Crosslinking reaction simulation device, and internal bubble estimation device

Abstract

Provided is a crosslinking reaction simulation device (1) comprising: a storage part (2); and a set conditions analysis part (8) that analyzes conditions set for a crosslinking reaction of a polymer part (12) of a target work model (WM), wherein the storage part (2) stores a polymer characteristic value (PM), target work model elastic modulus (Ks), and relationship data (CD) in which the conditions set for the crosslinking reaction of the polymer part (12) and the reaction time of the crosslinking reaction of the polymer part (12) are linked, and the set conditions analysis part (8) is provided with a polymer characteristic value acquisition part (181), a target work model elasticity acquisition part (182), a relationship data acquisition part (183), and a correction part (184) that corrects the relationships among the target work model elasticity (Ks), the set conditions, and the reaction time in the relationship data (CD) on the basis of the polymer characteristic value (PV).

Description

Crosslinking reaction simulation device and internal bubble estimation device

[0001] The present disclosure relates to a crosslinking reaction simulation device and an internal bubble estimation device.

[0002] Conventionally, the physical properties of chemical substances have been improved by crosslinking polymers. Rubber is an example of such a polymer. Rubber is made by adding sulfur or other crosslinking agents, vulcanization accelerators, etc. to raw rubber and then heating it, which forms a three-dimensional network of crosslinked structures between or within the rubber molecular chains.

[0003] A technique for estimating the degree of vulcanization of rubber, i.e., the reaction rate of the crosslinking reaction of rubber, is described in Non-Patent Document 1. The technique estimates the degree of vulcanization from the temperature history of rubber based on the Arrhenius equation, and estimates appropriate vulcanization conditions (mold temperature, vulcanization time).

[0004] Toshio Arimatsu, "Practical Design of Vulcanization Processes," Journal of the Society of Rubber Industry of Japan, Vol. 59, No. 3, (1986)

[0005] It is desirable to apply the above technology to crosslinking reactions of not only rubber but also resins, elastomers, etc. The reaction rate of the crosslinking reaction varies depending on the properties of the raw polymer. Therefore, in order to estimate appropriate setting conditions, it is conceivable to estimate the setting conditions using parameters that vary depending on the properties of the raw polymer. However, since the above technology does not include parameters that vary depending on the properties of the raw polymer, it is not possible to accurately estimate the reaction rate of the crosslinking reaction. Therefore, there is a problem in that it is difficult to appropriately estimate the setting conditions of the crosslinking reaction.

[0006] The present disclosure has been made in view of the above background, and aims to provide a crosslinking reaction simulation device and an internal bubble estimation device that are capable of appropriately estimating the setting conditions for the crosslinking reaction.

[0007] One aspect of the present disclosure is a crosslinking reaction simulation device comprising: a memory unit that stores data used in the simulation; and a setting condition analysis unit that analyzes setting conditions related to the crosslinking reaction of a polymer portion of a target work model, wherein the memory unit stores: a polymer characteristic value that is a characteristic value related to a raw material polymer that constitutes the polymer portion of the target work model; a target work model elastic modulus that is the elastic modulus of the polymer portion of the target work model; relationship data that correlates a set temperature in the crosslinking reaction of the polymer portion of the target work model having the target work model elastic modulus; and a reaction time in the crosslinking reaction of the polymer portion of the target work model having the target work model elastic modulus; and the setting condition analysis unit comprises: a polymer characteristic value acquisition unit that acquires the polymer characteristic value; a target work model elastic modulus acquisition unit that acquires the target work model elastic modulus; a relationship data acquisition unit that acquires the relationship data; and a correction unit that corrects the relationship between the target work model elastic modulus, the set temperature, and the reaction time in the relationship data based on the polymer characteristic value.

[0008] According to one aspect of the present disclosure, the setting conditions for the crosslinking reaction of the polymer portion can be calculated from the polymer characteristic values ​​of the raw polymer that is the target of the crosslinking reaction to obtain a product having the desired target work model elastic modulus.

[0009] Note that the symbols in parentheses in the claims indicate the correspondence with the specific means described in the embodiments described below, and do not limit the technical scope of the present disclosure.

[0010] 1-1。 FIG. 1-1 is a block diagram showing a crosslinking reaction simulation device of embodiment 1-1. FIG. 1-2 is a schematic diagram showing details of a storage unit according to embodiment 1-1. FIG. 1-3 is a diagram showing a state in which a target workpiece model is placed on a mold model in embodiment 1-1. FIG. 1-4 is a diagram showing a target workpiece model in embodiment 1-1. FIG. 1-5 is a flowchart showing the operation of a crosslinking reaction simulation device of embodiment 1-1. FIG. 1-6 is a diagram showing changes over time in temperature of each part in a state in which a target workpiece model is placed on a mold model in embodiment 1-1. FIG. 1-7 is a graph showing actual measured values ​​and predicted values ​​of the temperature of a polymer part in the prior art. FIG. 1-8 is a graph showing actual measured values ​​and predicted values ​​of the temperature of a polymer part in embodiment 1-1. FIG. 1-9 is a graph showing the amount of change in thermal conductivity with respect to the carbon black content in embodiment 1-1. FIG. 1-1 is a diagram for explaining the state of an analysis mesh in a state in which a target workpiece model is placed on a mold model in embodiment 1-1. FIG. 1-1 is a schematic diagram showing the structure of a vulcanization tester of embodiment 1-1. 1-1. (a) is a graph showing the change in torque with respect to reaction time, and (b) is a graph showing the change in reaction rate with respect to reaction time. (b) is a graph showing the change in reaction rate with respect to reaction time. (c) is a graph showing the calculation method of a slope coefficient in the crosslinking reaction analysis process of embodiment 1-1. (d) is a graph showing the calculation method of an equivalent reaction amount in the crosslinking reaction analysis process of embodiment 1-1. (e) is a graph showing the calculation method of an equivalent reaction amount from an equivalent reaction amount increase amount in the crosslinking reaction analysis process of embodiment 1-1. (f) is a graph showing the conversion method from an equivalent reaction amount increase amount to a reaction rate in the crosslinking reaction analysis process of embodiment 1-1. (f) is a graph showing the actual measured value and predicted value of the reaction rate of the polymer portion in the prior art. (g) is a graph showing the method of dividing the crosslinking reaction according to the reaction progress degree and calculating the activation energy for each division in embodiment 1-1. (h) is a graph showing the actual measured value and predicted value of the reaction rate of the polymer portion in embodiment 1-1. 1 is a diagram for explaining a method of expressing the amount of change in reaction rate with respect to reaction time in embodiment 1-1. FIG. 2 is a diagram showing the amount of change in reaction rate with respect to reaction time in embodiment 1-1. FIG. 3 is a diagram showing a second function in embodiment 1-1.It is a figure for explaining the state of the analysis mesh of the target work model in Embodiment 1-1. It is a figure showing a state where internal bubbles are generated in the polymer part. It is a figure showing the reaction rate at the blow point, the reaction rate at the time of demolding, and the final reaction rate in the crosslinking reaction. (a) is a graph showing the reaction rates at the blow points of a plurality of samples. (b) is a graph showing the torque measured by a vulcanization tester at the blow points of a plurality of samples. It is a figure for explaining the mechanism for suppressing internal bubbles. In Embodiment 1-1, it is a heat map showing the reaction rate distribution of the polymer part. In Embodiment 1-1, it is a figure showing the change amount of the elastic modulus in the direction orthogonal to the axis of the target work model with respect to the reaction time when the mold temperature is low. In Embodiment 1-1, it is a figure showing the change amount of the elastic modulus in the direction orthogonal to the axis of the target work model with respect to the reaction time when the mold temperature is medium. In Embodiment 1-1, it is a figure showing the change amount of the elastic modulus in the direction orthogonal to the axis of the target work model with respect to the reaction time when the mold temperature is high. In Embodiment 1-1, it is a figure showing the changes of the elastic modulus of the polymer part and the generation state of internal bubbles with respect to the reaction time. In Embodiment 1-1, it is a figure visually displaying the predicted results of the generation of internal bubbles. In Embodiment 1-1, it is a figure showing another example of the molding die model. It is a graph representing the change of the elastic modulus Ks of the target work model in Embodiment 1-1. (a) is a graph representing the change of the elastic modulus Ks of the target work model with respect to the set temperature of the crosslinking reaction. (b) is a graph showing the change of the elastic modulus Ks of the target work model with respect to the polymer characteristic value RKs. In Embodiment 1-1, it is a graph showing the process of calibrating the simulation results with the measured data. (a) is a simulation result showing the change of the elastic modulus with respect to the set temperature. (b) is the measured data showing the change of the elastic modulus with respect to the set temperature. (c) is a graph showing the state where the peak of the graph of the simulation result and the peak of the graph of the measured data are made to coincide. In Embodiment 1-1, it is a figure showing the state where the graph obtained in FIG. 36(c) is shifted to different values of RKs.1-3 is a diagram showing a comparison of simulation results and actual measurement data for different RKs in embodiment 1-1, where (a) is a comparison of simulation results and actual measurement data when RKs = 143, and (b) is a comparison of simulation results and actual measurement data when RKs = 178.

[0034] FIG. 1-4 is a diagram showing a method for adjusting the elastic modulus by adjusting the set temperature of the crosslinking reaction in embodiment 1-1.

[0035] FIG. 1-5 is a graph showing a process for calibrating simulation results with actual measurement data in embodiment 1-2, where (a) is a simulation result showing changes in elastic modulus with respect to reaction time, (b) is actual measurement data showing changes in elastic modulus with respect to reaction time, and (c) is a graph showing a state in which the peaks of the graph of the simulation results and the peaks of the graph of the actual measurement data are aligned.

[0036] FIG. 1-6 is a diagram showing a method for calibrating simulation results to actual measurement data by shifting the set temperature and reaction time using a table in embodiment 1-3.

[0037] FIG. 1-7 is a diagram showing a method for calibrating simulation results to actual measurement data by shifting the set temperature and reaction time using a table in embodiment 1-4.

[0038] 1-5 is a diagram showing a method for calibrating simulation results to match actual measurement data by shifting the set time using a table in embodiment 1-5. FIG. 1-6 is a diagram showing a method for calibrating simulation results to match actual measurement data by shifting the set temperature, reaction time, and elastic modulus using a table in embodiment 1-6. FIG. 1-7 is a diagram showing a method for calibrating simulation results to match actual measurement data by changing the value of the elastic modulus of a specific cell using a table. FIG. 1-8 is a block diagram showing a crosslinking reaction simulation device according to embodiment 2-1. FIG. 1-9 is a diagram showing details of a storage unit according to embodiment 2-1. FIG. 1-10 is a flowchart showing a main flow of a crosslinking reaction simulation device according to embodiment 2-1. FIG. 1-11 is a flowchart showing a simulation process according to embodiment 2-1. FIG. 1-12 is a part of a flowchart showing a prediction process according to embodiment 2-1. FIG. 1-13 is a diagram showing a correlation between a reaction rate obtained by simulation processing and a reaction rate predicted by machine learning in embodiment 2-1. FIG. 1-14 is a part of a flowchart showing a prediction process according to embodiment 2-2.1 is a part of a flowchart showing a preheating process according to embodiment 2-2. FIG. 2 is a part of a flowchart showing a preheating process according to embodiment 2-2. FIG. 3 is a part of a flowchart showing a preheating process according to a modified embodiment of embodiment 2-2. FIG. 4 is a part of a flowchart showing a preheating process according to a modified embodiment of embodiment 2-2. FIG. 5 is a flowchart showing a prediction process according to embodiment 2-3. FIG. 6 is a diagram showing the correlation between a reaction rate obtained by a simulation process and a reaction rate predicted by machine learning in embodiment 2-3. FIG. 7 is a part of a flowchart showing a prediction process according to embodiment 2-4. FIG. 8 is a part of a flowchart showing a preheating process according to embodiment 2-4. FIG. 9 is a part of a flowchart showing a prediction process according to embodiment 3-1. FIG. 10 is a flowchart showing a prediction process according to embodiment 3-2. FIG. 11 is a part of a flowchart showing a prediction process according to embodiment 3-3. FIG. 12 is a flowchart showing a prediction process according to embodiment 3-4.

[0011] (Embodiment 1-1) 1. Configuration of crosslinking reaction simulator 1 1-1. Overall configuration of crosslinking reaction simulator 1 The overall configuration of a crosslinking reaction simulator 1 (an example of an internal bubble estimation device) of embodiment 1-1 will be described with reference to Fig. 1. The crosslinking reaction simulator 1 of this embodiment simulates a crosslinking reaction that crosslinks molecular chains of raw material polymers, and predicts the setting conditions for the crosslinking reaction.

[0012] The raw polymer is not particularly limited as long as its molecular chains are capable of undergoing a crosslinking reaction. Any polymer can be appropriately selected, including thermosetting resins such as phenolic resin, epoxy resin, melamine resin, urea resin, unsaturated polyester resin, alkyd resin, silicone resin, and polyurethane; thermoplastic resins such as crosslinked polyethylene and crosslinked polypropylene; synthetic rubbers such as natural rubber, styrene-butadiene rubber, butadiene rubber, isoprene rubber, ethylene-propylene-diene rubber, fluororubber, acrylonitrile-butadiene rubber, chloroprene rubber, butyl rubber, and silicone rubber; and elastomers. In this embodiment, rubbers including natural rubber and synthetic rubber are used as the polymer. In the case of rubber, when heated with the addition of a crosslinking agent such as sulfur, the molecular chains that make up the rubber crosslink, causing a so-called vulcanization reaction.

[0013] As shown in FIG. 1A, the crosslinking reaction simulation device 1 includes a memory unit 2, a heat transfer analysis unit 3, a crosslinking reaction analysis unit 4, a structural analysis unit 5, an internal bubble estimation unit 6, a display unit 7, and a setting condition analysis unit 8.

[0014] 1b, the storage unit 2 stores data used in the simulation. The data includes a forming mold model MM, a target workpiece model WM, heat transfer analysis data TD used by the heat transfer analysis unit 3, crosslinking reaction analysis data RD used by the crosslinking reaction analysis unit 4, internal bubble estimation data BD used by the internal bubble estimation unit 6, and setting condition analysis data SD used by the setting condition analysis unit 8.

[0015] The molding die model MM will be described with reference to FIG. 2. As shown in FIG. 2, the molding die model MM includes a die 10 and a hot plate 11 attached to the die 10. The die 10 includes a lower die 10A located below and an upper die 10B attached to the lower die 10A from above. The hot plate 11 includes a lower hot plate 11A attached to the underside of the lower die 10A and an upper hot plate 11B attached to the upper surface of the upper die 10B. A lower cavity 100A opening upward is formed in the lower die 10A. An upper cavity 100B opening downward is formed in the upper die 10B. When the lower die 10A and the upper die 10B are assembled, a target workpiece model WM is placed in the space formed by the lower cavity 100A and the upper cavity 100B.

[0016] The target workpiece model WM will be described with reference to FIG. 3 . As shown in FIG. 3 , the target workpiece model WM includes a polymer portion 12. The polymer portion 12 includes a raw polymer. Furthermore, the polymer portion 12 may include additives such as antioxidants, carbon black, etc. The polymer portion 12 of this embodiment is configured to include a raw polymer and carbon black. The raw polymer is not particularly limited, and any material such as rubber or thermosetting resin can be appropriately selected. In this embodiment, the raw polymer is made of rubber. The polymer portion 12 of this embodiment is a rubber portion configured to exhibit vibration-damping performance, and the target workpiece model WM is a model of a vibration-damping rubber device. However, the polymer portion 12 may be configured to include a thermosetting resin as the raw polymer, or may not include carbon black.

[0017] In this embodiment, the polymer portion 12 is formed in a cylindrical shape along an axis A extending in the vertical direction. An outer joining member 13 (an example of a joining member) formed in a cylindrical shape extending in the vertical direction is disposed on the outer periphery of the polymer portion 12. The outer joining member 13 is made of metal, resin, or a composite of metal and resin. In this embodiment, the outer joining member 13 is made of metal and is joined to the outer periphery of the polymer portion 12. An inner joining member 14 (an example of a joining member) formed in a cylindrical shape extending in the vertical direction is disposed on the inner periphery of the polymer portion 12. The inner joining member 14 is made of metal, resin, or a composite of metal and resin. In this embodiment, the inner joining member 14 is made of metal and is joined to the inner periphery of the polymer portion 12. The vertical length of the inner joining member 14 is formed to be greater than the vertical length of the outer joining member 13. The upper and lower surfaces of the polymer portion 12 are formed to be concave. However, the shape of the polymer portion 12 is not limited to the above shape. Furthermore, the target workpiece model WM may be configured without both or either of the outer joining member 13 and the inner joining member 14 .

[0018] Returning to FIG. 2, with the target workpiece model WM placed on the mold model MM, gaps 15 are formed between the mold model MM and the outer joining member 13 and between the mold model MM and the inner joining member 14, respectively.

[0019] Returning to Figure 1, the heat transfer analysis data TD includes a thermal diffusivity characteristic TS that indicates the relationship between the mass ratio of carbon black to raw material polymer and the thermal diffusivity of the polymer portion 12, a contact heat transfer coefficient CH that is the heat transfer coefficient between the mold model MM and the outer joining member 13 and the inner joining member 14 when the target work model WM is placed in the mold model MM, an air heat transfer coefficient AC that is the heat transfer coefficient of the air surrounding the target work model WM when the target work model WM is demolded from the mold model MM, a temperature condition TC of the mold model MM input from a condition input unit 30 described later, and an outside air condition OC that includes the air temperature around the demolded target work model WM.

[0020] The crosslinking reaction analysis data RD defines an equivalent reaction amount as the ratio of the reaction amount of the crosslinking reaction at a target reaction temperature for a target reaction time to the reaction amount of the crosslinking reaction at a reference reaction temperature for a reference reaction time, and includes an equivalent reaction amount calculation model EM defined including a slope coefficient representing the slope of the Arrhenius plot, and a slope coefficient SC set according to the progress of the crosslinking reaction of the polymer portion 12. The crosslinking reaction analysis data RD further includes a first relationship data map DM1, a second relationship data map DM2, a first function F1, a second function F2, and a reference reaction curve RC, which will be described later.

[0021] The data BD for estimating internal bubbles includes a crosslinking reaction curve CC that defines the relationship between the elapsed time from the start of the crosslinking reaction and the torque, which is a value corresponding to the progress of the crosslinking reaction in the polymer part 12 of the target work model WM and can be measured by a crosslinking reaction characteristic tester using a test target polymer material corresponding to the polymer part 12.

[0022] The setting condition analysis data SD includes a polymer characteristic value PV and relationship data CD. The polymer characteristic value PV is a characteristic value related to the raw material polymer constituting the polymer portion 12 of the target workpiece model WM. The relationship data CD is data relating the target workpiece model elastic modulus Ks, which is the elastic modulus of the polymer portion 12 of the target workpiece model WM, the set temperature for the crosslinking reaction of the polymer portion 12 of the target workpiece model WM having the target workpiece model elastic modulus Ks, and the reaction time for the crosslinking reaction of the polymer portion 12 of the target workpiece model having the target workpiece model elastic modulus Ks. However, the target workpiece model elastic modulus Ks may be the elastic modulus in the axial direction (direction of axis A) of the target workpiece model WM, or may be the elastic modulus in the direction perpendicular to the axis (direction perpendicular to axis A) of the target workpiece model WM.

[0023] 1-3. Configuration of the heat transfer analysis unit 3 The heat transfer analysis unit 3 performs heat transfer analysis during a crosslinking reaction of the polymer portion 12 of the target workpiece model WM. As shown in FIG. 1, the heat transfer analysis unit 3 includes a condition input unit 30, a polymer thermal diffusivity determination unit 31, and an analysis unit 32.

[0024] The mass ratio of carbon black to raw material polymer in the target workpiece model WM and the temperature condition TC of the mold model MM are input to the condition input unit 30. The condition input unit 30 may be an input device such as a keyboard, mouse, trackball, or joystick, or may be an external storage medium such as a semiconductor memory or hard disk memory.

[0025] The polymer thermal diffusivity determination unit 31 determines the polymer thermal diffusivity, which is the thermal diffusivity of the polymer part 12 of the target work model WM, based on the mass ratio of carbon black input by the condition input unit 30 and the thermal diffusivity characteristic TS stored in the memory unit 2.

[0026] The analysis unit 32 performs a heat transfer analysis during the crosslinking reaction of the polymer part 12 of the target work model WM using the polymer thermal diffusivity determined by the polymer thermal diffusivity determination unit 31 and the temperature conditions TC stored in the memory unit 2, with the target work model WM placed on the molding die model MM.

[0027] 1-4. Configuration of the crosslinking reaction analysis unit 4 The crosslinking reaction analysis unit 4 analyzes the reaction rate of the crosslinking reaction of the polymer portion 12 using the results of the heat transfer analysis by the heat transfer analysis unit 3. As shown in FIG. 1 , the crosslinking reaction analysis unit 4 includes a temperature acquisition unit 40 and a reaction rate calculation processing unit 41.

[0028] The temperature acquisition unit 40 acquires the temperature at each time point for each element of the polymer part 12 of the target workpiece model WM in the crosslinking reaction as a result of the heat transfer analysis by the heat transfer analysis unit 3 .

[0029] The reaction rate calculation processing unit 41 calculates the equivalent reaction amount of the polymer portion 12 at each time based on the temperature at each time for each element of the polymer portion 12 in the acquired crosslinking reaction, the equivalent reaction amount calculation model EM, and the slope coefficient SC according to the progress of the crosslinking reaction at the target time, and calculates the reaction rate of the crosslinking reaction of the polymer portion 12 based on the calculated equivalent reaction amount.

[0030] 1-5. Configuration of the structural analysis unit 5 The structural analysis unit 5 performs structural analysis using the reaction rate of the crosslinking reaction of the polymer portion 12 analyzed by the crosslinking reaction analysis unit 4. As shown in Fig. 1, the structural analysis unit 5 includes a temperature acquisition unit 50, a reaction rate acquisition unit 51, an elastic modulus assignment unit 52, and a characteristic acquisition unit 53. However, the temperature acquisition unit 50 may be omitted.

