysis method will be described. In the present embodiment, molding defects at the position of interest are predicted based on the state in the flow locus of the resin.

The resin molding analysis method includes a learning phase and an operation phase. In the learning phase, as shown in FIG. 7, the state in the flow locus of the resin in which the molding defect occurs is learned by machine learning. Specifically, the state in the flow locus of the resin in which the molding defect occurs is machine-learned based on the flow locus of the resin at the time of resin molding at the position where the molding defect occurs and the state in the flow locus of the resin. Is accumulated. Molding defects in the molded product include internal defects (sinks, nests) in the molded product, surface defects (weld, silver) in the molded product, and defects (short, warp) in the entire molded product.

Further, in the learning phase, the shape feature amount is calculated from the three-dimensional CAD data or the analysis mesh. In addition, the recommended module is selected from the user information (license information). In addition, analysis conditions generated from parameter factors and the number of levels are generated. Then, a simulation is performed and the analysis is executed. The difference between the analysis result output by the simulation and the measured value is acquired, and the error is evaluated. Information on the occurrence of defects in the molded product is learned from the error between the analysis result and the measured value.

In the learning phase, a mold model for calculation is created. Then, a simulation of resin injection molding is executed based on the created mold model. Based on the execution simulation result, the flow path (trajectory) to an arbitrary point (attention position) is calculated, and the profile of the characteristic value (state in the flow locus of the resin) to the gate portion is calculated. The characteristic values (states in the flow locus of the resin) are, for example, wall thickness, temperature, flow velocity of the resin, elastic modulus, pressure, flow front speed, resin pressure, air temperature in contact with the flow front, gate passage speed, runner and Includes at least one of gate cross-sectional area, flow velocity vector, shrinkage strain, solidification timing and anisotropic physical properties.

In the learning phase, as input values, the profile of the evaluation items from the place where the molding defect occurs to the gate part, the resin material, the molding conditions, and the shape feature part (thin-walled part, thick-walled part, boss part, rib part). , Holes, distance from the surface, etc.) are entered. In the learning phase, the output values include the type of defect phenomenon (internal defects (sinks, nests) in the molded product, surface defects (weld, silver) in the molded product, defects in the entire molded product (short, warp), etc. ), And the defect rate, etc. are output. Then, in the learning phase, the profile pattern and the molding defect are related. Profile patterns and molding defects related by the learning phase are used in predicting molding defects.

In the operation phase, a mold model for calculation is created for any shape. Then, a simulation of resin injection molding is executed based on the created mold model. Based on the execution simulation result, the flow path (trajectory) to an arbitrary point (attention position) is calculated, and the profile of the characteristic value (state in the flow locus of the resin) to the gate portion is calculated. This will calculate the profile at each position for any shape.

Here, in the present embodiment, in the operation phase, the flow locus until the resin at the time of resin molding at the attention position of the resin molded product reaches the attention position is calculated, and attention is paid based on the state in the flow locus of the resin. Predict position molding defects.

Specifically, the states in the flow locus of the resin include wall thickness, temperature, flow velocity of the resin, elastic modulus, pressure, flow front speed, resin pressure, air temperature in contact with the flow front, gate passage speed, runner and gate. Based on at least one of the cross-sectional area, the flow velocity vector, the shrinkage strain, the solidification timing, and the anisotropic physical characteristics, the molding defect at the position of interest is predicted. Specifically, the states in the flow locus of the resin include wall thickness, temperature, flow velocity of the resin, elastic modulus, pressure, flow front speed, resin pressure, air temperature in contact with the flow front, gate passage speed, runner and gate disconnection. Prediction of molding defects at the position of interest based on changes in at least one of area, flow velocity vector, shrinkage strain, solidification timing and anisotropic physical properties.

For example, when predicting a molding defect at the position of interest as a state in the flow locus based on a change in wall thickness, it is based on a sudden change in wall thickness in the flow locus, the frequency of increase / decrease in wall thickness, and the like. It is predicted that there will be molding defects at the position of interest. When the wall thickness changes greatly and suddenly, it corresponds to the location of the expanding flow and the contracting flow on the trajectory of the actual flow. In this case, molding defects may occur at the downstream attention position due to the expanding or contracting flow.