[0031] The temperature acquisition unit 50 acquires, as a result of the heat transfer analysis by the heat transfer analysis unit 3, the temperature at each time point for each element of the polymer part 12 of the target workpiece model WM in the crosslinking reaction.

[0032] The reaction rate acquisition unit 51 acquires the reaction rate of each element of the polymer portion 12 calculated by the reaction rate calculation processing unit 41 of the crosslinking reaction analysis unit 4 .

[0033] The elastic modulus assigning unit 52 assigns an elastic modulus to the polymer portion 12 according to the obtained reaction rate.

[0034] The characteristic acquisition unit 53 acquires the characteristics of the target workpiece model WM by performing a structural analysis with the elastic modulus assigned to the polymer portion 12 .

[0035] 1-6. Configuration of Internal Bubbles Estimation Unit 6 The internal bubbles estimation unit 6 is applied to a crosslinking reaction process in which the polymer portion 12 of the target workpiece model WM is crosslinked in the mold model MM, and then the mold model MM is demolded, and estimates the generation of internal bubbles inside the polymer portion 12 of the target workpiece model WM that accompanies the demolding of the mold model MM. As shown in FIG. 1 , the internal bubbles estimation unit 6 includes a reaction rate acquisition unit 60, a torque calculation unit 61, and an estimation unit 62.

[0036] The reaction rate acquisition unit 60 acquires the reaction rate of the cross-linking reaction of the polymer portion 12 of the target workpiece model WM. The reaction rate is the reaction rate at the time when the mold 10 is opened, i.e., when the mold clamping pressure is released. When the polymer portion 12 is rubber, the reaction rate is also called the degree of vulcanization at demolding.

[0037] The torque calculation unit 61 calculates the torque (also called the demolding torque) corresponding to the acquired reaction rate based on the reaction rate (demolding reaction rate) acquired by the reaction rate acquisition unit 60 and the crosslinking reaction curve CC stored in the memory unit 2.

[0038] The estimation unit 62 compares the torque calculated by the torque calculation unit 61 (torque at demolding) with the blow point torque (described in detail later) calculated from the blow point vulcanization degree (described in detail later) in the polymer portion 12, and estimates whether or not bubbles will occur inside the polymer portion 12.

[0039] 1-7. Configuration of Display Unit 7 The display unit 7 displays a value obtained from the reaction rate according to the reaction time, based on the analysis results of the crosslinking reaction analysis unit 4. Furthermore, the display unit 7 displays characteristics according to the temperature of the mold model MM used in the crosslinking reaction of the polymer portion 12 and the in-mold reaction time from the start of the crosslinking reaction to the demolding of the mold model MM, based on the results of the structural analysis of the structural analysis unit 5.

[0040] The display unit 7 also displays the presence or absence of internal bubbles depending on the temperature of the mold model MM and the in-mold reaction time from the start of the crosslinking reaction to the demolding of the mold model MM, based on the estimation results of the internal bubble estimation unit 6. The display unit 7 also displays a value obtained from the reaction rate depending on the temperature of the mold model MM and the in-mold reaction time from the start of the crosslinking reaction to the demolding of the mold model MM, based on the analysis results of the crosslinking reaction analysis unit 4, along with the presence or absence of internal bubbles. The display unit 7 also displays the characteristics depending on the temperature of the mold model MM and the in-mold reaction time from the start of the crosslinking reaction to the demolding of the mold model MM, along with the presence or absence of internal bubbles, based on the results of the structural analysis of the structural analysis unit 5.

[0041] 1-8. Configuration of the setting condition analysis unit 8 The setting condition analysis unit 8 analyzes the setting conditions for the crosslinking reaction of the polymer portion 12 of the target workpiece model WM. The setting condition analysis unit 8 includes a polymer characteristic value acquisition unit 181, a target workpiece model elastic modulus acquisition unit 182, a relationship data acquisition unit 183, and a correction unit 184.

[0042] The polymer characteristic value acquisition unit 181 acquires the polymer characteristic value PV. The target workpiece model elastic modulus acquisition unit 182 acquires the target workpiece model elastic modulus Ks. The relationship data acquisition unit 183 acquires the relationship data CD. The correction unit 184 corrects the relationship between the target workpiece model elastic modulus Ks, the set temperature, and the reaction time in the relationship data CD based on the polymer characteristic value PV.

[0043] 2. Overall Operation of the Crosslinking Reaction Simulator 1 The overall operation of the crosslinking reaction simulator 1 according to this embodiment will be described with reference to Fig. 4. However, the following description is an example of the operation of the crosslinking reaction simulator 1, and the operation of the crosslinking reaction simulator 1 is not limited to the following description.

[0044] When the crosslinking reaction simulator 1 is started, a heat transfer analysis process S1 is executed. The heat transfer analysis process S1 obtains the temperature of each element of the polymer part 12 of the target workpiece model WM at each time point during the crosslinking reaction. Details will be described later.

[0045] Next, a crosslinking reaction analysis process S2 is executed. In the crosslinking reaction analysis process S2, the equivalent reaction amount of the polymer portion 12 at each time is calculated based on the temperature at each time for each element of the polymer portion 12 of the target workpiece model WM in the crosslinking reaction obtained as a result of the heat transfer analysis, the equivalent reaction amount calculation model EM, and a slope coefficient SC corresponding to the progress of the crosslinking reaction at the target time, and the reaction rate of the crosslinking reaction of the polymer portion 12 is calculated based on the calculated equivalent reaction amount. Details will be described later.

[0046] Next, a structural analysis process S3 is executed. In the structural analysis process S3, an elastic modulus is assigned in accordance with the reaction rate of each element of the polymer portion 12 calculated by the crosslinking reaction analysis unit 4, and structural analysis is performed with the elastic modulus assigned to the polymer portion 12, thereby acquiring the characteristics of the target workpiece model WM. Details will be described later.

[0047] Next, an internal bubble estimation process S4 is executed. In the internal bubble estimation process S4, a torque corresponding to the reaction rate of the polymer portion 12 is calculated based on the reaction rate of the polymer portion 12 calculated by the crosslinking reaction analysis process and the crosslinking reaction curve CC stored in the memory unit 2, and the generation of internal bubbles inside the polymer portion 12 of the target workpiece model WM is estimated based on the torque at the time of demolding. Details will be described later.

[0048] Next, a set condition analysis process S6 is executed. In the set condition analysis process S6, the relationship between the target workpiece model elastic modulus Ks, the set temperature, and the reaction time in the relationship data CD is corrected based on the polymer characteristic value PV. As a result, the set conditions for the crosslinking reaction of the polymer portion 12 are calculated from the polymer characteristic value PV of the raw material polymer that is the target of the crosslinking reaction to obtain a product having the desired target workpiece model elastic modulus Ks. Details will be described later.

[0049] Next, display processing S5 is executed. In display processing S5, the presence or absence of internal bubbles is displayed based on the estimation results by the internal bubble estimation unit 6, depending on the temperature of the mold model MM and the in-mold reaction time from the start of the crosslinking reaction to the demolding of the mold model MM. Furthermore, based on the analysis results by the crosslinking reaction analysis unit 4, a value obtained from the reaction rate depending on the temperature of the mold model MM and the in-mold reaction time from the start of the crosslinking reaction to the demolding of the mold model MM is displayed along with the presence or absence of internal bubbles. Furthermore, based on the results of the structural analysis by the structural analysis unit 5, characteristics depending on the temperature of the mold model MM and the in-mold reaction time from the start of the crosslinking reaction to the demolding of the mold model MM are displayed along with the presence or absence of internal bubbles. Furthermore, based on the analysis results of the setting conditions by the setting condition analysis unit 8, the setting conditions for the crosslinking reaction of the polymer portion 12 to obtain a product having the desired target workpiece model elastic modulus Ks are displayed.

[0050] When the display process S5 is completed, the operation of the crosslinking reaction simulator 1 is completed.

[0051] 3. Heat Transfer Analysis Process S1 3-1. Carbon Black Mass Ratio Next, details of the heat transfer analysis process S1 will be described with reference to FIG. 2 and FIGS. 5 to 8. As shown in FIG. 2, with the lower mold 10A and the upper mold 10B combined, the outer joining member 13 and the inner joining member 14 are placed in the space formed by the lower cavity 100A and the upper cavity 100B. Furthermore, raw material polymer is injected into the space formed by the lower cavity 100A and the upper cavity 100B. Thereafter, a cross-linking reaction of the molecular chains constituting the polymer portion 12 progresses within the mold 10.

[0052] As shown in FIG. 5 , the temperature of the hot plate 11 is set to a constant value. The temperature of the mold 10 is set to the same temperature as the hot plate 11. The temperature of the polymer portion 12 during injection is lower than the set temperature of the mold 10. Therefore, the temperature of the mold 10 temporarily drops after the raw polymer is injected. The temperature of the mold 10 then rises and becomes the same as the temperature of the hot plate 11. The temperature of the polymer portion 12 rises after being injected into the mold 10, and after a sufficient amount of time, becomes the same as the temperature of the mold 10. The temperatures of the inner and outer joining members 14 and 13 may be the same as the ambient temperature before being placed in the mold 10, or may be preheated. If the temperature is the same as the ambient temperature, it may drop to around 0°C, for example, in winter. If preheated, the temperature may be raised to any desired temperature, for example, to around 100°C. After the inner joining member 14 and the outer joining member 13 are placed in the mold 10, the temperatures of the inner joining member 14 and the outer joining member 13 rise due to heat transfer from the mold 10, and after the raw polymer is injected, the temperatures rise further and become the same as the temperature of the mold 10 after sufficient time has passed.

[0053] As shown in Figure 5, it takes some time for the temperature of the polymer portion 12 to reach the temperatures of the hot plate 11 and mold 10 and reach equilibrium. The following equation (1) is used to calculate the unsteady process until the equilibrium state is reached. Equation (1) is the so-called unsteady heat conduction equation.

[0054] where, t: time variable, x: position variable, T: temperature, α: thermal diffusivity

[0055] The thermal diffusivity α is expressed by the following formula (2).

[0056] Where, λ: thermal conductivity, ρ: density, C: specific heat

[0057] The thermal diffusivity α and the thermal conductivity λ correspond to the thermal diffusivity characteristic TS. However, the thermal diffusivity characteristic TS is not limited to the thermal diffusivity α and the thermal conductivity λ.

[0058] If the thermal diffusivity α is known, the thermal diffusivity α can be directly used to calculate the temperature of the polymer portion 12 according to the unsteady heat conduction equation (1). If the thermal diffusivity α is not known, the thermal diffusivity α can be calculated using the thermal conductivity λ, density ρ, and specific heat C according to equation (2), and the calculated α can then be used to calculate the temperature of the polymer portion 12 according to equation (1).

[0059] However, when calculating the temperature of the polymer portion 12 using Equation (1), if a value specified as the thermal diffusivity α of rubber in general standards such as JIS (Japanese Industrial Standards) is used, or if a value specified as the thermal conductivity λ in JIS is used even when Equation (2) is used, a problem occurs in which the value measured when the temperature of the polymer portion 12 of the target workpiece model WM placed in the mold 10 does not match with sufficient accuracy the value predicted when the temperature of the polymer portion 12 is calculated based on Equation (1), as shown in Figure 6. As a result of intensive investigations by the inventors, it was found that the cause was that the value of the thermal diffusivity α or thermal conductivity λ of the raw material polymer of the polymer portion 12 specified in JIS or the like was different from the value of the thermal diffusivity α or thermal conductivity λ of the actual polymer portion 12.

[0060] Therefore, the temperature of the polymer part 12 of the target work model WM to be measured was actually measured, and the thermal diffusivity α or thermal conductivity λ of the polymer part 12 was identified, thereby obtaining the thermal diffusivity α or thermal conductivity λ of the polymer part 12 of the target work model WM to be measured.

[0061] 7 shows the measured temperature of the polymer portion 12 of the target workpiece model WM and the temperature calculated using the measured values ​​of the thermal diffusivity α or the thermal conductivity λ obtained as described above. It was found that the measured temperature and the predicted temperature of the polymer portion 12 matched with high accuracy.

[0062] However, the thermal diffusivity α or thermal conductivity λ of the polymer portion 12 constituting the target workpiece model WM varies depending on the type of raw polymer of the polymer portion 12, the amount or type of additive material added to the raw polymer, etc. For this reason, it is cumbersome to actually measure the thermal diffusivity α or thermal conductivity λ for all polymer portions 12 with different compositions.

[0063] Therefore, the inventors focused on carbon black among the additive materials added to the polymer portion 12. The rubber constituting the polymer portion 12 is a material that does not easily conduct heat. On the other hand, carbon black is a material that easily conducts heat. For this reason, it is believed that the thermal diffusivity α or thermal conductivity λ of the polymer portion 12 is closely related to the mass ratio of carbon black to the raw material polymer constituting the polymer portion 12. Examples of carbon black include furnace black, acetylene black, and ketjen black.

[0064] 8 shows the relationship between the mass ratio of carbon black to raw polymer and the thermal conductivity λ of the polymer portion 12 when experiments were conducted using multiple types of raw polymer and different mass ratios of carbon black. In FIG. 8, the mass ratio of carbon black to raw polymer is expressed in parts by mass. However, the mass ratio of carbon black to raw polymer may also be expressed in % by mass.

[0065] As shown in Figure 8, the thermal conductivity λ of the polymer portion 12 has a linear relationship with the mass ratio of carbon black to raw polymer. It was also found that the thermal conductivity λ of the polymer portion 12 is independent of the type of raw polymer. This makes it possible to accurately predict the thermal conductivity λ of the polymer portion 12 based on the amount of carbon black mixed, regardless of the type of raw polymer. Furthermore, since the thermal diffusivity α can be calculated from the thermal conductivity λ using Equation (2), it is also possible to accurately predict the thermal diffusivity α.

[0066] 3-2. Contact Heat Transfer Coefficient CH As shown in FIG. 2, when the outer joining member 13 is disposed within the mold 10, an outer gap 15A may be formed between the inner surface of the mold 10 and the outer joining member 13. In this case, heat is transferred from the inner surface of the mold 10 to the outer joining member 13 via the air in the outer gap 15A. In the heat transfer analysis process S1, assuming that the inner surface of the mold 10 and the outer joining member 13 are in contact with each other, surfaces of the mold 10 and the outer joining member 13 that take into account the outer gap 15A are selected, and the contact heat transfer coefficient CH is set according to the size of the outer gap 15A. The contact heat transfer coefficient CH is determined by measuring temperature behavior when the size of the outer gap 15A is changed in an experiment using a test piece, and a value that can reproduce each temperature behavior is obtained by analysis.

[0067] Furthermore, when the inner joining member 14 is disposed in the mold 10, an inner gap 15B may be formed between the inner surface of the mold 10 and the inner joining member 14. In this case, heat is transferred from the inner surface of the mold 10 to the inner joining member 14 through the air in the inner gap 15B. In the heat transfer analysis process S1, a state in which the inner surface of the mold 10 and the inner joining member 14 are in contact is assumed, and surfaces of the mold 10 and the inner joining member 14 that take the inner gap 15B into consideration are selected, and the contact heat transfer coefficient CH is set according to the size of the inner gap 15B. The contact heat transfer coefficient CH is determined by measuring temperature behavior when the size of the inner gap 15B is changed in an experiment using a test piece, and a value that can reproduce each temperature behavior is used.

[0068] As described above, the storage unit 2 stores the contact heat transfer coefficient CH in the gaps 15A, 15B between the forming die model MM and the metal fittings in a state in which the target workpiece model WM is placed on the forming die model MM.

[0069] In the heat transfer analysis process S1, the analysis unit 32 is configured to perform heat transfer analysis with the target workpiece model WM placed on the mold model MM, using the polymer thermal diffusivity determined by the polymer thermal diffusivity determination unit 31, the temperature condition TC, and the contact heat transfer coefficient CH stored in the storage unit 2. This makes it possible to accurately predict the temperature of the polymer portion 12.

[0070] 3-3. Air Heat Transfer Coefficient AC Even after the target workpiece model WM is released from the mold model MM, the temperature of the polymer portion 12 of the target workpiece model WM does not immediately drop, and therefore the cross-linking reaction progresses in the polymer portion 12. Therefore, in this embodiment, in order to predict the temperature of the polymer portion 12 even after the target workpiece model WM is released from the mold model MM, the condition input unit 30 is configured to input the outside air conditions OC, which include the air temperature around the target workpiece model WM released from the mold model MM. Furthermore, the memory unit 2 stores the air heat transfer coefficient AC, which is the heat transfer coefficient of the air around the target workpiece model WM released from the mold model MM.

[0071] In this embodiment, the analysis unit 32 performs a heat transfer analysis in a state where the target workpiece model WM is released from the mold model MM after the polymer portion 12 of the target workpiece model WM has undergone a crosslinking reaction within the mold model MM, using the polymer thermal diffusivity determined by the polymer thermal diffusivity determination unit 31, the air heat transfer coefficient AC, and the outside air conditions OC stored in the storage unit 2. This makes it possible to accurately predict the temperature of the polymer portion 12 in a state where the target workpiece model WM has been released from the mold model MM.

[0072] As shown in FIG. 9 , an analytical mesh is created for the state in which the target workpiece model WM is placed on the forming mold model MM. The size of the analytical mesh may be predetermined or may be input from the condition input unit 30. The analytical mesh is created taking into consideration the characteristics (e.g., symmetry) of the shapes of the forming mold model MM and the target workpiece model WM. Eight-node solid elements were used as the mesh type. However, any model can be used for the mesh type, and for example, four-node solid elements may be used.

[0073] By using the analytical mesh, it is possible to predict the temperature at each node of the target workpiece model WM. In other words, it is possible to predict the temperature distribution of the polymer portion 12 of the target workpiece model WM. In particular, it is possible to predict the temperature at each node of the polymer portion 12 at each time.

[0074] The forming die model MM is not limited to the above, and may be, for example, a forming die model MMa shown in Fig. 33. The forming die model MMa includes a lower die 10A, an upper die 10B, a first intermediate die 110A, a second intermediate die 110B, a lower insert 16A, an upper insert 16B, a lower shaft portion 17A, an upper shaft portion 17B, and a heater 18. For ease of explanation, the gap 15 is omitted from Fig. 33.

[0075] Between the lower mold 10A and the upper mold 10B disposed above the lower mold 10A, there are disposed a first intermediate mold 110A disposed on the lower mold 10A side and a second intermediate mold 110B disposed on the upper mold 10B side. The first intermediate mold 110A and the second intermediate mold 110B are disposed radially outward from the target workpiece model WM. In other words, the first intermediate mold 110A and the second intermediate mold 110B are disposed outward from the outer joining member 13 of the target workpiece model WM.

[0076] The lower insert 16A and the upper insert 16B come into contact with the polymer portion 12 of the target workpiece model WM. As a result, the lower insert 16A forms the shape of the lower edge of the polymer portion 12, and the upper insert 16B forms the shape of the upper edge of the polymer portion 12.

[0077] The lower shaft portion 17A is disposed in a state in which it penetrates the lower mold 10A in the vertical direction. The lower shaft portion 17A is disposed in a state in which it is inserted from below into the inside of the inner joining member 14 of the target workpiece model WM. The upper surface of the lower shaft portion 17A abuts against the inner joining member 14 from below.

[0078] The upper shaft portion 17B is attached to the upper mold 10B and is arranged in a state where it is inserted from above into the inside of the inner joint member 14 of the target workpiece model WM. The lower surface of the upper shaft portion 17B abuts against the inner joint member 14 from above.

[0079] The inner joint member 14 is positioned by being sandwiched between the upper surface of the lower shaft portion 17A and the lower surface of the upper shaft portion 17B.

[0080] A heater 18 is disposed inside the lower mold 10A. The heater 18 may be an electric heating wire or a pipe through which a heated fluid (water or oil) flows. In the target workpiece model WMa, the temperature of the polymer portion 12 can be accurately predicted by using the set temperature of the lower insert 16A.

[0081] However, the heater 18 may be disposed in the upper die 10B, and in that case, the set temperature of the upper insert 16B may be used.

[0082] 4. Crosslinking Reaction Analysis Process S2 Next, the crosslinking reaction analysis process S2 will be described. Regarding the crosslinking reaction analysis process S2, the reaction rate will first be described. Next, the conventional technology will be described, followed by the problems with the conventional technology. After that, the crosslinking reaction analysis process of this embodiment will be described.

[0083] 4-1. Reaction Rate The reaction rate of the crosslinking reaction is measured as follows. As shown in FIG. 10 , the reaction rate of the rubber according to this embodiment is measured in accordance with JIS K6300-2 using a vulcanization tester 70 (an example of a crosslinking reaction tester). The reaction rate (degree of vulcanization) is calculated from the relationship between the torque and reaction time. The vulcanization tester 70 may be, for example, a Curastometer (registered trademark). However, the above-described JIS K6300-2 is only an example of a method for calculating the reaction rate, and the calculation method for the reaction rate is not limited to the above. Furthermore, the raw polymer constituting the polymer portion 12 is not limited to the polymer specified in JIS K6300-2, and any polymer can be selected as appropriate. JIS K6300-2 is based on ISO 6502:1999, but some provisions not included in the corresponding international standard have been added. Furthermore, some of the corresponding international standards have been changed and deleted.

[0084] Furthermore, when outputting vibration characteristics taking into account the reaction rate (degree of vulcanization), reaction behavior (vulcanization behavior) can be measured under desired vibration conditions using, for example, a rubber processability analyzer (RPA). The RPA is configured so that the frequency and amplitude can be set over a wide range.

[0085] As shown in Figure 10, a vulcanization tester 70 has a disk 75 attached to the tip of a rotor 74 placed between a lower die 72 and an upper die 73, each having a cavity 71. The lower die 72 and upper die 73 can be set to a predetermined temperature. The torsion angle of the rotor 74 can be selected from +3° to -3° or +1° to -1°. A measurement sample 76 is placed between the cavity 71 and the disk 75, and a crosslinking reaction is caused while a rotational torque is applied from the disk 75 at a constant temperature, and the progress of the crosslinking reaction is detected from changes in the torque.

[0086] FIG. 11 shows an example of measurement results. FIG. 11(a) is a graph (crosslinking reaction curve CC) showing the change in torque versus reaction time at a predetermined reaction temperature. Meanwhile, the reaction rate is defined as 0% for the minimum torque and 100% for the maximum torque. FIG. 11(b) shows a graph (reference reaction curve RC) showing the change in reaction rate versus reaction time at a predetermined reaction temperature. For the rubber according to this embodiment, the reaction rate is referred to as the degree of vulcanization. The degree of vulcanization is defined as the degree of vulcanization when the physical properties of the vulcanized rubber (such as elastic modulus, elongation, tensile strength, and hardness) are used as indicators. In FIG. 11(a), the region from reaction time 0 to the reaction time showing the minimum torque is a plastic region where the crosslinking reaction of the polymer portion 12 has not yet reacted. Therefore, in FIG. 11(b), the reaction rate in this region is set to 0%.

[0087] 4-2. Calculation method of reaction rate Next, a method of calculating the reaction rate will be described. In the above-mentioned Non-Patent Document 1, the equivalent reaction amount of the crosslinking reaction is estimated using the formula (4) relating to the equivalent reaction amount obtained by modifying the Arrhenius formula (3). The formula (4) relating to the equivalent reaction amount is calculated by dividing the reaction amount at a reaction temperature T and reaction time t by the reaction amount at a certain reference temperature T ０ , reference time t ０ It represents how many times the reaction amount in the reaction reversion period is. Note that equation (4) is called the equivalent vulcanization equation in the vulcanization reaction of rubber. By using this equivalent vulcanization equation, it is possible to predict the reaction rate in the reaction reversion period, which will be described later. As will be described in detail later, in this embodiment, it is possible to further improve the prediction accuracy in the reaction reversion period compared to when the equivalent vulcanization equation is simply used. Note that in general analysis software, the reaction rate in the reaction reversion period cannot be predicted using equations used to predict the behavior of thermosetting reactions of rubber, epoxy, etc. (for example, Kamal's reaction rate equation).

[0088] where k: reaction rate A: frequency factor E: activation energy R: gas constant T: reaction temperature

[0089] where U: equivalent reaction amount E: activation energy R: gas constant T: reaction temperature T ０ : Reference temperature t: Reaction time

[0090] The Arrhenius equation (3) is transformed as follows to derive the equivalent reaction amount equation (4).

[0091] The reaction amount Z of the crosslinking reaction at reaction time t is expressed by formula (5).

[0092] Where, Z: reaction amount t: reaction time

[0093] The reaction amount Z is calculated from the Arrhenius equation (3) and the above equation (5) as shown in the following equation (6).

[0094]

[0095] Here, a certain reference reaction temperature T ０ And the reaction time t ０ The reaction amount Z when ０ Considering this, Z ０ is expressed by the following equation (7).

[0096]

[0097] Reaction temperature T ０ and reaction time t ０ The reaction amount Z in ０ When the ratio U of the reaction amount Z at the reaction temperature T and reaction time t to the reaction amount Z is taken, the ratio U is expressed by the following formula (8).

[0098]

[0099] Substituting equations (6) and (7) into equation (8), equation (9) for the ratio U is obtained.

[0100]

[0101] In equation (9), t ０ When = 1, equation (9) is expressed as equation (4) above.

[0102] When the reaction temperature T changes moment by moment, the increase in the equivalent reaction amount for a small time period at the target time is calculated based on the temperature at the target time, the above formula (4), and the slope coefficient SC, the integrated value of the equivalent reaction amount from the start of the crosslinking reaction to the target time is calculated based on the increase in the equivalent reaction amount, and the reaction rate at the target time is calculated based on the integrated value of the equivalent reaction amount from the start of the crosslinking reaction to the target time.

[0103] Based on the formula (4), the equivalent reaction amount increase amount ΔU in the minute time Δt of the target time ｉ is expressed by the following equation (10).

[0104] Here, ΔU ｉ : Equivalent reaction increase amount in a small time of the target time Δt: Small time of the target time

[0105] When the reaction temperature T changes from moment to moment, the equivalent reaction amount U can be calculated by accumulating the temperature T for each unit time as shown in the following equation (11).

[0106] where U: equivalent reaction amount E: activation energy R: gas constant T: reaction temperature T ０ : Reference temperature Δt: Micro time of target time

[0107] The above-mentioned formula (4) is the formula inside Σ in the above-mentioned formula (11), and corresponds to each accumulated equivalent reaction amount.

[0108] The value of (-E / R) is required to calculate equation (4) or equation (11). In the prior art, (-E / R) was calculated as follows:

[0109] By taking the natural logarithm of both sides of Arrhenius equation (3), the following equation (12) is obtained.

[0110] where k: reaction rate A: frequency factor E: activation energy R: gas constant T: reaction temperature

[0111] For the above reaction amount, the reaction time t at a predetermined reaction rate γ γIt is considered that the following equation (13) holds between γ and the reaction rate k: The reaction rate γ can be set to any value between 0% and 100%, such as 5% or 90%, as appropriate.

[0112] where B: an arbitrary constant t γ : reaction time at a given reaction rate γ

[0113] The natural logarithm of both sides of equation (13) is taken, and this is substituted into equation (12) to transform the equation, thereby obtaining equation (14).

[0114]

[0115] From equation (14), ln(1 / t γ When the time at which the reaction rate reaches α% and each measured temperature are plotted with (E / R) on the vertical axis and (1 / T) on the horizontal axis, it is possible to approximate the time at which the reaction rate reaches α% and each measured temperature, and it is found that the slope of this line is (−E / R) (see FIG. 12).

[0116] In an actual reaction, the temperature changes over time as the reaction time elapses, as shown in Figure 13. In such a case, the equivalent reaction amount is calculated using equations (4) and (11) as follows:

[0117] First, the reaction time is divided into predetermined intervals (for example, 1 minute) as shown in Fig. 13. Next, the temperature is read at each predetermined interval.

[0118] Next, the equivalent reaction amount per predetermined interval is calculated using equation (4). The calculation is performed by substituting the predetermined interval Δt into equation (4). The calculation can be simplified by setting the predetermined interval to, for example, Δt = 1.

[0119] Next, ΔU per predetermined interval Δt calculated by equation (4) ｉ A specific calculation method will be described with reference to FIG.

[0120] As shown in FIG. 14, the reaction time t ｉ－１ , reaction temperature T ｉ－１ Equivalent reaction amount U ｉ－１ is calculated using equation (4). Similarly, the reaction time t ｉ , reaction temperature T ｉ Equivalent reaction amount Uｉ Calculate.

[0121] Next, U ｉ－１ × Δt and U ｉ × Δt. This gives the reaction time t ｉ－１ ~t ｉ In this embodiment, Δt=1 / 2t. Similarly, in accordance with equation (11), the equivalent reaction amount increase ΔU per predetermined time from t=1 to t=n is calculated. ｉ By integrating these, the equivalent reaction amount U can be obtained.

[0122] Next, a method for calculating the reaction rate from the equivalent reaction amount U will be described. First, the reaction temperature T ０ The reaction amount Z at reaction time t is expressed by equation (5). Next, from equation (5), the reaction temperature T ０ And the reaction time t ０ The reaction amount Z when ０ is expressed by the following formula (15): where the reaction temperature in formula (5) and formula (15) is the same value T ０ Therefore, from equation (3), each k has the same value.

[0123]

[0124] When the above-mentioned formulas (5) and (15) are substituted into the above-mentioned formula (8), the reaction rate k is reduced to obtain the following formula (16).

[0125]

[0126] In equation (16), t ０ = 1, equation (16) can be expressed as equation (17) below.

[0127] According to equation (17), t ０ When t = 1, the reaction time t can be read as the equivalent reaction amount U. As a result, the reaction temperature T ０In the graph showing the change in reaction rate versus reaction time, the reaction time t can be read as the equivalent reaction amount U. Figure 15 shows a graph with the reaction rate on the vertical axis and the equivalent reaction amount on the horizontal axis. Based on this graph, the reaction rate can be calculated from the equivalent reaction amount U obtained by calculation.

[0128] 4-3. Issues with the reaction rate calculation method However, when the reaction rate is estimated based on the above-mentioned method, there is a problem that the difference between the estimated value and the actual measured value is large, and sufficient accuracy cannot be obtained.

[0129] The change in reaction rate with reaction time is shown in Figure 16. When formula (4) is used, the reference temperature is set to 160°C, so when the reaction temperature is 160°C, the measured value and the predicted value match with good precision.

[0130] However, for reaction temperatures of 170°C and 180°C, the actual measured values ​​and predicted values ​​do not match sufficiently. In general, the predicted values ​​are larger than the actual measured values. The reason why the maximum values ​​of the reaction rate match is because, as mentioned above, the maximum value of the reaction rate is defined as 100%.

[0131] The reason why the actual measured value and the predicted value do not sufficiently match in the prior art will be explained with reference to Fig. 17. Fig. 17 shows a graph showing the change in reaction rate with reaction time. The crosslinking reaction according to this embodiment proceeds through the following process.

[0132] In the early stage of the reaction, when the reaction begins, the reaction rate gradually increases. After a certain amount of time has passed and the reaction reaches the accelerated stage, the reaction rate increases rapidly. Then, in the later stage of the reaction, the rate of increase in the reaction rate slows down and the reaction rate reaches a maximum value. After the reaction rate reaches its maximum value, the reaction enters the return stage, where the return reaction proceeds. During this return stage, the reaction rate gradually decreases.

[0133] In the above-mentioned technique, the period from the initial stage of the reaction to the reversion stage was regarded as one crosslinking reaction, so to speak. Therefore, the reaction rate was predicted using one activation energy E. As a result, it is thought that the actual measured value of the reaction rate did not sufficiently match the predicted value.

[0134] 4-4. Crosslinking Reaction Analysis Process of the Present Embodiment In the present embodiment, a plurality of activation energies are calculated in accordance with the progress of the reaction, and the equivalent reaction amount is predicted based on the obtained activation energies. Based on this concept, Equation (4) is transformed into Equation (18) below.

[0135] Here, E Ｎ : Activation energy for each stage of reaction divided into N stages

[0136] Figure 17 shows an example in which the reaction progress state is divided into four stages: early reaction progress, reaction promotion period, late reaction progress, and reaction return period. However, the reaction progress state may also be divided into any two to three, or five or more stages. Furthermore, the entire period including the reaction progress period and reaction return period may be divided into equal intervals. As an index for dividing into equal intervals, the progress of the crosslinking reaction over the entire period may be used as a standard, and the progress of the crosslinking reaction may be divided into 5% increments, for example. However, the index for division is not limited to 5%, and may be 1% to 4%, or 6% or more.

[0137] In this embodiment, the period from 0% to 5% of the reaction rate was defined as the initial period of the reaction, the period from 5% to 50% as the accelerated period of the reaction, the period from 50% to 100% as the late period of the reaction, and the period after the late period of the reaction where the reaction rate decreased from 100% to 90% as the reaction return period.

[0138] In each reaction step, the activation energy E Ｎ A term including E Ｎ As shown in FIG. 17, the value of (E / R) was calculated at the initial stage of the reaction (reaction rate 0% to 5%). １ / R) is approximately 12,000, and (E ２ / R) is about 9000, and (E ３ / R) is approximately 13,000, and (E ４ / R) was approximately 16,000. ＮThe value of (E / R) is not limited to the above value. When the period from the initial stage of the reaction to the reversion stage of the reaction was considered as one reaction stage, (E / R) was 11,000.

[0139] FIG. 18 shows the activation energy E １ ~E ４ The results of predicting the reaction rate using this method are shown together with the actual measured values. Not only did the actual measured value and the predicted value match at the reference temperature of 160°C, but the actual measured value and the predicted value also matched with good accuracy at reaction temperatures of 170°C and 180°C. This allows the reaction rate to be predicted with good accuracy.

[0140] In this embodiment, the slope coefficient SC is set to a different value during the reaction progress period until the crosslinking reaction rate reaches a peak and during the reaction reversion period after the reaction rate has exceeded the peak. Also, in this embodiment, the slope coefficient SC is set to a smaller value during the reaction reversion period than during the reaction progress period.

[0141] In this embodiment, the slope coefficient SC is set to different values ​​in the initial reaction period, the reaction acceleration period, the later reaction period, and the reaction return period. In this embodiment, the slope coefficient SC in the reaction acceleration period is set to a value larger than those in the initial reaction period and the later reaction period.

[0142] When the entire period of the crosslinking reaction is divided into equal intervals, the slope coefficient SC is set to a value corresponding to each interval. However, the slope coefficient SC calculated by the method described in the above-mentioned Non-Patent Document 1 may also be used.

[0143] 4-5. Method of expressing the reaction rate of the crosslinking reaction A method of expressing the reaction rate of the crosslinking reaction in this embodiment will be described with reference to Fig. 19. Fig. 19 is a graph showing the change in the reaction rate of the crosslinking reaction with respect to reaction time. The graph shown by the solid line is the actual measured value of the reaction rate.

[0144] The crosslinking reaction rate shown by the solid line gradually decreases after reaching a maximum value due to the progress of the back reaction. Therefore, for example, there will be a point where the reaction rate reaches 80% before reaching the maximum value, and a point where the reaction rate reaches 80% after reaching the maximum value. In this case, if we simply compare the reaction rate figures, it will be impossible to distinguish between the 80% value before the reaction rate reaches the maximum value and the 80% value after the reaction rate reaches the maximum value.

[0145] Therefore, in this embodiment, the reaction rate of the crosslinking reaction is defined as 0% at the start of the crosslinking reaction in the reaction progress period until the reaction rate of the crosslinking reaction reaches its peak, and the increasing rate of the reaction rate of the crosslinking reaction is defined in the range of 0% to 100%, and in the reaction return period when the reaction rate has exceeded its peak, the decreasing rate of the reaction rate from the peak reaction rate is defined as a value obtained by adding 100% to the decreasing rate.

[0146] To explain this using the graph in Figure 19, the line extending parallel to the horizontal axis from the point where the reaction rate is 100% is used as the axis of symmetry, and the graph of the reaction return period when the reaction rate has exceeded its peak is inverted as shown by arrow B. As a result, the point where the reaction rate is 80% after reaching its maximum value can be expressed as a reaction rate of 120%. As a result, the reaction rate before the reaction rate reaches its maximum value and the reaction rate after the reaction rate reaches its maximum value can be clearly distinguished (see Figure 20).

[0147] 4-6. Method for determining values ​​calculated by simulation Next, a method for determining values ​​calculated by simulation will be described. In this embodiment, the slope coefficient SC and the reaction rate are determined using one or both of a database and a function. By using a database, the accuracy of the estimated value can be improved. On the other hand, by using a function, the calculation speed can be improved.

[0148] 1, the storage unit 2 stores a first relationship data map DM1 that defines the correspondence relationship between the reaction rate and the slope coefficient SC of the crosslinking reaction of the polymer portion 12. The reaction rate calculation processing unit 41 determines the slope coefficient SC using the reaction rate at the previous time and the first relationship data map DM1.

[0149] The memory unit 2 stores a second relationship data map DM2 that defines the correspondence relationship between the equivalent reaction amount integrated value and the reaction rate at the target time. The reaction rate calculation processing unit 41 determines the reaction rate at the target time using the equivalent reaction amount integrated value at the target time and the second relationship data map DM2.

[0150] The second relationship data map DM2 is set from a reference reaction curve RC based on the relationship between the crosslinking reaction time and the torque generated in the test polymer material at a reference reaction temperature measured by a vulcanization testing machine 70 using a test polymer material corresponding to the polymer portion 12 of the target work model WM.

[0151] 1, the storage unit 2 stores a first function F1 that defines the correspondence relationship between the reaction rate of the crosslinking reaction of the polymer portion 12 and the slope coefficient SC. The reaction rate calculation processing unit 41 determines the slope coefficient SC using the reaction rate at the previous time and the first function F1.

[0152] The first function F1 is set to a different function according to a plurality of reaction rate divisions set according to the progress of the crosslinking reaction of the polymer portion 12. The number of divisions of the progress of the crosslinking reaction is arbitrary, and may be one or more than two. Any function can be appropriately selected as the function, such as an n-th order function including linear interpolation, a spline curve, etc. In this embodiment, for example, a sextic function is preferably used. The number of divisions of the first function F1 is determined by the activation energy E Ｎ This can be set independently of the classification used to calculate the value.

[0153] The storage unit 2 stores a second function F2 that defines the correspondence relationship between the equivalent reaction amount integrated value and the reaction rate at the target time. The reaction rate calculation processing unit 41 determines the reaction rate at the target time using the equivalent reaction amount integrated value at the target time and the second function F2.

[0154] The second function F2 is set from a reference reaction curve RC based on the relationship between the crosslinking reaction time and the torque generated in the test polymer material at a reference reaction temperature measured by a vulcanization tester 70 using a test polymer material corresponding to the polymer portion 12 of the target work model WM.

[0155] The second function F2 is set to a different function according to a plurality of reaction rate divisions set according to the progress of the crosslinking reaction of the polymer portion 12. The number of divisions of the progress of the crosslinking reaction is arbitrary, and may be one or more than two. Any function can be appropriately selected as the function, such as an n-th order function including linear interpolation, a spline curve, etc. In this embodiment, for example, a sextic function is preferably used. The number of divisions of the second function F2 is determined by the activation energy E Ｎ This can be set independently of the classification used to calculate the value.

[0156] 21 shows an example of the second function F2 when the progress of the crosslinking reaction is divided into four stages. In this embodiment, the stages are an early stage of reaction progress, an accelerated stage of reaction progress, a late stage of reaction progress, and a return stage of reaction. In each stage, the reaction rate and the reaction time are approximated by mutually different functions.

[0157] Next, an analytical mesh of the target workpiece model WM is created. The size of the analytical mesh may be predetermined or may be input from the condition input unit 30. When creating the analytical mesh, the analytical mesh as shown in FIG. 22 is created taking into consideration the shape characteristics (e.g., symmetry) of the target workpiece model WM. Eight-node solid elements were used as the mesh type. However, any model can be used as the mesh type, and for example, four-node solid elements may also be used.

[0158] Using the analytical mesh, the reaction rate of each node can be calculated from the temperature of each node at each time of only the polymer portion 12 of the target workpiece model WM.

[0159] 5. Structural Analysis Processing S3 5-1. Initial Processing Next, the structural analysis processing S3 will be described. The structural analysis unit 5 creates an analytical mesh for the target workpiece model WM. The analytical mesh for the target workpiece model WM may be the same as the analytical mesh created in the crosslinking reaction analysis processing S2, or it may be different. In this embodiment, the analytical mesh used in the structural analysis processing S3 is the same as the analytical mesh used in the crosslinking reaction analysis processing S2.

[0160] 5-2. Temperature and Reaction Rate Acquisition Next, the structural analysis unit 5 acquires the analysis results of the temperature distribution of the polymer portion 12 of the target workpiece model WM analyzed by the heat transfer analysis unit 3. The structural analysis unit 5 also acquires the analysis results of the reaction rate of the polymer portion 12 of the target workpiece model WM analyzed by the crosslinking reaction analysis unit 4.

[0161] 5-3. Elastic Modulus Prediction Next, the structural analysis unit 5 calculates the elastic modulus of each node at each time of the polymer portion 12 of the target workpiece model WM. First, a test piece with a crosslinking reaction rate of 100% is created, and the elastic modulus of this test piece is measured. Next, a correction coefficient is calculated based on the quotient obtained by dividing the torque at the target reaction rate by the torque at a reaction rate of 100%. The torque is a torque that can be measured by the vulcanization tester 70. The structural analysis unit 5 calculates the elastic modulus of each node at each time by multiplying the measured elastic modulus by the calculated correction coefficient.

[0162] Next, the structural analysis unit 5 predicts the characteristics of the target workpiece model WM. The input conditions used for the finite element analysis are the temperature of each node at each time, the elastic modulus of each node at each time, and the boundary conditions of the outer joining member 13 and the inner joining member 14.

[0163] After the input conditions are set, a force or strain is applied to the target workpiece model WM in a direction perpendicular to the axis A of the target workpiece model WM.

[0164] The structural analysis unit 5 outputs the overall elastic modulus of the target workpiece model WM. In this embodiment, the elastic modulus with respect to a statically applied force in a direction perpendicular to the axis A of the target workpiece model WM is predicted as the characteristic. However, the predicted characteristic is not limited to the elastic modulus with respect to a force or strain in a direction perpendicular to the axis A, and any physical property value can be predicted as the characteristic, such as the elastic modulus with respect to a force oscillating in a direction perpendicular to the axis A, the elastic modulus with respect to a statically applied force or strain in a direction along the axis A, or the elastic modulus with respect to a force oscillating in a direction along the axis A.

[0165] 6. Internal Air Bubble Estimation Process S4 Next, the internal air bubble estimation process S4 will be described. A rubber product, which is an example of this embodiment, is produced by molding and vulcanization. As a result, gases that were originally dissolved in the rubber or gases generated by the vulcanization reaction remain dissolved in the rubber under the high temperature and high pressure conditions of vulcanization. When the mold 10 is opened, the pressure applied to the rubber decreases, reducing the solubility of gases in the rubber. Therefore, if vulcanization is not fully progressed, internal air bubbles 80 may occur in the polymer portion 12 made of rubber (see FIG. 23).