In addition, when predicting molding defects at the position of interest based on changes in temperature as the state in the flow locus, the position of interest is based on the sudden temperature change in the flow locus, the frequency of temperature increase / decrease changes, and the like. Molding defects are predicted.

In addition, when predicting a molding defect at the attention position based on the change in the flow velocity as the state in the flow locus, the attention position is based on the sudden change in the flow velocity in the flow locus, the frequency of increase / decrease in the flow velocity, and the like. Molding defects are expected.

For example, in the case of the molding model 10 as shown in FIG. 2, the state in the flow locus from the gate portion 11 to which the resin is supplied to the attention position 13 is acquired. In the example shown in FIG. 2, the position where the weld occurs is set as the attention position 13. The attention position 13 can be set to any position. In the example shown in FIG. 3, the relationship between the flow locus from the start point 12 of the gate portion 11 to the attention position 13 and the wall thickness is shown. In the example shown in FIG. 4, the relationship between the flow locus from the start point 12 of the gate portion 11 to the attention position 13 and the temperature is shown. In the example shown in FIG. 5, the relationship between the flow locus from the start point 12 of the gate portion 11 to the attention position 13 and the flow velocity (linear velocity) is shown.

As shown in FIG. 6, when the horizontal axis is time, the vertical axis is the resin temperature or elastic modulus, and the attention position is the thick portion, a defect phenomenon occurs in the thick portion by confirming the change in the physical quantity at the attention position. Predict the occurrence of sink marks or nests.

When the horizontal axis is the flow length, the vertical axis is the wall thickness, temperature or pressure, and the position of interest is the thin part, the change in physical quantity at the position of interest is confirmed, and the change before and after the resin reaches the position of interest. By confirming, it is predicted that a short shot as a defective phenomenon will occur in the thin-walled portion. The flow length may be the distance from the gate.

When the horizontal axis is the flow length, the vertical axis is the flow front velocity or the resin pressure, and the attention position is the surface portion, the surface portion is defective by confirming the change before and after the resin reaches the attention position. It is predicted that surface gloss will occur as a phenomenon. When the horizontal axis is the flow length, the vertical axis is the resin temperature or the flow velocity of the resin, and the attention position is the surface portion, the surface portion is defective by confirming the change before and after the resin reaches the attention position. It is predicted that the floating of fibers will occur as a phenomenon.

In addition, when the horizontal axis is the flow length, the vertical axis is the resin temperature or the flow front speed, and the attention position is the surface part, the surface part is defective by confirming the change before and after the resin reaches the attention position. It is predicted that the flow mark as a phenomenon will occur. When the horizontal axis is the flow length, the vertical axis is the air temperature in contact with the flow front portion, and the attention position is the surface portion, the change in physical quantity at the attention position is confirmed, and before and after the resin reaches the attention position. By confirming the change, it is predicted that burning as a defective phenomenon will occur on the surface portion.

When the horizontal axis is time, the vertical axis is the gate passage speed or the cross-sectional area between the runner and the gate, and the attention position is the gate, the change in physical quantity at the attention position is confirmed, and the resin is at the attention position. By confirming the change before and after reaching, it is predicted that jetting as a defective phenomenon will occur at the gate portion.

When the horizontal axis is the flow length, the vertical axis is the resin temperature, the resin pressure or the flow velocity vector, and the attention position is the confluence of the flow, the change in the physical quantity at the attention position is confirmed and the resin reaches the attention position. By confirming the change before and after the operation, it is predicted that a weld line as a defective phenomenon will occur at the confluence of the flow.

When the horizontal axis is the flow length, the vertical axis is the resin temperature, and the attention position is the edge part, the change in the physical quantity at the attention position and the change before and after the resin reaches the attention position should be confirmed. Therefore, it is predicted that an ear flow as a defective phenomenon will occur at the edge portion.