[0166] If the vulcanization time is extended while the rubber is placed in the mold 10 and the vulcanization is allowed to proceed, the internal bubbles 80 will no longer be generated. In the vulcanization of an actual product, points within the product that are different in distance from the mold 10 (the heat source) will receive different total heat amounts due to differences in temperature rise history, and the vulcanization progress will be different even for the same vulcanization time. Therefore, it is necessary to heat and vulcanize the product in the mold 10 until the generation of internal bubbles 80 is no longer observed in the slowest vulcanization portion. The degree of vulcanization of the rubber in the slowest vulcanization portion at the time the mold 10 is opened when the rubber is molded for a vulcanization time until the generation of internal bubbles 80 is no longer observed in this slowest vulcanization portion is called the blow point vulcanization degree (hereinafter referred to as the blow point) (see Figure 24).

[0167] In order to suppress internal bubbles 80, it is preferable to make the vulcanization time in the mold 10 as long as possible. However, if the vulcanization time in the mold 10 is extended excessively, the manufacturing efficiency of the product decreases. For this reason, it is important to estimate the blow point in order to determine the vulcanization time of the product.

[0168] 25(a) shows the blow point reaction rate when different types of fillers are added to different types of polymers in different amounts. The blow point reaction rate is the reaction rate of the crosslinking reaction of the polymer portion 12 at the blow point.

[0169] Eight types of polymers, A to H, were used as raw material polymers. Samples A1 to A3 were made using raw material polymer A with different filler blend amounts. Similarly, samples B1 to B3 were made using raw material polymer B with different filler blend amounts, and samples C1 to C3 were made using raw material polymer C with different filler blend amounts. Because the samples used were commercially available products, the exact type and blend amount of filler are unknown.

[0170] 25(a), the blow point reactivity rate varies greatly depending on the type of raw polymer and the amount of filler blended. For this reason, it was found that it is difficult to predict the blow point of the polymer portion 12 using the blow point reactivity rate.

[0171] 26, when the pressure P1 at which the internal cells 80 try to expand and the pressure P2 at which the polymer portion tries to suppress the internal cells 80 become P1 < P2, the internal cells 80 are thought to be suppressed by the rubber and disappear. Therefore, the blow point reactivity was converted into a torque measured by the vulcanization tester 70 corresponding to the blow point reactivity, and the results were compared for each sample.

[0172] 25(b) shows the blow point torque when different types of fillers are added to different types of polymers. As described above, the blow point torque is obtained by converting the blow point reactivity into a torque measured by the vulcanization tester 70 corresponding to the blow point reactivity.

[0173] As shown in FIG. 25(b), the torque (blow point torque) at which internal bubbles 80 no longer occur was approximately the same regardless of the type of raw polymer or the amount of filler blended. In other words, by using an easily measurable torque, the blow point can be easily predicted. The vertical axis of FIG. 25(b) shows the threshold value. The estimation unit 62 of the internal bubble estimation unit 6 estimates that internal bubbles 80 have occurred in the polymer portion 12 of the target workpiece model WM if the torque when the target workpiece model WM is demolded does not exceed the threshold value.

[0174] The threshold is set to a uniform value for multiple types of raw polymers. The threshold is a torque corresponding to a reaction rate greater than the reaction rate of the crosslinking reaction at the blow point of the polymer portion 12. However, multiple groups may be set, and a threshold may be set for each group. The multiple groups may be divided by any criteria.

[0175] The reaction rate acquisition unit 51 acquires the reaction rate for each part of the polymer portion 12, the torque calculation unit 61 calculates the torque for each part of the polymer portion 12, and the estimation unit 62 estimates the occurrence of internal bubbles 80 for each part of the polymer portion 12.

[0176] 7. Setting Condition Analysis Process S6 Next, the setting condition analysis process S6 will be described. The elastic modulus Ks of the target workpiece model WM can change depending on the setting conditions for the crosslinking reaction of the polymer portion 12 of the target workpiece model WM. The setting conditions are not particularly limited and can include, for example, primary vulcanization conditions such as the set temperature of the mold 10 for the crosslinking reaction, the reaction time of the crosslinking reaction, the temperature of the raw polymer when the raw polymer is injected into the mold 10, the temperature of the insert, the temperature of the outer joining member 13, the temperature of the inner joining member 14, the temperature of the lower shaft portion, and the temperature of the upper shaft portion, as well as secondary vulcanization conditions. The secondary vulcanization conditions may include the primary vulcanization conditions described above or may include other vulcanization conditions. The secondary vulcanization conditions are an example of additional crosslinking reaction conditions.

[0177] An example of the setting conditions will be described below, taking the set temperature of the mold 10 as an example. As shown in Fig. 34(a) , the elastic modulus Ks of the target workpiece model WM has a correlation with the set temperature in the cross-linking reaction of the polymer portion 12 of the target workpiece model WM.

[0178] On the other hand, the elastic modulus Ks of the target workpiece model WM has a correlation with a polymer characteristic value PV, which is a characteristic value related to the polymer portion 12 of the target workpiece model WM. The polymer characteristic value PV is not particularly limited and includes a characteristic value related to the elasticity of the raw material polymer, and further includes at least one of the scorch time and Mooney viscosity of the raw material polymer.

[0179] As an example of a polymer characteristic value PV, RKs will be described. RKs is defined as the elastic modulus of a test specimen, which is obtained by molding and crosslinking a raw polymer into a predetermined shape, applying a load to the test specimen in a compressive direction at a predetermined displacement rate using a predetermined jig, and calculating the elastic modulus by a predetermined method.

[0180] An example of a method for measuring RKs is described below. A raw polymer is placed in a mold having a cavity with a diameter of 30.0 mm and a thickness of 30 mm, and crosslinked by holding at 150°C for 20 minutes to produce a cylindrical sample. The produced sample is used to measure the static spring constant of the crosslinked raw polymer. A load of 10 N is applied to the sample in the thickness direction, and the sample is compressed at a displacement rate of 0.33 mm / sec, and the spring constant is calculated using the forward stroke method based on JIS K6385:2012.

[0181] Another example of the polymer characteristic value PV is TS, which is defined as the difference in deformation between the first load and the second load, when a test piece is prepared by molding and crosslinking a raw polymer into a predetermined shape, applying a predetermined first load in the thickness direction, and then applying a second load smaller than the first load.

[0182] An example of a method for measuring TS is described below. A raw polymer is placed in a mold having a cavity with a diameter of 30.0 mm and a thickness of 30 mm, and crosslinked by holding at 150°C for 20 minutes to produce a cylindrical sample. A load of 70 kg (an example of a first load) is applied to the produced sample in the thickness direction of the sample. Then, a load of 20 kg (an example of a second load) is applied, and the difference in the amount of deformation in the thickness direction of the sample when a load of 70 kg and a load of 20 kg are applied is measured.

[0183] 34(b) shows the change in the modulus of elasticity Ks with respect to R Ks, which is an example of a characteristic value of the polymer portion 12. The modulus of elasticity Ks of the target workpiece model WM has a correlation with R Ks. In this embodiment, the modulus of elasticity Ks of the target workpiece model WM and R Ks have a linear relationship.

[0184] The polymer characteristic value PV of the raw polymer varies for each raw polymer lot. Since the polymer characteristic value PV can be measured in advance for each raw polymer lot, it is possible to know the polymer characteristic value PV before the crosslinking reaction is performed. Therefore, it is considered possible to adjust the elastic modulus Ks of the target workpiece model WM after the crosslinking reaction to a desired value by changing the setting conditions for the crosslinking reaction according to the polymer characteristic value PV of the raw polymer.

[0185] In order to adjust the elastic modulus Ks of the target work model WM, it is necessary to clarify the correlation between the set temperature of the crosslinking reaction and the elastic modulus Ks of the target work model WM for each raw material polymer having a different polymer characteristic value PV.

[0186] 4, it is possible to predict the correlation between the elastic modulus Ks of the target workpiece model WM and the reaction time of the crosslinking reaction under a plurality of set conditions (for example, the set temperature of the mold 10). Based on this correlation, it is possible to calculate the correlation between the elastic modulus Ks of the target workpiece model WM and the set conditions of the crosslinking reaction at a predetermined reaction time.

[0187] 35(a) shows the correlation between the elastic modulus Ks of the target workpiece model WM calculated as described above and the set temperature of the mold 10. The vertical axis represents the ratio of the elastic moduli in the direction perpendicular to the axis, and is not the value of the elastic modulus Ks itself. The graph shown in FIG. 35(a) does not include information about RKs, which is an example of the polymer characteristic value PV of the raw material polymer.

[0188] 35(b) shows actual measurement data showing the change in the elastic modulus in the direction perpendicular to the axis with respect to the set temperature of the mold. The measured RKs of the raw polymer was 191 [N / mm].

[0189] As shown in Fig. 35(c), the graph shown in Fig. 35(a) is shifted in the vertical and / or horizontal directions so that the peak positions of the graph shown in Fig. 35(a) and the peak positions of the graph shown in Fig. 35(b) coincide with each other. In other words, the graph is shifted in the direction of the elastic modulus perpendicular to the axis and / or in the direction of the set temperature. This allows the simulation results obtained by executing the heat transfer analysis process S1, the crosslinking reaction analysis process S2, and the structural analysis process S3 to be calibrated to the actual measurement data, as shown in Fig. 35(c).

[0190] As shown in Figure 34(b), the elastic modulus Ks and RKs have a linear relationship. Therefore, by shifting the graph corresponding to RKs = 191 [N / mm] in the vertical axis direction (the direction of the elastic modulus perpendicular to the axis), it is possible to obtain the correlation between the elastic modulus perpendicular to the axis and the set temperature for raw polymers having different RKs (see Figure 36).

[0191] FIG. 36 shows a graph corresponding to a raw polymer with RKs = 191 [N / mm], a graph corresponding to a raw polymer with RKs = 215 [N / mm], and a graph corresponding to a raw polymer with RKs = 180 [N / mm]. The graph corresponding to the raw polymer with RKs = 215 [N / mm] can be obtained by shifting the graph corresponding to the raw polymer with RKs = 191 [N / mm] upward in the vertical axis direction (the direction of the elastic modulus perpendicular to the axis). Similarly, the graph corresponding to the raw polymer with RKs = 180 [N / mm] can be obtained by shifting the graph corresponding to the raw polymer with RKs = 191 [N / mm] downward in the vertical axis direction (the direction of the elastic modulus perpendicular to the axis). In this way, a correlation between the elastic modulus perpendicular to the axis and the set temperature can be obtained for a raw polymer with any RKs.

[0192] 37(a) shows the simulation results of the elastic modulus in the direction perpendicular to the axis and the set temperature calculated as above, and the actual measurement data of the raw polymer with RKs = 143 [N / mm]. The simulation results are in good agreement with the actual measurement data.

[0193] Similarly, as shown in FIG. 37(b), simulation results that closely match the measured data of the raw polymer with RKs=178 [N / mm] can be obtained.

[0194] 8. Display Process S5 Next, the display process S5 will be described. As shown in Fig. 1, the display unit 7 displays a value obtained from the reaction rate according to the reaction time based on the analysis result of the crosslinking reaction analysis unit 4.

[0195] Furthermore, based on the analysis results of the structural analysis unit 5, the display unit 7 displays characteristics corresponding to the temperature of the mold model MM used in the crosslinking reaction of the polymer part 12 and the in-mold reaction time from the start of the crosslinking reaction to the demolding of the mold model MM.

[0196] Furthermore, the display unit 7 displays whether or not internal bubbles 80 have occurred, based on the estimation results of the internal bubble estimation unit 6, depending on the temperature of the mold model MM and the in-mold reaction time from the start of the crosslinking reaction to the demolding of the mold model MM.

[0197] Furthermore, based on the analysis results of the crosslinking reaction analysis unit 4, the display unit 7 displays a value obtained from the reaction rate according to the temperature of the mold model MM and the in-mold reaction time from the start of the crosslinking reaction to the demolding of the mold model MM, along with the presence or absence of internal bubbles 80. However, the content displayed by the display unit 7 is not limited to the above.

[0198] 8-1. Display Method Using Tables (1) Displaying Reaction Rates With reference to Tables 1 to 4, a method for displaying the reaction rate of the crosslinking reaction using a table will be described.

[0199]

[0200] Table 1 summarizes the relationship between the mold temperature and the reaction time for the crosslinking reaction rate in the element of the target workpiece model WM that has the smallest crosslinking reaction rate.

[0201] The reaction times listed in Table 1 represent reaction times when the target workpiece model WM is placed inside the mold 10. The reaction rate values ​​listed in Table 1 are calculated based on the temperature history of both the reaction time when the target workpiece model WM is placed inside the mold 10 and the time after the target workpiece model WM is released from the mold 10. Since the crosslinking reaction continues even after the target workpiece model WM is released from the mold 10, the above considerations are necessary. In the following explanation, the same applies to the reaction times and reaction rates listed in Tables 2 to 4.

[0202] In Table 1, the region surrounded by a thick line in the upper right corner has a reaction rate greater than 115%. This region is undesirable because the reverse reaction of the crosslinking reaction has progressed excessively. However, the reaction rate range can be set arbitrarily, and a value other than 115% may be used as the threshold.

[0203] The region surrounded by a thick ruled line in the lower left of Table 1 has a reaction rate lower than 95%. This region is not preferable because the crosslinking reaction has not progressed sufficiently. However, the reaction rate range can be set arbitrarily, and a value other than 95% may be used as the threshold.

[0204] In Table 1, the regions surrounded by thick lines, excluding the upper right region and the lower left region, are preferable regions from the viewpoint of the reaction rate of the cross-linking reaction. Note that Table 1 does not include information on internal bubbles 80, which will be described later.

[0205] Table 2 summarizes the relationship between the mold temperature and the reaction time for the average value of the crosslinking reaction of all elements of the target workpiece model WM.

[0206]

[0207] In Table 2, the region surrounded by a thick ruled line in the upper right corner has a reaction rate greater than 115%. This region is undesirable because the reverse reaction of the crosslinking reaction has progressed excessively. However, the reaction rate range can be set arbitrarily, and a value other than 115% may be used as the threshold value.

[0208] The region surrounded by a thick ruled line in the lower left of Table 2 has a reaction rate lower than 95%. This region is not preferable because the crosslinking reaction has not progressed sufficiently. However, the reaction rate range can be set arbitrarily, and a value other than 95% may be used as the threshold.

[0209] In Table 2, the regions other than the upper right region and the lower left region are preferable regions from the viewpoint of the reaction rate of the cross-linking reaction. Note that Table 2 does not include information on the internal bubbles 80, which will be described later.

[0210] Table 3 summarizes the relationship between the mold temperature and the reaction time for the crosslinking reaction rate in the element of the target workpiece model WM that has the highest crosslinking reaction rate.

[0211]

[0212] In Table 3, the region surrounded by a thick line in the upper right corner has a reaction rate greater than 115%. This region is undesirable because the reverse reaction of the crosslinking reaction has progressed excessively. However, the reaction rate range can be set arbitrarily, and a value other than 115% may be used as the threshold.

[0213] The region surrounded by a thick ruled line in the lower left of Table 3 has a reaction rate lower than 95%. This region is not preferable because the crosslinking reaction has not progressed sufficiently. However, the reaction rate range can be set arbitrarily, and a value other than 95% may be used as the threshold.

[0214] In Table 3, the regions other than the upper right region and the lower left region are preferable regions from the viewpoint of the reaction rate of the cross-linking reaction. Note that Table 3 does not include information on the internal bubbles 80, which will be described later.

[0215] Table 4 is Table 2 with the addition of information regarding the internal bubbles 80.

[0216]

[0217] In Table 4, the region surrounded by a thick ruled line in the upper right corner has a reaction rate greater than 115%. This region is undesirable because the reverse reaction of the crosslinking reaction has progressed excessively. However, the reaction rate range can be set arbitrarily, and a value other than 115% may be used as the threshold.

[0218] In Table 4, the area surrounded by the double ruled line indicates that an internal bubble 80 was generated in the polymer portion 12 of the target workpiece model WM.

[0219] In the area below the dashed line in the double-lined area in the lower left of Table 4, the reaction rate is less than 95%. This area is not preferable because the crosslinking reaction has not progressed sufficiently. However, the reaction rate range can be set arbitrarily, and a value other than 95% may be used as the threshold.

[0220] In Table 4, among the regions surrounded by double ruled lines, the region above the dashed line indicates that the reaction rate of the crosslinking reaction was within the range of 95% to 115%, but internal bubbles 80 occurred in the polymer portion 12 of the target workpiece model WM. Since internal bubbles 80 occurred in the polymer portion 12 in this region, it is not suitable for use as a product.

[0221] In this embodiment, Table 4 was created by adding information about the internal bubbles 80 to the prediction results shown in Table 2, but this is not limited to this, and information about the internal bubbles 80 can be added to the prediction results of the reaction rate at any temperature.

[0222] (2) Form of Displaying Elastic Modulus Table 5 is a table summarizing the relationship between the elastic modulus of the target workpiece model WM, the mold temperature, and the reaction time.

[0223] The reaction times listed in Table 5 represent the reaction times when the target workpiece model WM is placed inside the mold 10. The elastic modulus values ​​listed in Table 5 are calculated based on the temperature history of both the reaction time when the target workpiece model WM is placed inside the mold 10 and the time after the target workpiece model WM is released from the mold 10. Since the crosslinking reaction continues even after the target workpiece model WM is released from the mold 10, the above considerations are necessary. Note that in the following explanation, the same applies to the reaction times and elastic moduli listed in Figures 28 to 31.

[0224]

[0225] The elastic modulus is the elastic modulus when a static force is applied in a direction perpendicular to the axis A of the target workpiece model WM. The values ​​in the table are the ratios to the elastic modulus when the cross-linking reaction rate is 100%. However, the direction of the force applied to the target workpiece model WM is not limited, and the elastic modulus when a force is applied in any direction can be used, such as when a static force is applied in a direction parallel to the axis A of the target workpiece model WM, when a static force is applied in a direction rotating around the axis A of the target workpiece model WM, or when a static force is applied in a direction tilting relative to the axis A of the target workpiece model WM. Furthermore, the force applied to the target workpiece model WM is not limited to a static force; for example, it may be the elastic modulus when vibration is applied to the target workpiece model WM.

[0226] Table 5 summarizes the relationship between the elastic modulus obtained as a result of performing a structural analysis in which an elastic modulus corresponding to the degree of vulcanization was assigned to each element of the target workpiece model WM, and the mold temperature and reaction time. The meanings of the thick lines, double lines, and dashed lines in Table 5 are the same as those in Table 4, so duplicate explanations will be omitted. In Table 5, cells filled with "-" indicate where a calculation error occurred in the elastic modulus simulation.

[0227] Additionally, any information can be added to Table 5, such as information about the internal bubbles 80.

[0228] In this embodiment, lines are used to separate the numbers written in the table, but this is not limited to this. The numbers in the table can be separated by any method, such as changing the color of the numbers written in the table, changing the font, changing the typeface to italics or bold, changing the background color of the cells in which the numbers are written, or changing the background pattern of the cells.

[0229] 8-2. Display Method Using Heat Map Diagram 90 A form of displaying the reaction rate of the crosslinking reaction using a heat map diagram 90 will be described with reference to Fig. 27. Fig. 27 shows a heat map diagram 90 that displays the distribution of the reaction rate of the crosslinking reaction for the target workpiece model WM in shades of gray.

[0230] In the heat map diagram 90, the outer bonding member 13 and the inner bonding member 14 of the target workpiece model WM are displayed in white. Also, the polymer portion 12 of the target workpiece model WM is displayed divided into regions 91 to 94 according to the magnitude of the crosslinking reaction rate, and each of the divided regions 91 to 94 is displayed in shades of gray according to the magnitude of the crosslinking reaction rate.

[0231] Region 91 has the highest cross-linking reaction rate and is displayed in the darkest color tone. Region 92 has the second highest cross-linking reaction rate and is displayed in the second darkest color tone. Region 93 has the third highest cross-linking reaction rate and is displayed in the third darkest color tone. Region 94 has the lowest cross-linking reaction rate and is displayed in the lightest color tone.

[0232] The display unit 7 displays a heat map diagram 90 to the operator. The operator can intuitively understand the distribution of the reaction rate of the crosslinking reaction in the polymer portion 12 by visually checking the regions 91 to 94 displayed in shaded areas, which indicate the reaction rate of the crosslinking reaction in the polymer portion 12 of the target work model WM.

[0233] The heat map diagram 90 can be created and output for all reaction conditions (e.g., temperature, time, etc.) related to the target workpiece model WM. Alternatively, the heat map diagram 90 may be created and output only for predetermined reaction conditions.

[0234] The method of displaying the heat map 90 by the display unit 7 is not particularly limited, and for example, the display unit 7 may display the above-mentioned Table 5, and the operator may select each cell in the displayed Table 5, thereby displaying the heat map 90 for the reaction conditions corresponding to the selected cell. Alternatively, the operator may input reaction conditions for the crosslinking reaction, thereby displaying the heat map 90 corresponding to the input reaction conditions.

[0235] In this embodiment, the heat map 90 is configured so that the polymer portion 12 is divided into regions 91 to 94, but the present invention is not limited to this, and the polymer portion 12 may be divided into two to three regions, or five or more regions. Furthermore, in the heat map 90, regions with high reaction rates may be displayed in lighter colors, and regions with low reaction rates may be displayed in darker colors.

[0236] In addition, in the heat map diagram 90, areas with high response rates may be displayed in warm colors such as red, and areas with low response rates may be displayed in cool colors such as blue, or areas with high response rates may be displayed in cool colors such as blue, and areas with low response rates may be displayed in warm colors such as red.

[0237] 8-3. Graphical Display Method (1) Display of Elastic Modulus A graphical display of the elastic modulus of the target workpiece model WM will be described with reference to FIGS. 28 to 30. FIGS. 28 to 30 are graphs of the elastic modulus of the target workpiece model WM versus the reaction time of the crosslinking reaction. FIG. 28 is a graph showing the crosslinking reaction of the polymer portion 12 of the target workpiece model WM at the lowest mold 10 temperature, FIG. 29 is a graph showing the crosslinking reaction of the polymer portion 12 of the target workpiece model WM at an intermediate mold 10 temperature, and FIG. 30 is a graph showing the crosslinking reaction of the polymer portion 12 of the target workpiece model WM at the highest mold 10 temperature. In FIGS. 28 to 30, the solid lines indicate the measured values ​​of the elastic modulus, and the dashed lines indicate the predicted values.

[0238] As shown in FIGS. 28 to 30, the modulus of elasticity increases after the crosslinking reaction starts, reaches a maximum value, and then decreases after reaching the maximum value.

[0239] In FIG. 28, where the temperature of the mold 10 is the lowest, the measured and predicted values ​​of the elastic modulus matched with good precision.

[0240] In Figure 29, where the temperature of the mold 10 is moderate, the measured and predicted values ​​of the elastic modulus did not match for the first two minutes after the start of the reaction, but after two minutes, the measured and predicted values ​​matched accurately. Up to two minutes after the start of the reaction, the predicted elastic modulus was smaller than the measured value. However, since the region within two minutes after the start of the reaction is not expected to be used as a product, the impact on the product is minimal.

[0241] In Figure 30, where the mold 10 temperature is the highest, the measured and predicted values ​​of the elastic modulus did not match for up to two minutes after the start of the reaction, but after two minutes, the measured and predicted values ​​matched accurately. Up to two minutes after the start of the reaction, the predicted elastic modulus was smaller than the measured value. The difference between the measured and predicted values ​​was larger than when the mold 10 temperature was intermediate. However, as mentioned above, the region within two minutes after the start of the reaction is not expected to be used as a product, so there is little impact on the product.

[0242] (2) Displaying both the elastic modulus and the internal bubbles 80 With reference to Fig. 31 , a description will be given of a displaying, using a graph, the elastic modulus of the polymer portion 12 of the target workpiece model WM and the predicted results for the occurrence of internal bubbles 80. The temperature of the mold 10 is set to a medium temperature, as in Fig. 29 described above.

[0243] 31 shows a graph of the elastic modulus versus reaction time when a force is applied in a direction perpendicular to the axis A of the target workpiece model WM. Square symbols indicate the elastic modulus in the element where the temperature of the polymer portion 12 of the target workpiece model WM is the highest. Circle symbols indicate the elastic modulus in the element where the temperature of the polymer portion 12 of the target workpiece model WM is the median. Triangular symbols indicate the elastic modulus in the element where the temperature of the polymer portion 12 of the target workpiece model WM is the lowest. In addition, in FIG. 31, open symbols indicate cases where it is predicted that an internal bubble 80 will occur, and filled symbols indicate cases where it is predicted that an internal bubble 80 will not occur.

[0244] The graphs for the element where the temperature of the polymer portion 12 is the highest value and the element where the temperature is the median value are graphs for the reaction return period after the reaction progress period.

[0245] The cross-linking reaction proceeds most quickly in the element where the temperature of the polymer portion 12 is the highest. Therefore, it was predicted that internal bubbles 80 would not be generated when the reaction time exceeded 3 minutes, i.e., 3 minutes and 30 seconds had elapsed. In the element where the temperature of the polymer portion 12 was the highest, the reversal reaction of the cross-linking reaction also proceeded quickly, and therefore the modulus of elasticity was lowest when the reaction time was 6 minutes.

[0246] In the element where the temperature of the polymer portion 12 is the median, the crosslinking reaction progresses at a moderate level. Therefore, it was predicted that internal bubbles 80 would not be generated when the reaction time exceeded 3 minutes 30 seconds and 4 minutes had elapsed. The elastic modulus after 6 minutes of reaction time was greater than that of the element where the temperature of the polymer portion 12 was the highest.

[0247] In the element where the temperature of the polymer portion 12 is the lowest, the crosslinking reaction progresses the slowest. Therefore, before the reaction time is 4 minutes, the reaction is in the progress phase, and it is predicted that the modulus of elasticity will increase as the reaction time passes. After the reaction time has passed 4 minutes, the reversal reaction of the crosslinking reaction will progress, and it is predicted that the modulus of elasticity will gradually decrease. In the element where the temperature of the polymer portion 12 is the lowest, it is predicted that no internal bubbles 80 will occur when the reaction time exceeds 5 minutes and 5 minutes and 30 seconds have passed. At this point, it is predicted that the internal bubbles 80 will have disappeared in the entire region of the polymer portion 12 of the target workpiece model WM. It is predicted that the modulus of elasticity at the reaction time of 6 minutes will be the largest in the entire region of the polymer portion 12.

[0248] As described above, it was predicted that the internal bubbles 80 would have disappeared in the entire region of the polymer portion 12 of the target workpiece model WM when the reaction time had reached 5 minutes and 30 seconds, but considering the risk, the reaction time of 6 minutes was adopted as the reaction time of this target workpiece model WM. At the reaction time of 6 minutes, the variation in the elastic modulus of the polymer portion 12 was predicted to be 12%.

[0249] Thus, according to this embodiment, it is possible to predict the reaction time at which the internal bubbles 80 no longer occur for the target workpiece model WM, and also to predict the variation in the elastic modulus of the polymer portion 12 during that reaction time.

[0250] (3) Form for visually displaying internal bubbles 80 With reference to FIG. 32, a form for visually displaying the predicted results of the generation of internal bubbles 80 in the polymer portion 12 of the target workpiece model WM will be described.

[0251] 32 is a diagram using a cross-sectional view of the target workpiece model WM to indicate, by using shading, the degree of likelihood of internal bubbles 80 occurring in the polymer portion 12. The outer bonding member 13 and the inner bonding member 14 of the target workpiece model WM are shown by the darkest pattern in FIG.

[0252] The polymer portion 12 of the target workpiece model WM is shown with the lightest pattern in Fig. 32. An elliptical region P and an elliptical region Q located within region P are displayed inside the polymer portion 12. Region P is displayed with a pattern that is slightly darker than the light pattern representing the polymer portion 12. Region Q is displayed with a pattern that is slightly darker than region P.

[0253] An example will be described in which, as a result of the internal bubble estimation process S4, region P is determined to be a region where internal bubbles 80 are relatively likely to occur, and region Q is determined to be a region where internal bubbles 80 are more likely to occur than region P. Regions P and Q actually correspond to regions where internal bubbles 80 are likely to occur (see FIG. 23 ).

[0254] When the internal bubble estimation process S4 is completed, the display unit 7 displays the target workpiece model WM shown in FIG. 32. By looking at FIG. 32 displayed on the display unit 7, the operator visually recognizes the outer and inner bonding members 13 and 14, which are indicated by relatively dark patterns, and the polymer portion 12, which is indicated by a relatively light pattern. The operator can intuitively understand that the outer and inner bonding members 13 and 14, which are indicated by dark patterns, are different from the polymer portion 12. This allows the operator to intuitively understand that in order to search for the internal bubble 80, it is sufficient to check the polymer portion 12, which is indicated by a relatively light pattern. However, the outer and inner bonding members 13 and 14 may also be displayed in white.

[0255] Next, the worker visually recognizes that regions P and Q, which are displayed with a relatively dark pattern, are displayed within the polymer portion 12, which is displayed with a light pattern. This allows the worker to intuitively understand that internal bubbles 80 are likely to occur in regions P and Q of the polymer portion 12. Furthermore, because region Q is displayed with a darker pattern than region P, the worker can intuitively understand that internal bubbles 80 are most likely to occur in region P. This allows the worker to efficiently consider the manufacturing conditions for the target workpiece model WM by simply searching for internal bubbles 80 in regions where internal bubbles 80 are likely to occur. However, regions where internal bubbles 80 are likely to occur may also be displayed with a relatively light pattern.

[0256] 8-4. Embodiment using a graph for correlation between set temperature and elastic modulus Ks Figure 38 shows a graph relating to the correlation between elastic modulus Ks and set temperature for crosslinking reaction for raw material polymers having multiple (three in this embodiment) different RKs.

[0257] The horizontal axis of the graph shown in Figure 38 is the set temperature of the mold in which the crosslinking reaction is carried out. A predetermined manufacturing range is set for the set temperature. The crosslinking reaction is carried out at a set temperature within the manufacturing range.

[0258] The vertical axis of the graph shown in Figure 38 is the elastic modulus Ks of the target workpiece model WM. A standard is set for the elastic modulus Ks. The elastic modulus Ks of the target workpiece model WM is required to be a value within the range between the upper and lower limits of the standard. If Ks is greater than the upper limit of the standard, or if Ks is smaller than the lower limit of the standard, the target workpiece model WM is deemed to be a defective product.

[0259] There is a median between the upper and lower limits of the specification. The elastic modulus Ks is preferably a value between the upper and lower limits of the specification, and more preferably near the median.

[0260] 38, when RKs has a relatively large value, if the crosslinking reaction is carried out at a set temperature close to the lower limit of the manufacturing range, it is conceivable that the elastic modulus Ks of the target workpiece model WM will be greater than the upper limit of the specification. In this case, by setting the set temperature closer to the upper limit of the manufacturing range based on the simulation results, it is possible to make the elastic modulus Ks of the target workpiece model WM smaller than the upper limit of the specification.

[0261] On the other hand, when RKs has a relatively small value, if the crosslinking reaction is carried out at a set temperature close to the upper limit of the manufacturing range, it is conceivable that the elastic modulus Ks of the target workpiece model WM will be smaller than the lower limit of the specification. In this case, by setting the set temperature closer to the lower limit of the manufacturing range based on the simulation results, it is possible to make the elastic modulus Ks of the target workpiece model WM larger than the lower limit of the specification.

[0262] Furthermore, when RKs is in the middle range, if the crosslinking reaction is carried out at a set temperature within the manufacturing range, the elastic modulus Ks of the target workpiece model WM may fall within the range between the upper and lower limits of the specification. Even in such cases, the set temperature can be adjusted based on the simulation results so that the elastic modulus Ks of the target workpiece model WM is close to the median value between the upper and lower limits of the specification.

[0263] The display unit 7 may display to the operator the graph shown in Fig. 38. Furthermore, the display unit 7 may display to the operator the numerical value of the set temperature that is appropriately set in Fig. 38.

[0264] The setting condition analysis unit may be configured to transmit the value of the set temperature appropriately set in Figure 38 to a crosslinking reaction device (not shown) without displaying it on the display unit 7, and the crosslinking reaction device may then perform a crosslinking reaction on the raw polymer based on the received value.

[0265] (Embodiment 1-2) Next, embodiment 1-2 will be described with reference to Fig. 39. This embodiment differs from embodiment 1-2 in that the reaction time of the crosslinking reaction of the polymer portion 12 is used as the polymer characteristic value PV.

[0266] By executing the heat transfer analysis process S1, the crosslinking reaction analysis process S2, and the structural analysis process S3 described in Figure 4 of embodiment 1-1, it is possible to predict the correlation between the elastic modulus Ks of the target work model WM and the reaction time of the crosslinking reaction under multiple set conditions (e.g., the set temperature of the mold 10).

[0267] Figure 39(a) shows the correlation between the elastic modulus Ks of the target workpiece model WM calculated as described above and the reaction time of the crosslinking reaction. The vertical axis represents the ratio of the elastic modulus in the direction perpendicular to the axis, not the value of the elastic modulus Ks itself. The graph shown in Figure 35(a) does not include information about RKs, which is an example of the polymer characteristic value PV of the raw polymer.

[0268] On the other hand, Fig. 39(b) shows actual measurement data showing the change in the elastic modulus in the direction perpendicular to the axis with respect to the reaction time of the crosslinking reaction. Although not described in detail, the raw polymer has a predetermined RKs.

[0269] As shown in Fig. 39(c), the graph shown in Fig. 39(a) is shifted in the vertical and / or horizontal directions so that the peak positions of the graph shown in Fig. 39(a) and the peak positions of the graph shown in Fig. 39(b) coincide with each other. In other words, the graph is shifted in the direction of elastic modulus perpendicular to the axis and / or in the direction of reaction time. This allows the simulation results obtained by executing the heat transfer analysis process S1, the crosslinking reaction analysis process S2, and the structural analysis process S3 to be calibrated to the actual measurement data, as shown in Fig. 39(c).

[0270] Note that the configuration other than the above is substantially the same as in embodiment 1-1, and therefore a duplicated description will be omitted. Furthermore, among the symbols used in the following embodiments, the same symbols as those used in the previous embodiments represent the same components as those in the previous embodiments, unless otherwise specified.

[0271] (Embodiment 1-3) Next, embodiment 1-3 will be described with reference to Fig. 40. In this embodiment, a table is used to calibrate the simulation results, and the simulation results are calibrated by shifting the setting conditions and the reaction time.

[0272] First, in this embodiment, by executing the heat transfer analysis process S1, the crosslinking reaction analysis process S2, and the structural analysis process S3 described in FIG. 4 of embodiment 1-1, Table 6 corresponding to Table 5 of embodiment 1-1 is created.

[0273] As shown in Figure 40, Table 6 summarizes the relationship between the elastic modulus Ks of the target workpiece model WM, the set temperature for the crosslinking reaction, and the reaction time for the crosslinking reaction. The set temperature is not particularly limited, and may be the temperature of the mold 10, the temperature of the insert, or the temperature of another component. Furthermore, the elastic modulus Ks may be the elastic modulus of the target workpiece model WM in the direction perpendicular to the axis or in the axial direction.

[0274] As shown in Table 6, when the set temperature is T1 and the reaction time is t1, the elastic modulus Ks of the target workpiece model WM is α1. Hereinafter, when the set temperature is T1 and the reaction time is tm, the elastic modulus Ks of the target workpiece model WM is αn, where l, m, and n are natural numbers greater than or equal to 1. Table 6 does not include information about RKs.

[0275] For the Table 6 obtained as described above, a crosslinking reaction is carried out using a raw material polymer having a predetermined RKs, and actual measurement data is obtained for the modulus of elasticity, set temperature, and reaction time. Based on the obtained actual measurement data, the set temperature and reaction time listed in Table 6 are changed. Specifically, the set temperature and reaction time are shifted to match the value of the modulus of elasticity Ks of the actual measurement data. In this way, Table 7 is created.

[0276] As shown in Table 7, when the set temperature is T'1 and the reaction time is t', the elastic modulus Ks of the target workpiece model WM is α1. Hereinafter, when the set temperature is T'1 and the reaction time is t'm, the elastic modulus Ks of the target workpiece model WM is αn, where 1, m, and n are natural numbers greater than or equal to 1. This allows the simulation results obtained by executing the heat transfer analysis process S1, the crosslinking reaction analysis process S2, and the structural analysis process S3 to be calibrated to match the actual measurement data.

[0277] The configuration other than that described above is substantially the same as that of embodiment 1-1, so a duplicated description will be omitted.

[0278] (Embodiment 1-4) Next, embodiment 1-4 will be described with reference to Fig. 41. In this embodiment, a table is used to calibrate the simulation results, and the simulation results are calibrated by shifting the reaction time.

[0279] First, in this embodiment, by executing the heat transfer analysis process S1, the crosslinking reaction analysis process S2, and the structural analysis process S3 described in FIG. 4 of embodiment 1-1, Table 6 corresponding to Table 5 of embodiment 1-1 is created.

[0280] For the Table 6 obtained as described above, a crosslinking reaction is carried out using a raw material polymer having a predetermined RKs, and actual measurement data is obtained for the modulus of elasticity, set temperature, and reaction time. Based on the obtained actual measurement data, the reaction time listed in Table 6 is changed. Specifically, the reaction time is shifted to match the value of the modulus of elasticity Ks of the actual measurement data. Table 8 is thus prepared.

[0281] As shown in Table 8, when the set temperature is T1 and the reaction time is t'1, the elastic modulus Ks of the target workpiece model WM is α1. Hereinafter, when the set temperature is T1 and the reaction time is t'm, the elastic modulus Ks of the target workpiece model WM is αn, where l, m, and n are natural numbers greater than or equal to 1. This allows the simulation results obtained by executing the heat transfer analysis process S1, the crosslinking reaction analysis process S2, and the structural analysis process S3 to be calibrated to match the actual measurement data.

[0282] The configuration other than that described above is substantially the same as that of embodiment 1-1, so a duplicated description will be omitted.

[0283] (Embodiment 1-5) Next, embodiment 1-5 will be described with reference to Fig. 42. In this embodiment, a table is used to calibrate the simulation results, and the simulation results are calibrated by shifting the set temperature.

[0284] First, in this embodiment, by executing the heat transfer analysis process S1, the crosslinking reaction analysis process S2, and the structural analysis process S3 described in FIG. 4 of embodiment 1-1, Table 6 corresponding to Table 5 of embodiment 1-1 is created.

[0285] For the Table 6 obtained as described above, a crosslinking reaction is carried out using a raw material polymer having a predetermined RKs, and actual measurement data is obtained for the modulus of elasticity, set temperature, and reaction time. Based on the obtained actual measurement data, the set temperature listed in Table 6 is changed. Specifically, the set temperature is shifted to match the value of the modulus of elasticity Ks of the actual measurement data. In this way, Table 9 is created.

[0286] As shown in Table 9, when the set temperature is T'1 and the reaction time is t1, the elastic modulus Ks of the target workpiece model WM is α1. Hereinafter, when the set temperature is T'1 and the reaction time is tm, the elastic modulus Ks of the target workpiece model WM is αn, where 1, m, and n are natural numbers greater than or equal to 1. This allows the simulation results obtained by executing the heat transfer analysis process S1, the crosslinking reaction analysis process S2, and the structural analysis process S3 to be calibrated to match the actual measurement data.

[0287] The configuration other than that described above is substantially the same as that of embodiment 1-1, so a duplicated description will be omitted.

[0288] (Embodiment 1-6) Next, embodiment 1-6 will be described with reference to Fig. 43. In this embodiment, a table is used to calibrate the simulation results, and the simulation results are calibrated by shifting the set temperature, reaction time, and target workpiece model elastic modulus Ks.

[0289] First, the heat transfer analysis process S1, the crosslinking reaction analysis process S2, and the structural analysis process S3 shown in FIG. 4 of the embodiment 1-1 are executed to create Table 6, which corresponds to Table 5 of the embodiment 1-1.

[0290] For the Table 6 obtained as described above, a crosslinking reaction is carried out using a raw material polymer having a predetermined RKs, and actual measurement data is obtained for the modulus of elasticity, set temperature, and reaction time. Based on the obtained actual measurement data, the set temperature, reaction time, and modulus of elasticity Ks listed in Table 6 are changed. Specifically, the set temperature, reaction time, and modulus of elasticity Ks are shifted to match the value of the modulus of elasticity Ks of the actual measurement data. In this way, Table 10 is created.

[0291] As shown in Table 10, when the set temperature is T'1 and the reaction time is t'1, the elastic modulus Ks of the target workpiece model WM is α'1. Hereinafter, when the set temperature is T'l and the reaction time is t'm, the elastic modulus Ks of the target workpiece model WM is α'n, where l, m, and n are natural numbers greater than or equal to 1. This allows the simulation results obtained by executing the heat transfer analysis process S1, the crosslinking reaction analysis process S2, and the structural analysis process S3 to be calibrated to match the actual measurement data.

[0292] This embodiment differs from the above-described embodiments 1-3 in that the elastic modulus Ks of the target workpiece model WM is also shifted. For example, as in embodiments 1-3, the elastic modulus Ks is fixed, and simulation results and actual measurement data are generated. In this case, it is conceivable that the set temperature T'l and reaction time t'm may be numerical values ​​that include decimals. In such a case, by shifting the elastic modulus Ks so that the corrected set temperature T'l and reaction time t'm become integers, Table 10 can be made easier for operators to understand.

[0293] The configuration other than that described above is substantially the same as that of embodiment 1-1, so a duplicated description will be omitted.

[0294] (Embodiment 1-7) Next, embodiment 1-7 will be described with reference to Fig. 44. In this embodiment, a table is used to calibrate the simulation results, and the simulation results are calibrated by shifting the target workpiece model elastic modulus Ks for one cell in the table.

[0295] First, in this embodiment, by executing the heat transfer analysis process S1, the crosslinking reaction analysis process S2, and the structural analysis process S3 described in FIG. 4 of embodiment 1-1, Table 6 corresponding to Table 5 of embodiment 1-1 is created.

[0296] For the Table 6 obtained as described above, a crosslinking reaction is carried out using a raw polymer having a predetermined RKs, and actual measurement data is obtained for the elastic modulus, set temperature, and reaction time. Based on the obtained actual measurement data, only one datum for the elastic modulus Ks listed in Table 6 is changed. Specifically, only one value of the elastic modulus Ks is shifted to match the value of the elastic modulus Ks of the actual measurement data. In this way, Table 11 is created.

[0297] As shown in Table 11, when the set temperature is T2 and the reaction time is t3, the elastic modulus Ks of the target workpiece model WM is β'3. This allows the simulation results obtained by executing the heat transfer analysis process S1, the crosslinking reaction analysis process S2, and the structural analysis process S3 to be calibrated to match the actual measurement data. According to this embodiment, it is only necessary to configure the desired value of the elastic modulus Ks, thereby shortening the calculation time.

[0298] The configuration other than that described above is substantially the same as that of embodiment 1-1, so a duplicated description will be omitted.

[0299] 45 to 50, an embodiment 2-1 will be described. In this embodiment, the elastic modulus Ks of a target workpiece model is predicted using a simulation result related to the reaction rate of a crosslinking reaction and machine learning. In the machine learning, a number of trained models up to the same as the number of elements of the target workpiece model WM is generated, and a number of reaction rates of the crosslinking reaction up to the same as the number of elements is used.

[0300] The crosslinking reaction simulator 1a according to this embodiment includes an arithmetic processing unit 9 that performs machine learning. The crosslinking reaction simulator 1a according to this embodiment generates a trained model TM by performing machine learning using training data TRD for learning selected from data related to the reaction rate of the crosslinking reaction of the polymer portion 12, and predicts the target workpiece model elastic modulus Ks using this trained model TM.