In addition, when the horizontal axis is the flow length, the vertical axis is the shrinkage strain, solidification timing or anisotropic physical properties, and the position of interest is the position of the center of gravity or the gate of the model, the entire calculation model including the position of interest can be confirmed. , It is predicted that warp deformation as a defective phenomenon will occur at the center of gravity position or the gate position of the model.

When a molding defect is predicted at the position of interest, it is possible to improve the state of the flow locus so that the molding defect does not occur. For example, in the part where the wall thickness, temperature and flow velocity change suddenly, it is easy to cause molding defects, so the wall thickness of the molded product may be changed or the position of the gate part may be changed so as to eliminate the sudden change part. do. In addition, since the meeting angle of the weld is also affected by the temperature and pressure of the resin that has reached that point, it is possible to determine at which position the resin flow is bad for the weld.

In addition, since the occurrence of molding defects can be predicted from the shape of the curve, which is the relationship between the flow locus and the wall thickness, temperature, and flow velocity, the occurrence of molding defects can be suppressed by improving the shape of the curve. Is possible.

Further, in the present embodiment, as shown in FIG. 7, the resin in which the molding defect occurs is based on the flow locus of the resin at the time of resin molding at the position where the molding defect has occurred and the state in the resin flow locus. The state in the flow locus is machine-learned, and the molding defect at the position of interest is predicted based on the result of the machine learning and the state in the flow locus of the resin.

Further, in the present embodiment, as shown in FIG. 8, the flow characteristics of the resin obtained by simulating resin molding with a predetermined mold and the resin obtained by simulating resin molding with an arbitrary mold are performed. Based on the flow characteristics, the flow characteristics of the resin with respect to the shape of the mold are machine-learned, and based on the result of the machine learning, the flow characteristics of the resin up to the position of interest are predicted, and the predicted flow characteristics and the flow of the resin are predicted. Based on the state in the locus, the molding defect at the position of interest is predicted.

The flow characteristics of the resin include the flow length and the flow ratio of the resin. Further, the predetermined mold is, for example, a spiral flow mold having a constant wall thickness. Here, even when the resin is supplied under the same pressure, if the shape of the mold is different, the flow length (flow ratio) is different even with the same resin and the same molding conditions. Therefore, we would like to obtain the value of the flow length (flow ratio) when the numerical value of the flow ratio presented by the resin manufacturer based on the resin catalog value is molded in an arbitrary mold. In other words, the resin manufacturer uses a spiral flow mold, which is a spiral mold, in order to clearly indicate the flow length (flow ratio), which is an index of moldability, and keeps the molding conditions constant. It is presented as an index of the moldability of the resin by the length (flow length) of whether the resin flows or the ratio of the length to the wall thickness (flow ratio). Therefore, under the same material and the same molding conditions, the longer the flow length, the easier the resin to flow. Molding makers and analysis engineers use the flow length (flow ratio) of resin makers as a material for determining whether molding is possible. However, since the flow resistance differs between the spiral flow mold and the mold molded in-house, it cannot be judged from the numbers as they are. Therefore, it is desirable to obtain a numerical value of the flow length that can be used as a judgment index when molding with an in-house mold. For this purpose, as shown in FIG. 9, any gold can be obtained by inputting the volume, area, average wall thickness, flow locus, etc. of the flow area flowing at the same pressure and outputting the flow length (flow ratio). The value of the flow length (flow ratio) when molded in the mold is acquired.

(Machine learning)
An example of machine learning used in the learning phase will be described.

Machine learning is performed by methods such as regression, decision tree, neural network, Bayes, clustering, and ensemble learning.

(Forecast processing of flow length (flow ratio))
The process of predicting the flow length (flow ratio) will be described with reference to FIGS. 9 to 11. The flow length (flow ratio) prediction process is executed by the computer 1 (processor 2).