[0301] As shown in Fig. 45 , the crosslinking reaction simulator 1a according to this embodiment includes a calculation processing unit 9. The calculation processing unit 9 performs machine learning. In Fig. 45 , the heat transfer analysis unit 3, the crosslinking reaction analysis unit 4, the structural analysis unit 5, the internal bubble estimation unit 6, and the setting condition analysis unit 8 are the same as those in Fig. 1a, and therefore the same reference numerals are used and redundant explanations will be omitted.

[0302] The calculation processing unit 9 generates a trained model TM by performing machine learning using the training data TRD. The calculation processing unit 9 also predicts a target variable using the trained model TM. The target variable is not particularly limited, and any target variable can be appropriately selected, such as the final reaction rate of the polymer portion 12, the reaction rate of the polymer portion 12 at the time of demolding, the final equivalent reaction amount of the polymer portion 12, the equivalent reaction amount of the polymer portion at the time of demolding, the elastic modulus Ks of the target workpiece model WM, or the occurrence of internal bubbles in the polymer portion 12.

[0303] As shown in Figure 46, the memory unit 2 stores training data TRD for learning selected from data regarding the reaction rate or equivalent reaction amount of the crosslinking reaction of the polymer portion 12, a learned model TM generated by performing machine learning using the training data TRD, and a threshold value TH.

[0304] Next, the operation of the crosslinking reaction simulator 1a according to this embodiment will be described. Figure 47 shows a main flow of the operation of the crosslinking reaction simulator 1a according to this embodiment. As shown in Figure 47, when the crosslinking reaction simulator 1a according to this embodiment is started, a simulation process S10 is executed. Next, a prediction process S20 is executed, and the operation ends. However, before executing the simulation process S10, a step of determining analysis conditions by experimental design may be executed. By using data selected by experimental design, the simulation process S10 can be executed efficiently.

[0305] As shown in FIG. 48, the simulation process S10 according to this embodiment is the same as the main flow shown in FIG. 4 of the first embodiment, so the same processes are given the same reference numerals and redundant explanations will be omitted.

[0306] 49 shows a flowchart of the prediction process according to this embodiment. When the prediction process (S20) starts, the calculation processing unit 9 generates training data (S210). More specifically, the training data is extracted from the multiple reaction rates calculated by the reaction rate calculation processing unit 41. The training data may be extracted randomly from the multiple reaction rates, or may be extracted based on a predetermined rule that takes into account the distribution of data related to the multiple reaction rates, or may be extracted based on, for example, experimental design.

[0307] Next, the calculation processing unit 9 generates a first trained model (S211). The first trained model is an example of a trained model. The calculation processing unit 9 generates a first trained model for each element corresponding to multiple analysis meshes of the target work model WM. In other words, up to the same number of first trained models as the number of analysis meshes of the target work model WM are generated. The first trained model predicts the final reaction rate of the polymer portion 12 from the setting conditions for the crosslinking reaction of the polymer portion 12. Note that the final reaction rate is the final reaction rate of element number n.

[0308] Next, the calculation processing unit 9 generates a second trained model (S211). The second trained model is an example of a trained model. The calculation processing unit 9 predicts the elastic modulus Ks of the target workpiece model from the set conditions and the final reaction rates of each element of the multiple analysis meshes of the target workpiece model WM, which are predicted based on the multiple first trained models.

[0309] Next, the calculation processing unit 9 predicts the reaction rate (S213). As a result, the calculation processing unit 9 predicts the final reaction rate for each element corresponding to the multiple analysis meshes of the target workpiece model WM based on the set conditions and the first and second trained models.

[0310] Next, the calculation processing unit 9 calculates the coefficient of determination R ２ is calculated (S214). ２ The following equation (19) is used, where the coefficient of determination R ２ Alternatively, a different definition may be used. The index of prediction accuracy is the coefficient of determination R ２However, the index is not limited to the above, and any index such as mean absolute error (MEA) or root mean square error (RMSE) can be appropriately selected.

[0311]

[0312] where y is the sample value, f is the estimated value based on the forecast, and y ａｖ is the mean value of the sample values.

[0313] Next, the calculation processing unit 9 calculates the coefficient of determination R of the first trained model. ２ It is determined whether (first) is smaller than a threshold value TH (S215). The value of the threshold value TH is not particularly limited, and any value can be appropriately adopted. The threshold value TH is preferably 0.6 to 0.9, and more preferably 0.65 to 0.75. In this embodiment, the threshold value TH is set to 0.7.

[0314] Coefficient of determination R of the first trained model ２ If (first) is smaller than the threshold value TH (S215: Y), the simulation process (S218) is executed for all conditions. The contents of the simulation process (S218) are the same as the simulation process (S10) of FIG. 48, so a duplicated explanation will be omitted. As a result, the target workpiece model elastic modulus Ks is calculated. This completes the prediction process (S20).

[0315] On the other hand, the first trained model R ２ is not smaller than the threshold value TH (S215: N), the calculation processing unit 9 calculates the coefficient of determination R ２ It is determined whether (second) is smaller than a threshold value (S216).

[0316] Coefficient of determination R of the second trained model ２ If (second) is smaller than the threshold value TH (S216: Y), the simulation process (S218) is executed for all conditions. As a result, the target workpiece model elastic modulus Ks is calculated. This completes the prediction process (S20).

[0317] On the other hand, the coefficient of determination R of the second trained model ２If (second) is not smaller than the threshold value TH (S216: N), the calculation processing unit 9 predicts the final reaction rate of each element corresponding to the multiple analysis meshes of the target workpiece model WM using the set conditions and the first trained model. Next, the calculation processing unit 9 predicts the target workpiece model elastic modulus Ks using the set conditions and the final reaction rate of each element corresponding to the multiple analysis meshes of the target workpiece model WM using the second trained model (S217). This completes the prediction process (S20).

[0318] Figure 50 shows the correlation when the horizontal axis represents the final reaction rate of element number n of the polymer portion 12 obtained by simulation processing, and the vertical axis represents the final reaction rate of element number n predicted by machine learning. The closer the multiple points shown in Figure 51 are to the straight line extending diagonally upward to the right from the origin, the higher the correlation between the simulation results and the machine learning results, indicating higher prediction accuracy of the machine learning. According to this embodiment, a high correlation was obtained when the reaction rate was 1.0 or higher.

[0319] According to this embodiment, the final reaction rate can be predicted by machine learning, and the elastic modulus Ks of the target workpiece model can be predicted by machine learning, thereby reducing the calculation time for the elastic modulus Ks of the target workpiece model.

[0320] Furthermore, by using the target work model elastic modulus Ks predicted as described above to predict the setting conditions in the same manner as in embodiment 1, the calculation time for the setting conditions can be shortened overall.

[0321] (Embodiment 2-2) Next, embodiment 2-2 will be described with reference to Fig. 51 to Fig. 53. In this embodiment, the elastic modulus Ks of a target workpiece model is predicted using simulation results related to the reaction rate of a crosslinking reaction and machine learning. In the machine learning, a trained model that predicts multiple statistics for the reaction rate of a crosslinking reaction is generated, and the same number of reaction rates of the crosslinking reaction as the statistics are used.

[0322] 51 shows a part of a flowchart of the prediction process of this embodiment. When the prediction process (S20) of this embodiment starts, the calculation processing unit 9 creates training data (S2200). The process of S2200 of this embodiment is the same as S210 of FIG. 49, so a duplicated explanation will be omitted.

[0323] Next, the calculation processing unit 9 generates a first trained model (S2201), a third trained model (S2202), and a fourth trained model (S2203). The processing of S2201 is the same as S210 in Figure 49, so a duplicated explanation will be omitted. However, the fourth trained model may be generated after S2207 and before S2209, which will be described later.

[0324] The third trained model predicts the elastic modulus Ks of the target workpiece model based on the set conditions and at least one of the average value of the final reaction rates of each element corresponding to the multiple analysis meshes of the target workpiece model WM, the maximum value, the minimum value, and the standard deviation of the final reaction rates of each element corresponding to the multiple analysis meshes of the target workpiece model WM. However, the statistical quantities to be predicted are not limited to those described above, and any statistical quantity can be predicted.

[0325] The fourth trained model is an example of a trained model. The calculation processing unit 9 predicts at least one of the average value, maximum value, minimum value, and standard deviation for the final reaction rate of each element corresponding to multiple analysis meshes of the target workpiece model WM. The number of fourth trained models is the total of those that predict at least one of the average value, maximum value, minimum value, and standard deviation of the final reaction rate. Therefore, the number of fourth trained models is at most one. When using one fourth trained model, the average value is preferably used as the predicted statistic. However, the predicted statistic is not limited to the above, and any statistic may be predicted.

[0326] Next, the calculation processing unit 9 predicts the reaction rate (S2204). As a result, the calculation processing unit 9 predicts the final reaction rate for each element corresponding to the multiple analysis meshes of the target workpiece model WM based on the set conditions and the first trained model, the third trained model, and the fourth trained model. However, the reaction rate using the fourth trained model may be predicted after S2207 and before S2209, which will be described later.

[0327] Next, the calculation processing unit 9 calculates the coefficient of determination R ２ (S2205), where the coefficient of determination R ２ The prediction may be made after S2207 and before S2209, which will be described later.

[0328] Next, the calculation processing unit 9 calculates the coefficient of determination R of the first trained model. ２ It is determined whether (first) is smaller than a threshold value TH (S2206).

[0329] Coefficient of determination R of the first trained model ２ If (first) is smaller than the threshold value TH (S2206: Y), the simulation process (S2208) is executed for all conditions. The contents of the simulation process (S2208) are the same as the simulation process (S10) in FIG. 48, so redundant explanations will be omitted. As a result, the target workpiece model elastic modulus Ks is calculated. This completes the prediction process (S20) (see FIG. 52).

[0330] On the other hand, the first trained model R ２ is not smaller than the threshold value TH (S2206: N), the calculation processing unit 9 calculates the coefficient of determination R ２ It is determined whether (third) is smaller than the threshold value (S2207).

[0331] Coefficient of determination R of the third trained model ２If (third) is smaller than the threshold value TH (S2207: Y), the simulation process (S2208) is executed for all conditions. As a result, the target workpiece model elastic modulus Ks is calculated. This completes the prediction process (S20) (see FIG. 52).

[0332] On the other hand, the coefficient of determination R of the third trained model ２ If (third) is not smaller than the threshold value TH (S2207: N), as shown in FIG. 52 , the calculation processing unit 9 calculates the coefficient of determination R ２ It is determined whether (fourth) is smaller than the threshold value TH (S2209).

[0333] Coefficient of determination R of the fourth trained model ２ If (fourth) is not smaller than the threshold value TH (S2209: N), the calculation processing unit 9 selects the third trained model and the fourth trained model (S2210).

[0334] Next, the calculation processing unit 9 uses the fourth trained model to predict at least one of the average value, maximum value, minimum value, and standard deviation for the final reaction rate of each element corresponding to multiple analysis meshes of the target work model WM (S2211).

[0335] Next, the calculation processing unit 9 executes a target workpiece model elastic modulus prediction process (S2212). As a result, the calculation processing unit 9 predicts the target workpiece model elastic modulus Ks based on the third trained model, the setting conditions, and at least one of the average value, maximum value, minimum value, and standard deviation for the final reaction rate of each element corresponding to the multiple analysis meshes of the target workpiece model WM. However, the statistical quantity is not limited to the above, and any statistical quantity can be predicted. This completes the prediction process (S20).

[0336] On the other hand, the coefficient of determination R ２ If (fourth) is smaller than the threshold value TH (S2209: Y), as shown in FIG. 53, the calculation processing unit 9 selects the first trained model (S2213).

[0337] Next, the calculation processing unit 9 calculates the final reaction rates for the elements corresponding to the multiple analysis meshes of the target workpiece model WM (S2214).

[0338] Next, the calculation processing unit 9 calculates statistics for elements corresponding to a plurality of analysis meshes of the target workpiece model WM (S2215).

[0339] Next, the calculation processing unit 9 calculates the coefficient of determination R based on the predicted value of the statistics of the final reaction rate calculated in S2215 and the analytical value. ２ (statistics) is calculated (S2216).

[0340] Next, the calculation processing unit 9 calculates the coefficient of determination R ２ It is determined whether (statistics) is smaller than a threshold value TH (S2217).

[0341] Coefficient of determination R ２ If (statistic) is smaller than the threshold value TH (S2217: Y), the simulation process (S2220) is executed for all conditions. This calculates the target workpiece model elastic modulus Ks. This completes the prediction process (S20) (see FIG. 52).

[0342] On the other hand, the coefficient of determination R ２ If (statistics) is not smaller than the threshold value TH (S2217: N), the calculation processing unit 9 selects the third trained model (S2218).

[0343] Next, the calculation processing unit 9 predicts the target workpiece model elastic modulus Ks using the third trained model and the statistics calculated in S2215 (S2219). This completes the prediction process (S20) (see FIG. 52).

[0344] According to this embodiment, a minimum of two fourth trained models are used, thereby shortening the calculation time for the target work model elastic modulus Ks.

[0345] (Modification of Embodiment 2-2) Next, a modification of Embodiment 2-2 will be described with reference to Fig. 54 to Fig. 55. Fig. 54 and Fig. 55 show a part of a flowchart of the preheating process according to this modification. The flowchart shown in Fig. 54 differs from the flowchart in Fig. 51 in that S2201 and S2206 are omitted.

[0346] The flowchart in FIG. 55 differs from the flowchart in FIG. 53 in that S2201, S2204, S2205, and S2206 are included. In FIG. 55, a first trained model is generated (S2201). Next, the response rate is predicted (S2204). After that, the coefficient of determination R ２ Next, the calculation processing unit 9 calculates the coefficient of determination R of the first trained model (S2205). ２ It is determined whether (first) is smaller than a threshold value TH (S2206). ２ If (first) is not smaller than the threshold value TH (S2206: N), the first trained model is selected (S2213).

[0347] On the other hand, the coefficient of determination R of the first trained model ２ If (first) is smaller than the threshold value TH (S2206: Y), a simulation process (S2220) is executed.

[0348] Other than the above, the embodiment is the same as in embodiment 2-2, so a duplicated description will be omitted.

[0349] According to this embodiment, the first trained model is adopted when the accuracy of the third trained model and the fourth trained model is not sufficiently high. The first trained model generates multiple predictive models, so it takes time to generate the models. Therefore, according to this embodiment, when the accuracy of the third trained model and the fourth trained model is sufficiently high, the generation of the first trained model can be omitted, thereby speeding up processing overall.

[0350] 56 to 57, an embodiment 2-3 will be described. In this embodiment, the elastic modulus Ks of a target workpiece model is predicted using machine learning and simulation results related to the equivalent reaction amount of a crosslinking reaction. In the machine learning, a maximum number of trained models equal to the number of elements of the target workpiece model WM is generated, and a maximum number of equivalent reaction amounts equal to the number of elements is used.

[0351] When the prediction process (S230) according to this embodiment starts, the calculation processing unit 9 generates training data (S230). The calculation processing unit 9 extracts training data from the multiple equivalent reaction amounts calculated by the reaction rate calculation processing unit 41. The training data may be extracted randomly from the multiple equivalent reaction amounts, or may be extracted based on a predetermined rule that takes into account the distribution of data related to the multiple equivalent reaction amounts, or may be extracted based on, for example, experimental design.

[0352] Next, the calculation processing unit 9 generates a fifth trained model (S231). The fifth trained model is an example of a trained model. The calculation processing unit 9 generates a fifth trained model for each element corresponding to multiple analysis meshes of the target workpiece model WM. In other words, up to the same number of fifth trained models as the number of analysis meshes of the target workpiece model WM are generated. The fifth trained model predicts the equivalent reaction amount of the polymer portion 12 from the set conditions of the crosslinking reaction of the polymer portion 12.

[0353] Next, the calculation processing unit 9 generates a sixth trained model (S232). The sixth trained model is an example of a trained model. The calculation processing unit 9 predicts the target workpiece model elastic modulus Ks from the set conditions and the equivalent reaction amounts of each element of the multiple analysis meshes of the target workpiece model WM predicted based on the multiple sixth trained models. For example, the predicted multiple equivalent reaction amounts may be converted into multiple final reaction amounts, and the target workpiece model elastic modulus Ks may be predicted based on the converted multiple final reaction amounts.

[0354] Next, the calculation processing unit 9 executes an equivalent reaction amount prediction process (S233). As a result, the prediction processing unit predicts an equivalent reaction amount for each element corresponding to multiple analysis meshes of the target workpiece model WM based on the set conditions and the fifth trained model. However, the order of the process of generating the sixth trained model (S232) and the equivalent reaction amount prediction process (S233) is arbitrary.

[0355] Using the above equation (17), the graph shown in Figure 11(b), in which the vertical axis represents the reaction rate and the horizontal axis represents the reaction time, can be converted into a graph in which the vertical axis represents the reaction rate and the horizontal axis represents the equivalent reaction amount. This allows the predicted equivalent reaction amount to be easily converted into the reaction rate. The converted reaction rate can be used to determine the presence of internal bubbles, as described below.

[0356] Next, the calculation processing unit 9 calculates the coefficient of determination R ２ is calculated (S234).

[0357] Next, the calculation processing unit 9 calculates the coefficient of determination R ２ It is determined whether (fifth) is smaller than the threshold value TH (S235).

[0358] Coefficient of determination R of the fifth trained model ２ If (5) is smaller than the threshold value TH (S235: Y), the simulation process (S238) is executed for all conditions. As a result, the target workpiece model elastic modulus Ks is calculated. This completes the prediction process (S20).

[0359] On the other hand, the fifth trained model R ２ is not smaller than the threshold value TH (S235: N), the calculation processing unit 9 calculates the coefficient of determination R ２ (Sixth) is determined to be smaller than the threshold value (S236).

[0360] Coefficient of determination R of the sixth trained model ２If (6th) is smaller than the threshold value TH (S236: Y), the simulation process (S238) is executed for all conditions. As a result, the target workpiece model elastic modulus Ks is calculated. This completes the prediction process (S20).

[0361] On the other hand, the coefficient of determination R ２ If (6th) is not smaller than the threshold value TH (S236: N), the calculation processing unit 9 predicts the equivalent reaction amount of each element corresponding to the multiple analysis meshes of the target workpiece model WM using the set conditions and the fifth trained model. Next, the calculation processing unit 9 predicts the target workpiece model elastic modulus Ks using the set conditions and the equivalent reaction amount of each element corresponding to the multiple analysis meshes of the target workpiece model WM using the sixth trained model (S237). This completes the prediction process (S20).

[0362] Figure 57 shows the correlation when the horizontal axis represents the final equivalent reaction amount of element number n of the polymer portion 12 obtained by simulation processing, and the vertical axis represents the final equivalent reaction amount of element number n predicted by machine learning. The closer the multiple points shown in Figure 55 are to the straight line extending diagonally upward to the right from the origin, the higher the correlation between the simulation results and the machine learning results, and the higher the prediction accuracy of the machine learning. According to this embodiment, the prediction accuracy of the equivalent reaction amount predicted by machine learning can be improved.

[0363] According to this embodiment, it is possible to improve the prediction accuracy of the equivalent reaction amount predicted by machine learning, thereby improving the prediction accuracy of the target workpiece model elastic modulus Ks.

[0364] Furthermore, by using the target workpiece model elastic modulus Ks predicted as described above to predict the setting conditions in the same manner as in embodiment 1, the prediction accuracy of the setting conditions can be improved overall, and the calculation time can be shortened.

[0365] (Embodiment 2-4) Next, embodiment 2-4 will be described with reference to Fig. 58 to Fig. 60. In this embodiment, the elastic modulus Ks of a target workpiece model is predicted using simulation results related to the equivalent reaction amount of a crosslinking reaction and machine learning. In the machine learning, a trained model that predicts multiple statistics for the equivalent reaction amount of a crosslinking reaction is generated, and the same number of equivalent reaction amounts of the crosslinking reaction as the statistics are used.

[0366] 58 shows a part of a flowchart of the prediction process of this embodiment. When the prediction process (S20) of this embodiment starts, the calculation processing unit 9 creates training data (S2400). The process of S2400 of this embodiment is the same as S2200 of FIG. 51, so a duplicated explanation will be omitted.

[0367] Next, the calculation processing unit 9 generates a fifth trained model (S2401), a seventh trained model (S2402), and an eighth trained model (S24203). The processing of S2401 is the same as S2231 in Figure 54, so a duplicated explanation will be omitted. However, the eighth trained model may be generated after S2407 (described later) and before S2409 (between S2407 and S2409).

[0368] The seventh trained model predicts the elastic modulus Ks of the target workpiece model based on the set conditions and at least one of the average value of the equivalent reaction amount of each element corresponding to the multiple analysis meshes of the target workpiece model WM, and the maximum value, minimum value, and standard deviation of the equivalent reaction amount of each element corresponding to the multiple analysis meshes of the target workpiece model WM.

[0369] The eighth trained model is an example of a trained model. The calculation processing unit 9 predicts at least one of the average value, maximum value, minimum value, and standard deviation for the equivalent reaction quantity of each element corresponding to the multiple analysis meshes of the target workpiece model WM. The number of eighth trained models is the total number of models that predict at least one of the average value, maximum value, minimum value, and standard deviation of the equivalent reaction quantity. Therefore, the number of eighth trained models is at most one. However, the statistical quantity to be predicted is not limited to the above, and any statistical quantity can be predicted.

[0370] Next, the calculation processing unit 9 predicts the equivalent reaction amount (S2404). As a result, the calculation processing unit 9 predicts the equivalent reaction amount for each element corresponding to the multiple analysis meshes of the target workpiece model WM based on the set conditions and the fifth trained model, the seventh trained model, and the eighth trained model. However, the equivalent reaction amount using the eighth trained model may be predicted after S2407 described below and before S2409 (between S2407 and S2409).

[0371] Next, the calculation processing unit 9 calculates the coefficient of determination R ２ (S2405), where the coefficient of determination R ２ The prediction may be made after S2407 and before S2409 (between S2407 and S2409), which will be described later.

[0372] Next, the calculation processing unit 9 calculates the coefficient of determination R ２ It is determined whether (fifth) is smaller than the threshold value TH (S2406).

[0373] Coefficient of determination R of the fifth trained model ２ If (fifth) is smaller than the threshold value TH (S2406: Y), a simulation process (S2408) is executed for all conditions. The contents of the simulation process (S2408) are the same as the simulation process (S10) of FIG. 48, so a duplicated explanation will be omitted. As a result, the target workpiece model elastic modulus Ks is calculated. This completes the prediction process (S20) (see FIG. 59).

[0374] On the other hand, the first trained model R ２ is not smaller than the threshold value TH (S2206: N), the calculation processing unit 9 calculates the coefficient of determination R ２ It is determined whether (third) is smaller than the threshold value (S2207).

[0375] Coefficient of determination R of the fifth trained model ２If (third) is smaller than the threshold value TH (S2407: Y), the simulation process (S2408) is executed for all conditions. As a result, the target workpiece model elastic modulus Ks is calculated. This completes the prediction process (S20) (see FIG. 59).

[0376] On the other hand, the coefficient of determination R ２ If (fifth) is not smaller than the threshold value TH (S2407: N), as shown in FIG. 57, the calculation processing unit 9 calculates the coefficient of determination R ２ It is determined whether (8th) is smaller than the threshold value TH (S2409).

[0377] Coefficient of determination R of the 8th trained model ２ If (8th) is not smaller than the threshold value TH (S2409: N), the calculation processing unit 9 selects the seventh trained model and the eighth trained model (S2410).

[0378] Next, the calculation processing unit 9 uses the eighth learned model to predict at least one of the average value, maximum value, minimum value, and standard deviation for the equivalent reaction amount of each element corresponding to multiple analysis meshes of the target work model WM (S2411).

[0379] Next, the calculation processing unit 9 executes a target workpiece model elastic modulus prediction process (S2412). As a result, the calculation processing unit 9 predicts the target workpiece model elastic modulus Ks based on the seventh trained model, the setting conditions, and at least one of the average value, maximum value, minimum value, and standard deviation for the final reaction rate of each element corresponding to the multiple analysis meshes of the target workpiece model WM. However, the statistical quantity is not limited to the above, and any statistical quantity can be predicted. This completes the prediction process (S20).

[0380] On the other hand, the coefficient of determination R of the 8th trained model ２ If (8th) is smaller than the threshold value TH (S2409: Y), the calculation processing unit 9 selects the fifth trained model (S2413), as shown in FIG. 58.

[0381] Next, the calculation processing unit 9 calculates the equivalent reaction amount for all elements corresponding to the plurality of analysis meshes of the target workpiece model WM (S2414).

[0382] Next, the calculation processing unit 9 calculates statistics for all elements corresponding to the plurality of analysis meshes of the target workpiece model WM (S2415).

[0383] Next, the calculation processing unit 9 calculates the coefficient of determination R based on the predicted value of the statistics of the final reaction rate calculated in S2415 and the analytical value. ２ (Statistics) is calculated (S2416).

[0384] Next, the calculation processing unit 9 calculates the coefficient of determination R ２ It is determined whether (statistics) is smaller than the threshold value TH (S2417).

[0385] Coefficient of determination R ２ If (statistic) is smaller than the threshold value TH (S2417: Y), the simulation process (S2420) is executed for all conditions. This calculates the target workpiece model elastic modulus Ks. This completes the prediction process (S20) (see FIG. 57).

[0386] On the other hand, the coefficient of determination R ２ If (statistics) is not smaller than the threshold value TH (S2417: N), the calculation processing unit 9 selects the seventh trained model (S2418).

[0387] Next, the calculation processing unit 9 predicts the target workpiece model elastic modulus Ks using the seventh trained model and the statistics calculated in S2415 (S2419). This completes the prediction process (S20) (see FIG. 59).

[0388] According to this embodiment, at least one eighth trained model is used, thereby shortening the calculation time for the target work model elastic modulus Ks.

[0389] (Modification of Embodiment 2-4) As in the modification of embodiment 2-2 described above, S2401 and S2406 shown in FIG. 58 may be omitted, and S2401, S2404, S2405, and S2406 may be executed in FIG.

[0390] According to this embodiment, the fifth trained model is adopted when the accuracy of the seventh trained model and the eighth trained model is not sufficiently high. The fifth trained model generates multiple predictive models, so it takes time to generate the models. Therefore, according to this embodiment, when the accuracy of the seventh trained model and the eighth trained model is sufficiently high, the generation of the fifth trained model can be omitted, thereby speeding up processing overall.

[0391] (Embodiment 3-1) Next, embodiment 3-1 will be described with reference to Fig. 61. In this embodiment, the occurrence of internal bubbles in a target workpiece model WM is predicted using machine learning and simulation results related to the reaction rate of a crosslinking reaction. In the machine learning, a maximum number of trained models equal to the number of elements in the target workpiece model WM is generated, and a maximum number of reaction rates of the crosslinking reaction equal to the number of elements is used.

[0392] 61 shows a flowchart of the prediction process according to this embodiment. When the prediction process (S20) starts, the calculation processing unit 9 generates training data (S250). The process of S250 in this embodiment is the same as S210 in FIG. 49, so a duplicated description will be omitted.

[0393] Next, the calculation processing unit 9 generates a ninth trained model (S251). The ninth trained model is an example of a trained model. The calculation processing unit 9 generates a ninth trained model for each element corresponding to multiple analysis meshes of the target workpiece model WM. In other words, up to the same number of ninth trained models as the number of analysis meshes of the target workpiece model WM are generated. The ninth trained model predicts the demolding reaction rate of the polymer portion 12 from the set conditions for the crosslinking reaction of the polymer portion 12.

[0394] Next, the prediction processing unit executes a demolding reaction rate prediction process (S252), which predicts the demolding reaction rate for each element corresponding to the multiple analysis meshes of the target workpiece model WM based on the set conditions and the ninth trained model.

[0395] Next, the calculation processing unit 9 calculates the coefficient of determination R based on the demolding reaction rate calculated in S252 and the analytical value. ２(9th) is calculated (S253).

[0396] Next, the calculation processing unit 9 calculates the coefficient of determination R ２ It is determined whether (9th) is smaller than the threshold value TH (S254).

[0397] Coefficient of determination R ２ If (9th) is smaller than the threshold value TH (S254: Y), the simulation process (S256) is executed for all conditions. The occurrence of internal bubbles is predicted. This completes the prediction process (S20).

[0398] On the other hand, the coefficient of determination R ２ If (9th) is not smaller than the threshold value TH (S254: N), the calculation processing unit 9 estimates the occurrence of internal bubbles in the target workpiece model WM based on the demolding reaction rate calculated in S252 (S255, internal bubble estimation process). The content of S255 is the same as S4 in Fig. 4, so a duplicated explanation will be omitted. This completes the prediction process (S20).

[0399] According to this embodiment, it is possible to predict the occurrence of internal bubbles in the target workpiece model WM by machine learning, which reduces the number of analyses required to predict the occurrence of internal bubbles in the target workpiece model WM and shortens the calculation time.

[0400] (Embodiment 3-2) Next, embodiment 3-2 will be described with reference to FIG. 62 . In this embodiment, the occurrence of internal bubbles in a target workpiece model WM is predicted using machine learning and simulation results related to the reaction rate of the crosslinking reaction. In the machine learning, a trained model that predicts multiple statistics for the reaction rate of the crosslinking reaction is generated, and the same number of reaction rates of the crosslinking reaction as the statistics are used. However, the minimum value of the reaction amount at demolding may also be used as one of the multiple statistics.

[0401] 62 shows a flowchart of the prediction process of this embodiment. When the prediction process (S20) of this embodiment starts, the calculation processing unit 9 creates training data (S260). The process of S260 of this embodiment is the same as S210 of FIG. 49, so a duplicated explanation will be omitted.

[0402] Next, the calculation processing unit 9 generates a tenth trained model (S261). The tenth trained model is an example of a trained model. The calculation processing unit 9 predicts at least one of the average value, maximum value, minimum value, and standard deviation for the demolding reaction rate of each element corresponding to multiple analysis meshes of the target workpiece model WM. The number of tenth trained models is the total of those that predict at least one of the average value, maximum value, minimum value, and standard deviation of the demolding reaction rate. Therefore, the number of tenth trained models is at most one. However, the statistical quantity is not limited to the above, and any statistical quantity can be predicted.

[0403] Next, the prediction processing unit executes a demolding reaction rate prediction process (S262), which predicts the demolding reaction rate for each element corresponding to the multiple analysis meshes of the target workpiece model WM based on the set conditions and the tenth trained model.

[0404] Next, the calculation processing unit 9 calculates the coefficient of determination R based on the demolding reaction rate calculated in S262 and the analytical value. ２ (10th) is calculated (S253).

[0405] Next, the calculation processing unit 9 calculates the coefficient of determination R ２ It is determined whether (10th) is smaller than the threshold value TH (S264).

[0406] Coefficient of determination R ２ If the tenth value is smaller than the threshold value TH (S264: Y), the simulation process (S266) is executed for all conditions. This predicts the occurrence of internal bubbles. This completes the prediction process (S20).

[0407] On the other hand, the coefficient of determination R ２ If (10th) is not smaller than the threshold value TH (S264: N), the calculation processing unit 9 estimates the occurrence of internal bubbles in the target workpiece model WM based on the demolding reaction rate calculated in S262 (S265, internal bubble estimation process). The content of S265 is the same as S4 in Fig. 4, so a duplicated explanation will be omitted. This completes the prediction process (S20).

[0408] According to this embodiment, a minimum of two and a maximum of four tenth trained models are used, thereby shortening the calculation time required to predict the occurrence of internal bubbles in the target workpiece model WM.

[0409] (Embodiment 3-3) Next, embodiment 3-3 will be described with reference to Fig. 63. In this embodiment, the occurrence of internal bubbles in a target workpiece model WM is predicted using machine learning and simulation results related to the equivalent reaction amount of a crosslinking reaction. In the machine learning, the same number of trained models as the number of elements in the target workpiece model WM are generated, and a maximum number of equivalent reaction amounts equal to the number of elements is used.

[0410] When the prediction process (S20) according to this embodiment starts, the calculation processing unit 9 generates training data (S270). The process of S270 according to this embodiment is the same as S2200 in Fig. 51, so a duplicated description will be omitted.

[0411] Next, the calculation processing unit 9 generates an eleventh trained model (S271). The eleventh trained model is an example of a trained model. The calculation processing unit 9 generates an eleventh trained model for each element corresponding to the multiple analysis meshes of the target workpiece model WM. In other words, the same number of eleventh trained models as the number of analysis meshes of the target workpiece model WM are generated. The eleventh trained model predicts the equivalent reaction amount of the polymer portion 12 at the time of demolding from the set conditions of the crosslinking reaction of the polymer portion 12.

[0412] Next, the calculation processing unit 9 executes an equivalent reaction amount prediction process (S272). As a result, the prediction processing unit predicts the equivalent reaction amount at demolding for each element corresponding to the multiple analysis meshes of the target workpiece model WM based on the set conditions and the eleventh trained model.

[0413] Next, the calculation processing unit 9 calculates the coefficient of determination R based on the equivalent reaction amount at the time of demolding calculated in S272 and the analytical value. ２ (11th) is calculated (S273).

[0414] Next, the calculation processing unit 9 calculates the coefficient of determination R ２ It is determined whether (11th) is smaller than the threshold value TH (S274).

[0415] Coefficient of determination R２ If (11th) is smaller than the threshold value TH (S274: Y), the simulation process (S276) is executed for all conditions. As a result, the target workpiece model elastic modulus Ks is calculated. This completes the prediction process (S20).

[0416] On the other hand, the coefficient of determination R ２ If (11) is not smaller than the threshold value TH (S274: N), the calculation processing unit 9 estimates the occurrence of internal bubbles in the target workpiece model WM based on the equivalent reaction amount at demolding calculated in S272 (S275, internal bubble estimation process). The content of S275 is the same as S4 in Figure 4, so duplicated explanations will be omitted. This completes the prediction process (S20).

[0417] According to this embodiment, it is possible to improve the prediction accuracy of the equivalent reaction amount predicted by machine learning, which in turn improves the prediction accuracy of the generation of internal bubbles in the target workpiece model WM.

[0418] (Embodiment 3-4) Next, embodiment 3-4 will be described with reference to FIG. 64. In this embodiment, the occurrence of internal bubbles in a target workpiece model WM is predicted using machine learning and simulation results related to the equivalent reaction amount of a crosslinking reaction. In the machine learning, a trained model is generated that predicts multiple statistics for the equivalent reaction amount of a crosslinking reaction at demolding, and the same number of equivalent reaction amounts of the crosslinking reaction as the statistics are used. However, the minimum value of the reaction amount at demolding may also be used as one of the multiple statistics.

[0419] 64 shows a flowchart of the prediction process of this embodiment. When the prediction process (S20) of this embodiment starts, the calculation processing unit 9 creates training data (S280). The process of S280 of this embodiment is the same as S2200 of FIG. 51, so a duplicated explanation will be omitted.

[0420] Next, the calculation processing unit 9 generates a twelfth trained model (S281). The twelfth trained model is an example of a trained model. The calculation processing unit 9 predicts at least one of the average value, maximum value, minimum value, and standard deviation for the equivalent reaction quantity of each element corresponding to the multiple analysis meshes of the target workpiece model WM. The number of twelfth trained models is the total of those that predict at least one of the average value, maximum value, minimum value, and standard deviation of the equivalent reaction quantity. Therefore, the number of twelfth trained models is at most one. However, the statistical quantity is not limited to the above, and any statistical quantity can be predicted.

[0421] Next, the calculation processing unit 9 executes an equivalent reaction amount prediction process (S282). As a result, the prediction processing unit predicts the equivalent reaction amount at demolding for each element corresponding to the multiple analysis meshes of the target workpiece model WM based on the set conditions and the twelfth trained model.

[0422] Next, the calculation processing unit 9 calculates the coefficient of determination R based on the equivalent reaction amount at demolding calculated in S282 and the analytical value. ２ (12th) is calculated (S283).

[0423] Next, the calculation processing unit 9 calculates the coefficient of determination R ２ It is determined whether (12th) is smaller than the threshold value TH (S284).

[0424] Coefficient of determination R ２ If (12th) is smaller than the threshold value TH (S284: Y), the simulation process (S286) is executed for all conditions. As a result, the target workpiece model elastic modulus Ks is calculated. This completes the prediction process (S20).

[0425] On the other hand, the coefficient of determination R ２ If (12th) is not smaller than the threshold value TH (S284: N), the calculation processing unit 9 estimates the generation of internal bubbles in the target workpiece model WM based on the demolding equivalent reaction amount calculated in S282 (S285, internal bubble estimation process). The content of S285 is the same as S4 in Figure 4, so duplicated explanations will be omitted. This completes the prediction process (S20).

[0426] According to this embodiment, a minimum of two and a maximum of four fourth trained models are used, thereby reducing the time required to calculate the occurrence of internal bubbles in the target workpiece model WM.

[0427] The present disclosure is not limited to the above-described embodiments, and can be applied to various embodiments without departing from the spirit of the present disclosure.

[0428] In this embodiment, the internal bubble estimation unit 6 is configured to include a torque calculation unit 61. However, this is not limited thereto. Alternatively, the internal bubble estimation unit 6 may include a reaction rate acquisition unit 60 that acquires the reaction rate (vulcanization degree) of the crosslinking reaction of the polymer portion 12 of the target workpiece model WM, and an estimation unit 62 that estimates the generation of bubbles in the polymer portion 12 of the target workpiece model WM based on the reaction rate at the time of demolding. In this case, before performing reaction analysis of the vulcanization reaction, the vulcanization degree at a threshold torque is calculated from the reference reaction curve RC, and this vulcanization degree can be used as the blow point vulcanization degree. The blow point vulcanization degree may also be measured directly using a known measuring device such as a blow point analyzer or blow point tester. Then, reaction analysis of the vulcanization reaction is performed, and the calculation result (vulcanization degree) of the reaction rate acquisition unit 60 is compared with the blow point vulcanization degree to estimate the generation of internal bubbles. For example, if the vulcanization degree does not exceed the blow point vulcanization degree, it is estimated that internal bubbles 80 have been generated in the polymer portion 12 of the target workpiece model WM.

[0429] For example, the crosslinking reaction simulator can take the following forms [1] to

[36] .

[0430] [1] A crosslinking reaction simulation device comprising: a storage unit for storing data used in a simulation; a heat transfer analysis unit for performing a heat transfer analysis during a crosslinking reaction of a polymer part of a target workpiece model; and a crosslinking reaction analysis unit for analyzing a reaction rate of the crosslinking reaction of the polymer part using the results of the heat transfer analysis, wherein the storage unit stores: the target workpiece model having the polymer part comprising a raw material polymer; and an equivalent reaction amount calculation model that defines the ratio of the reaction amount of the crosslinking reaction at a target reaction temperature and a target reaction time to the reaction amount of the crosslinking reaction at a reference reaction temperature as an equivalent reaction amount, and that includes a slope coefficient that represents the slope of an Arrhenius plot; and the slope coefficient that is set according to the progress of the crosslinking reaction of the polymer part, wherein the crosslinking reaction analysis unit comprises: a temperature acquisition unit that acquires the temperature at each time for each element of the polymer part of the target workpiece model in the crosslinking reaction as a result of the heat transfer analysis; a reaction rate calculation processing unit that calculates the equivalent reaction amount of the polymer portion for each time based on the acquired temperature for each element of the polymer portion in the crosslinking reaction for each time, the equivalent reaction amount calculation model, and the slope coefficient according to the progress of the crosslinking reaction at a target time, and calculates the reaction rate of the crosslinking reaction of the polymer portion based on the calculated equivalent reaction amount.

[0431] [2] The crosslinking reaction simulation device according to [1], wherein the polymer part is a rubber part configured to exhibit vibration-damping performance, and the target workpiece model is a model of a vibration-damping rubber device.

[0432] [3] The crosslinking reaction simulator according to [1] or [2], wherein the slope coefficient is set to a different value in a reaction progress period until the reaction rate of the crosslinking reaction reaches a peak and in a reaction return period after the reaction rate has exceeded the peak.

[0433] [4] The crosslinking reaction simulator according to [3], wherein the entire period including the reaction progress period and the reaction return period is divided into equal intervals, and the slope coefficient is set to a value corresponding to each division.

[0434] [5] The crosslinking reaction simulator according to [3], wherein the slope coefficient is set to a different value for each of an initial reaction progress period when the crosslinking reaction starts to proceed, a reaction acceleration period when the reaction progress rate becomes faster than in the initial reaction progress period, and a later reaction progress period when the reaction progress rate becomes slower than in the reaction acceleration period, in the crosslinking reaction until the reaction rate reaches a peak.

[0435] [6] The crosslinking reaction simulator according to [3], wherein the gradient coefficient is set to a smaller value in the reaction reversion period than in the reaction progress period.

[0436] [7] The crosslinking reaction simulation device according to [3], wherein the slope coefficient is set to a smaller value in the reaction return period than in the reaction progress period, the reaction progress period is divided into an initial reaction progress period in which the crosslinking reaction starts to progress, a reaction promotion period in which the reaction progress rate becomes faster than in the initial reaction progress period, and a later reaction progress period in which the reaction progress rate becomes slower than in the reaction promotion period, and the reaction promotion period is set to a larger value than in the initial reaction progress period and the later reaction progress period.

[0437] [8] The crosslinking reaction simulation device according to [1] or [2], which determines the slope coefficient to be used at a target time based on the reaction rate at the previous time, calculates an equivalent reaction amount increase in a small time period at the target time based on the temperature at the target time, an equivalent reaction amount calculation model, and the slope coefficient, calculates an equivalent reaction amount integrated value from the start of the crosslinking reaction to the target time based on the equivalent reaction amount increase, and calculates the reaction rate at the target time based on the equivalent reaction amount integrated value from the start of the crosslinking reaction to the target time.

[0438] [9] The crosslinking reaction simulation device according to [8], wherein a first relationship data map defining a correspondence relationship between the reaction rate of the crosslinking reaction of the polymer portion and the slope coefficient is stored, and the slope coefficient is determined using the reaction rate at a previous time and the first relationship data map.

[0439]

[10] The crosslinking reaction simulation device according to [8], wherein a first function defining a correspondence relationship between the reaction rate of the crosslinking reaction of the polymer portion and the slope coefficient is stored, and the slope coefficient is determined using the reaction rate at a previous time and the first function.

[0440]

[11] The crosslinking reaction simulator according to

[10] , wherein the first function is set to a different function according to a plurality of reaction rate sections set according to the progress of the crosslinking reaction of the polymer section.

[0441]

[12] The crosslinking reaction simulator according to [8], further comprising: storing a second relationship data map defining a correspondence relationship between the equivalent reaction amount integrated value and the reaction rate at a target time; and determining the reaction rate at the target time using the equivalent reaction amount integrated value at the target time and the second relationship data map.

[0442]

[13] The crosslinking reaction simulator according to [8], which stores a second function that defines a correspondence relationship between the equivalent reaction amount integrated value and the reaction amount at a target time, and determines a reaction rate at a target time using the equivalent reaction amount integrated value at the target time and the second function.

[0443]

[14] The crosslinking reaction simulation device described in

[12] , wherein the second relationship data map is set from a reference reaction curve based on the relationship between the crosslinking reaction time and the torque generated in the test target polymer material at a reference reaction temperature measured by a crosslinking reaction characteristic tester using a test target polymer material corresponding to the polymer part of the target workpiece model.