In the learning phase shown in FIG. 10, in step S1, actually measured values of the flow length (flow ratio) for each resin, molding conditions, and spiral flow specifications are acquired. In step S2, a molding simulation is performed under the same molding conditions and the same mold specifications as in step S1, and the analysis value of the flow length (flow ratio) is calculated.

In step S3, the liquidity results of the measured value of the flow length (flow ratio) acquired in step S1 and the analysis value of the flow length (flow ratio) calculated in step S2 are compared with the measured value. The correction coefficient with the analysis value is obtained.

In step S4, the analysis value (arbitrary model value) of the flow length (flow ratio) is calculated by performing the analysis by matching the resin and the molding conditions with an arbitrary mold model. In step S5, the analysis value of the flow length (flow ratio) and the analysis value (arbitrary model value) of the flow length (flow ratio) are compared, and the difference in the shape characteristics of the flow range and the difference in the flow length are obtained. Are compared and accumulated as machine learning data. Then, the processes of steps S1 to S5 are repeated for various resins, molds, and molding conditions, and a plurality of data are accumulated in the database.

In the operation phase shown in FIG. 11, in step S11, molding simulation is performed with any resin, any mold, and any molding conditions. In step S12, based on the simulation result, the position up to the same position as the flow length (spiral flow value) of the resin is displayed as the flowable position.

(Effect of this embodiment)
Next, the effect of this embodiment will be described.

In the present embodiment, as described above, molding defects in the region of interest can be predicted based on the flow trajectory until the resin reaches the position of interest, so that a plurality of analysis results in the region of interest can be determined from multiple aspects. Therefore, it is not necessary to determine the occurrence of molding defects. Thereby, it is possible to easily predict the internal defect (sink, nest) of the molded product, the surface defect of the molded product, and the defect (warp) of the entire molded product.

Further, in the present embodiment, as described above, the states in the flow locus of the resin include the wall thickness, the temperature, the flow velocity of the resin, the elastic modulus, the pressure, the flow front speed, the resin pressure, the air temperature in contact with the flow front, and the gate. A molding defect at the position of interest is predicted based on at least one of the passage speed, the cross-sectional area of the runner and the gate, the flow velocity vector, the shrinkage strain, the solidification timing, and the anisotropic physical characteristics. As a result, the wall thickness of the place where the resin passes, the temperature state when the resin flows, the state of the flow velocity of the resin, the elastic characteristic, the pressure, the flow front speed, the resin pressure, the air temperature in contact with the flow front, and the gate passage speed. , The state of the resin in the region of interest, which is the final destination, can be determined based on the cross-sectional area of the runner and gate, the flow velocity vector, the shrinkage strain, the solidification timing, or the anisotropic physical characteristics. It is possible to easily predict the molding defect of the resin at the attention position which is the final arrival position.

Further, in the present embodiment, as described above, the states in the flow locus of the resin include the wall thickness, the temperature, the flow velocity of the resin, the elastic modulus, the pressure, the flow front speed, the resin pressure, the air temperature in contact with the flow front, and the gate. Prediction of molding defects at the position of interest based on changes in at least one of the pass-through speed, runner and gate cross-sectional area, flow velocity vector, shrinkage strain, solidification timing and anisotropic physical properties. As a result, the wall thickness during flow, temperature, sudden change in resin flow velocity, elastic modulus, pressure, flow front speed, resin pressure, air temperature in contact with flow front, gate passage speed, runner and gate cross-sectional area, Based on the flow velocity vector, shrinkage strain, solidification timing, and anisotropic physical properties, it is possible to easily predict the molding failure of the resin that occurs at the downstream attention position.

Further, in the present embodiment, as described above, the flow locus of the resin in which the molding defect occurs is based on the flow locus of the resin at the time of resin molding at the position where the molding defect occurs and the state in the resin flow locus. The state inside is machine-learned, and the molding defect at the position of interest is predicted based on the result of the machine learning and the state in the flow locus of the resin. As a result, the relationship between the state in the flow locus of the resin and the state of the position of interest can be accumulated by machine learning, so that molding defects in the region of interest can be easily and accurately performed based on the state in the flow locus of the resin. It can be predicted well.

Further, in the present embodiment, as described above, the flow characteristics of the resin obtained by simulating the resin molding with a predetermined mold and the flow characteristics of the resin obtained by simulating the resin molding with an arbitrary mold. Based on the above, the flow characteristics of the resin with respect to the shape of the mold are machine-learned, and based on the result of the machine learning, the flow characteristics of the resin up to the position of interest are predicted, and the predicted flow characteristics and the flow trajectory of the resin are included. Based on the state of, the molding defect of the attention position is predicted. As a result, even if there is an error between the flow characteristics of a predetermined mold and the flow characteristics of an arbitrary mold, the flow characteristics of the resin up to the position of interest in any mold based on the results of machine learning. Can be predicted accurately.

(Modification example)
It should be noted that the embodiments disclosed this time are exemplary in all respects and are not considered to be restrictive. The scope of the present invention is shown by the scope of claims rather than the description of the above-described embodiment, and further includes all modifications (modifications) within the meaning and scope equivalent to the scope of claims.

For example, in the above embodiment, an example of a configuration in which a state in a flow locus of a resin in which a molding defect occurs is accumulated by machine learning is shown, but the present invention is not limited to this. In the present invention, the state in the flow locus of the resin in which the molding defect occurs may be accumulated based on the actually measured data. Further, machine learning may be performed and the data may be accumulated so as to complement the actually measured data.

Further, in the present invention, the plurality of data may be stored in a computer that estimates the data, or may be stored in a device such as an external server provided separately from the computer that estimates the data.

Further, in the above embodiment, for convenience of explanation, the processing operation of the computer is described by using a flow-driven flowchart in which the processing operations are sequentially performed along the processing flow, but the present invention is not limited to this. In the present invention, the processing operation of the computer may be performed by event-driven type (event-driven type) processing in which the processing is executed in event units. In this case, it may be completely event-driven, or it may be a combination of event-driven and flow-driven.

1 Computer 3a Program 7 Recording medium

Claims (7)

The flow locus until the resin at the time of resin molding at the attention position of the resin molded product reaches the attention position is calculated.
A resin molding analysis method for predicting molding defects at the position of interest based on the state in the flow locus of the resin. The states in the flow locus of the resin include wall thickness, temperature, resin flow velocity, elastic modulus, pressure, flow front speed, resin pressure, air temperature in contact with the flow front, gate passage speed, runner and gate cross-sectional area, and flow velocity. The resin molding analysis method according to claim 1, wherein a molding defect at the position of interest is predicted based on at least one of a vector, shrinkage strain, solidification timing, and anisotropic physical characteristics. The states in the flow locus of the resin include wall thickness, temperature, resin flow velocity, elastic modulus, pressure, flow front speed, resin pressure, air temperature in contact with the flow front, gate passage speed, runner and gate cross-sectional area, and flow velocity. The resin molding analysis method according to claim 2, wherein a molding defect at the position of interest is predicted based on a change in at least one of a vector, a shrinkage strain, a solidification timing, and an anisotropic physical property. Based on the resin flow locus at the time of resin molding at the position where the molding defect occurred and the state in the resin flow locus, the state in the resin flow locus where the molding defect occurs is machine-learned.
The resin molding analysis method according to any one of claims 1 to 3, which predicts a molding defect at the position of interest based on the result of machine learning and the state in the flow locus of the resin. Based on the resin flow characteristics obtained by simulating resin molding with a predetermined mold and the resin flow characteristics obtained by simulating resin molding with an arbitrary mold, the resin with respect to the shape of the mold Machine learning the flow characteristics,
A claim that predicts the flow characteristics of the resin up to the attention position based on the result of machine learning, and predicts the molding defect of the attention position based on the predicted flow characteristics and the state in the flow locus of the resin. Item 6. The resin molding analysis method according to any one of Items 1 to 4. A program for causing a computer to execute the resin molding analysis method according to any one of claims 1 to 5. A recording medium on which the program according to claim 6 is recorded and readable by a computer.

Images