[0444]

[15] The crosslinking reaction simulation device described in

[13] , wherein the second function is set from a reference reaction curve based on the relationship between the crosslinking reaction time and the torque generated in the test target polymer material at a reference reaction temperature measured by a crosslinking reaction characteristic tester using a test target polymer material corresponding to the polymer part of the target workpiece model.

[0445]

[16] The crosslinking reaction simulator according to

[15] , wherein the second function is set to a different function according to a plurality of reaction rate ranges set according to the progress of the crosslinking reaction of the polymer portion.

[0446]

[17] The crosslinking reaction simulator according to [1] or [2], wherein the reaction rate of the crosslinking reaction is defined as a value obtained by adding a degree of decrease in the reaction rate from the peak to 100% in a reaction progress period until the reaction rate of the crosslinking reaction reaches a peak, with the degree of increase in the reaction rate of the crosslinking reaction being 0% at the start of the crosslinking reaction, to 100% in a reaction return period when the reaction rate has exceeded the peak.

[0447]

[18] The crosslinking reaction simulation device according to [1] or [2], further comprising a display unit that displays a value obtained from the reaction rate according to the target reaction time based on the analysis result of the crosslinking reaction analysis unit.

[0448]

[19] The crosslinking reaction simulation device according to [1] or [2], further comprising a structural analysis unit that performs structural analysis using the reaction rate of the crosslinking reaction of the polymer part analyzed by the crosslinking reaction analysis unit, wherein the structural analysis unit comprises: a reaction rate acquisition unit that acquires the reaction rate of each element of the polymer part calculated by the reaction rate calculation processing unit of the crosslinking reaction analysis unit; an elastic modulus assignment unit that assigns an elastic modulus in the polymer part according to the acquired reaction rate; and a characteristic acquisition unit that acquires the characteristics of the target workpiece model by performing structural analysis with the elastic modulus assigned to the polymer part.

[0449]

[20] The crosslinking reaction simulation device according to

[19] , further comprising a display unit that displays the characteristics according to the temperature of the mold model used in the crosslinking reaction of the polymer part and the in-mold reaction time from the start of the crosslinking reaction to the release of the mold model, based on the results of the structural analysis.

[0450]

[21] The crosslinking reaction simulation device according to

[19] , wherein the characteristic of the target workpiece model is the elastic modulus of the target workpiece model.

[0451]

[22] The crosslinking reaction simulation device according to [1] or [2], wherein the storage unit further stores: a mold model; the target workpiece model having the polymer portion further comprising carbon black; and thermal diffusivity characteristics representing the relationship between the mass ratio of the carbon black to the raw material polymer and the thermal diffusivity of the polymer portion; and the heat transfer analysis unit comprises: a condition input unit for inputting the mass ratio of the carbon black to the raw material polymer in the target workpiece model and the temperature conditions of the mold model; a polymer thermal diffusivity determination unit for determining a polymer thermal diffusivity, which is the thermal diffusivity of the polymer portion of the target workpiece model, based on the mass ratio input by the condition input unit and the thermal diffusivity characteristics stored in the storage unit; and an analysis unit for performing heat transfer analysis using the polymer thermal diffusivity determined by the polymer thermal diffusivity determination unit and the temperature conditions stored in the storage unit, with the target workpiece model placed on the mold model.

[0452]

[23] The crosslinking reaction simulator according to

[22] , wherein the thermal diffusivity characteristics are independent of the type of the raw material polymer.

[0453]

[24] The crosslinking reaction simulator according to

[22] , wherein the thermal diffusivity characteristics are a relationship established from actual measured values ​​obtained by experiments using a plurality of types of raw material polymers at different mass ratios of carbon black.

[0454]

[25] The crosslinking reaction simulator according to

[22] , wherein the thermal diffusivity characteristic has a linear relationship with the mass ratio of the carbon black.

[0455]

[26] The crosslinking reaction simulation device described in

[22] , wherein the target workpiece model comprises the polymer part and a joining member joined to the polymer part, the memory unit further stores a contact heat transfer coefficient, which is the heat transfer coefficient between the mold model and the joining member, in a state where the target workpiece model is placed on the mold model, and the analysis unit performs the heat transfer analysis using the polymer thermal diffusivity determined by the polymer thermal diffusivity determination unit, the temperature conditions stored in the memory unit, and the contact heat transfer coefficient, in a state where the target workpiece model is placed on the mold model.

[0456]

[27] The crosslinking reaction simulation device described in

[22] , wherein the condition input unit is configured to input outside air conditions including the air temperature surrounding the target work model demolded from the mold model, the memory unit further stores an air heat transfer coefficient which is the heat transfer coefficient of air, and the analysis unit performs the heat transfer analysis using the polymer thermal diffusivity determined by the polymer thermal diffusivity determination unit, the air heat transfer coefficient stored in the memory unit, and the outside air conditions in a state where the target work model is demolded from the mold model after the polymer part of the target work model has undergone a crosslinking reaction within the mold model.

[0457]

[28] The crosslinking reaction simulation device according to [1] or [2], wherein the memory unit further stores a molding die model, and further comprises an internal bubble estimation unit that is applied to a crosslinking reaction process in which the polymer part of the target workpiece model is subjected to a crosslinking reaction within the molding die model and then the molding die model is demolded, and that estimates the generation of bubbles inside the polymer part of the target workpiece model due to the demolding of the molding die model, and the internal bubble estimation unit comprises: a reaction rate acquisition unit that acquires a reaction rate of the crosslinking reaction of the polymer part of the target workpiece model; and an estimation unit that estimates the generation of bubbles inside the polymer part of the target workpiece model based on the reaction rate at the time of demolding.

[0458]

[29] The memory unit further stores a crosslinking reaction curve that defines the relationship between the elapsed time from the start of the crosslinking reaction and a torque that is a value corresponding to the progress of the crosslinking reaction in the polymer part of the target workpiece model and that can be measured by a crosslinking reaction characteristic tester using a test target polymer material that corresponds to the polymer part, and the internal bubble estimation unit further includes a torque calculation unit that calculates the torque corresponding to the acquired reaction rate based on the reaction rate acquired by the reaction rate acquisition unit and the crosslinking reaction curve stored in the memory unit, and the estimation unit estimates the generation of bubbles inside the polymer part of the target workpiece model based on the torque at the time of demolding. A crosslinking reaction simulation device as described in

[28] .

[0459]

[30] The crosslinking reaction simulation device described in

[29] , wherein the estimation unit estimates that bubbles have occurred inside the polymer part of the target workpiece model if the torque at the time of demolding does not exceed a preset threshold value.

[0460]

[31] The crosslinking reaction simulator according to

[30] , wherein the threshold value is set to a uniform value for a plurality of types of raw material polymers.

[0461]

[32] The crosslinking reaction simulator according to

[30] , wherein the threshold value is the torque corresponding to a reaction rate greater than the reaction rate of the crosslinking reaction at the blow point of the polymer material.

[0462]

[33] The crosslinking reaction simulation device described in

[29] , wherein the reaction rate acquisition unit acquires the reaction rate for each part of the polymer part, the torque calculation unit calculates the torque for each part of the polymer part, and the estimation unit estimates the generation of bubbles for each part of the polymer part.

[0463]

[34] The crosslinking reaction simulation device according to

[29] , further comprising a display unit that displays whether or not bubbles have occurred, based on the estimation result by the internal bubble estimation unit, in accordance with the temperature of the molding die model and the in-mold reaction time from the start of the crosslinking reaction to the demolding of the molding die model.

[0464]

[35] The crosslinking reaction simulation device described in

[34] , wherein the display unit displays a value obtained from the reaction rate according to the temperature of the mold model and the in-mold reaction time from the start of the crosslinking reaction to the demolding of the mold model, based on the analysis results by the crosslinking reaction analysis unit, along with the presence or absence of bubbles.

[0465]

[36] The crosslinking reaction simulation device according to

[34] further comprises a structural analysis unit that acquires the characteristics of the target workpiece model by performing structural analysis using the reaction rate analyzed by the crosslinking reaction analysis unit, and the display unit displays the characteristics according to the temperature of the mold model and the in-mold reaction time from the start of the crosslinking reaction to the demolding of the mold model, along with the presence or absence of bubbles, based on the results of the structural analysis.

Claims

1. A crosslinking reaction simulation device comprising: a memory unit for storing data used in a simulation; and a setting condition analysis unit for analyzing setting conditions related to a crosslinking reaction of a polymer part of a target work model, wherein the memory unit stores: a polymer characteristic value which is a characteristic value related to a raw material polymer constituting the polymer part of the target work model; a target work model elastic modulus which is the elastic modulus of the polymer part of the target work model; relationship data correlating the setting conditions in the crosslinking reaction of the polymer part of the target work model having the target work model elastic modulus; and a reaction time in the crosslinking reaction of the polymer part of the target work model having the target work model elastic modulus, and the setting condition analysis unit comprises: a polymer characteristic value acquisition unit for acquiring the polymer characteristic value; a target work model elastic modulus acquisition unit for acquiring the target work model elastic modulus; a relationship data acquisition unit for acquiring the relationship data; and a correction unit for correcting the relationship between the target work model elastic modulus, the setting conditions, and the reaction time of the relationship data based on the polymer characteristic value.

2. A crosslinking reaction simulation device according to claim 1, wherein the polymer characteristic values ​​include a characteristic value relating to the elasticity of the raw polymer, and further include at least one of the scorch time and Mooney viscosity of the raw polymer.

3. The crosslinking reaction simulation device of claim 1, wherein the memory unit further stores a molding model, the target work model comprises the polymer part and a joining member joined to the polymer part, and the correction unit corrects the relationship between the target work model elastic modulus, the setting conditions, the reaction time, the temperature at the time of injection of the polymer part into the mold, the temperature of the joining member, and at least one of the conditions for an additional crosslinking reaction on the polymer part, of the relationship data based on the polymer characteristic value.

4. The apparatus further comprises a heat transfer analysis unit which performs a heat transfer analysis during a crosslinking reaction of a polymer part of a target workpiece model, a crosslinking reaction analysis unit which uses the results of the heat transfer analysis to analyze a reaction rate of the crosslinking reaction of the polymer part, and a structural analysis unit which performs a structural analysis using the reaction rate of the crosslinking reaction of the polymer part analyzed by the crosslinking reaction analysis unit, and the memory unit further stores: the target workpiece model having the polymer part comprising a raw material polymer; an equivalent reaction amount calculation model which defines the ratio of the reaction amount of the crosslinking reaction at a target reaction time at a target reaction temperature to the reaction amount of the crosslinking reaction at a reference reaction time at a reference reaction temperature as an equivalent reaction amount and is defined including a slope coefficient which represents the slope of an Arrhenius plot; and the slope coefficient which is set according to the progress of the crosslinking reaction of the polymer part, and the crosslinking reaction analysis unit comprises: a temperature acquisition unit which acquires the temperature at each time for each element of the polymer part of the target workpiece model in the crosslinking reaction as a result of the heat transfer analysis; 2. The crosslinking reaction simulation device according to claim 1, further comprising: a reaction rate calculation processing unit that calculates the equivalent reaction amount of the polymer part for each time based on the acquired temperature for each element of the polymer part in the crosslinking reaction for each time, the equivalent reaction amount calculation model, and the slope coefficient corresponding to the progress of the crosslinking reaction at a target time, and calculates a reaction rate of the crosslinking reaction of the polymer part based on the calculated equivalent reaction amount, wherein the structural analysis unit comprises: a reaction rate acquisition unit that acquires the reaction rate of each element of the polymer part calculated by the reaction rate calculation processing unit of the crosslinking reaction analysis unit; an elastic modulus assignment unit that assigns an elastic modulus in the polymer part according to the acquired reaction rate; and a target workpiece model elastic modulus acquisition unit that acquires the target workpiece model elastic modulus by performing structural analysis with the elastic modulus assigned to the polymer part.

5. The apparatus further comprises: a heat transfer analysis unit that performs a heat transfer analysis during a crosslinking reaction of a polymer part of a target workpiece model; a crosslinking reaction analysis unit that performs an analysis of a reaction rate of the crosslinking reaction of the polymer part using a result of the heat transfer analysis; a structure analysis unit that performs a structure analysis using the reaction rate of the crosslinking reaction of the polymer part analyzed by the crosslinking reaction analysis unit; and an arithmetic processing unit that performs machine learning; and the storage unit further stores: the target workpiece model having the polymer part containing a raw material polymer; an equivalent reaction amount calculation model that defines the ratio of the reaction amount of the crosslinking reaction at a target reaction time at a target reaction temperature to the reaction amount of the crosslinking reaction at a reference reaction time at a reference reaction temperature as an equivalent reaction amount and is defined to include a slope coefficient that represents the slope of an Arrhenius plot; the slope coefficient; training data selected from data related to the reaction rate or the equivalent reaction amount; and a trained model generated by performing machine learning using the training data; and the crosslinking reaction analysis unit further comprises: a temperature acquisition unit that acquires the temperature at each time for each element of the polymer part of the target workpiece model in the crosslinking reaction as a result of the heat transfer analysis; 2. The crosslinking reaction simulation device according to claim 1, further comprising: a reaction rate calculation processing unit that calculates the equivalent reaction amount of the polymer part for each time based on the acquired temperature for each element of the polymer part in the crosslinking reaction, the equivalent reaction amount calculation model, and the slope coefficient at a target time, and calculates a reaction rate of the crosslinking reaction of the polymer part based on the calculated equivalent reaction amount; wherein the structural analysis unit comprises: a reaction rate acquisition unit that acquires the reaction rate of each element of the polymer part calculated by the reaction rate calculation processing unit of the crosslinking reaction analysis unit; an elastic modulus assignment unit that assigns an elastic modulus in accordance with the acquired reaction rate in the polymer part; and a target workpiece model elastic modulus acquisition unit that acquires the target workpiece model elastic modulus by performing structural analysis with the elastic modulus assigned to the polymer part; and wherein the arithmetic processing unit generates the trained model by performing the machine learning using the training data, and predicts a target variable using the trained model.

6. The crosslinking reaction simulation device according to claim 5, wherein the gradient coefficient is set according to the progress of the crosslinking reaction of the polymer portion.

7. A crosslinking reaction simulation device according to claim 5 or 6, wherein the calculation processing unit generates the trained model that predicts the reaction rate of the crosslinking reaction of the polymer portion.

8. A crosslinking reaction simulation device as described in claim 7, wherein the calculation processing unit processes a reaction rate prediction process that predicts the reaction rate of the crosslinking reaction of the polymer portion, and a target work model elastic modulus prediction process that predicts the target work model elastic modulus.

9. A crosslinking reaction simulation device as described in claim 8, wherein the reaction rate prediction process predicts the final reaction rate of the polymer portion of the target work model from the set conditions, and the target work model elastic modulus prediction process predicts the target work model elastic modulus based on the set conditions and the final reaction rate.

10. A crosslinking reaction simulation device as described in claim 5 or 6, wherein the calculation processing unit generates the trained model using the equivalent reaction amount of the polymer portion as a target variable.

11. A crosslinking reaction simulation device as described in claim 10, wherein the calculation processing unit processes an equivalent reaction amount prediction process that predicts the equivalent reaction amount of the polymer portion, and a target work model elastic modulus prediction process that predicts the target work model elastic modulus.

12. A crosslinking reaction simulation device as described in claim 11, wherein the equivalent reaction amount prediction process predicts the equivalent reaction amount from the set conditions, and the target work model elastic modulus prediction process predicts the target work model elastic modulus based on the set conditions and the equivalent reaction amount.

13. A crosslinking reaction simulation device as described in claim 12, wherein the equivalent reaction amount prediction process predicts statistics for the equivalent reaction amount in the elements of the polymer part from the set conditions, and the target work model elastic modulus prediction process predicts the target work model elastic modulus based on the set conditions and the statistics for the equivalent reaction amount in the elements.

14. A crosslinking reaction simulation device as described in claim 5 or 6, wherein the memory unit further stores a molding die model, and further comprises an internal bubble estimation unit that is applied to a crosslinking reaction process in which the polymer part of the target work model is crosslinked in the molding die model and then the molding die model is demolded, and estimates the generation of bubbles inside the polymer part of the target work model due to the demolding of the molding die model, the internal bubble estimation unit comprises: a reaction rate acquisition unit that acquires a reaction rate of the crosslinking reaction of the polymer part of the target work model; and an estimation unit that estimates the generation of bubbles inside the target work model based on the reaction rate at the time of demolding, and the calculation processing unit predicts the generation of bubbles inside the target work model using the learned model.

15. The crosslinking reaction simulation device according to claim 14, wherein the calculation processing unit generates the trained model that predicts the reaction rate at the time of demolding of the polymer part.

16. A crosslinking reaction simulation device as described in claim 15, wherein the calculation processing unit processes a reaction rate prediction process that predicts the reaction rate at the time of demolding of the polymer part, and an internal bubble estimation process that estimates the generation of the bubbles inside the target work model.

17. A crosslinking reaction simulation device as described in claim 16, wherein the reaction rate prediction process predicts the reaction rate of the polymer part of the target work model at the time of demolding from the set conditions, and the internal bubble estimation process estimates the occurrence of bubbles inside the target work model based on the set conditions and the reaction rate at the time of demolding.

18. The crosslinking reaction simulation device according to claim 14, wherein the calculation processing unit generates the trained model using the equivalent reaction amount of the polymer portion as a target variable.

19. A crosslinking reaction simulation device as described in claim 18, wherein the calculation processing unit processes an equivalent reaction amount prediction process that predicts the equivalent reaction amount of the polymer portion, and an internal bubble estimation process that estimates the generation of the bubbles inside the target work model.

20. A crosslinking reaction simulation device as described in claim 19, wherein the equivalent reaction amount prediction process predicts the equivalent reaction amount of the polymer portion of the target work model from the set conditions, and the internal bubble estimation process estimates the occurrence of the bubbles inside the target work model based on the set conditions and the equivalent reaction amount.

21. A crosslinking reaction simulation device as described in claim 1, wherein the polymer characteristic value is the elastic modulus of the test piece calculated by a specified method using a test piece formed from the raw polymer into a specified shape, applying a load to the test piece in a direction that compresses the test piece at a specified displacement rate using a specified jig.

22. The crosslinking reaction simulation device of claim 1, wherein the memory unit further stores a molding die model, and further comprises an internal bubble estimation unit that is applied to a crosslinking reaction process in which the polymer part of the target work model is crosslinked within the molding die model and then the molding die model is demolded, and that estimates the generation of bubbles inside the polymer part of the target work model due to the demolding of the molding die model, and the internal bubble estimation unit comprises: a reaction rate acquisition unit that acquires a reaction rate of the crosslinking reaction of the polymer part of the target work model; and an estimation unit that estimates the generation of bubbles inside the target work model based on the reaction rate at the time of demolding.

23. The crosslinking reaction simulation device according to claim 1, wherein the polymer portion is a rubber portion configured to exhibit vibration isolation performance, and the target workpiece model is a model of a vibration isolation rubber device.

24. An internal bubble estimation device that is applied to a crosslinking reaction process in which a polymer part of a target work model is subjected to a crosslinking reaction in a mold model and then the mold model is demolded, and that estimates the generation of bubbles inside the polymer part of the target work model due to the demolding of the mold model, the device comprising: a memory unit that stores a crosslinking reaction curve that defines the relationship between an elapsed time from the start of a crosslinking reaction and a torque that corresponds to the progress of the crosslinking reaction in the polymer part of the target work model and that can be measured by a crosslinking reaction characteristic tester using a test target polymer material corresponding to the polymer part; a reaction rate acquisition unit that acquires a reaction rate of the crosslinking reaction of the polymer part of the target work model; an estimation unit that estimates the generation of bubbles inside the polymer part of the target work model based on the reaction rate at the time of demolding or an equivalent reaction amount defined as the ratio of the reaction amount of the crosslinking reaction at a target reaction time at a target reaction temperature to the reaction amount of the crosslinking reaction at a reference reaction time at a reference reaction temperature; and a calculation processing unit that performs machine learning, the memory unit further comprising: training data for learning selected from data on the reaction rate or data on the equivalent reaction amount; and a trained model generated by performing machine learning using the training data, wherein the calculation processing unit generates the trained model by performing the machine learning using the training data, predicts the reaction rate or the equivalent reaction amount using the trained model, and estimates the occurrence of bubbles inside the polymer portion of the target work model based on the reaction rate or the equivalent reaction amount.

25. A crosslinking reaction simulation device comprising: a memory unit for storing data used in a simulation; a heat transfer analysis unit for performing a heat transfer analysis during a crosslinking reaction of a polymer part of a target workpiece model; a crosslinking reaction analysis unit for analyzing a reaction rate of the crosslinking reaction of the polymer part using a result of the heat transfer analysis; and a calculation processing unit for performing machine learning, wherein the memory unit stores: the target workpiece model having the polymer part containing a raw material polymer; an equivalent reaction amount calculation model that defines the ratio of the reaction amount of the crosslinking reaction at a target reaction time at a target reaction temperature to the reaction amount of the crosslinking reaction at a reference reaction time at a reference reaction temperature as an equivalent reaction amount and is defined including a slope coefficient representing the slope of an Arrhenius plot; the slope coefficient; training data selected from data related to the reaction rate or the equivalent reaction amount; and a trained model generated by performing machine learning using the training data; and the crosslinking reaction analysis unit stores: a temperature acquisition unit for acquiring the temperature at each time for each element of the polymer part of the target workpiece model in the crosslinking reaction as a result of the heat transfer analysis; a reaction rate calculation processing unit that calculates the equivalent reaction amount of the polymer part for each time based on the acquired temperature for each element of the polymer part in the crosslinking reaction for each time, the equivalent reaction amount calculation model, and the slope coefficient at a target time, and calculates a reaction rate of the crosslinking reaction of the polymer part based on the calculated equivalent reaction amount, wherein the calculation processing unit generates the trained model by performing the machine learning using the training data, and predicts a target variable using the trained model.

26. The crosslinking reaction simulation device according to claim 25, wherein the gradient coefficient is set according to the progress of the crosslinking reaction of the polymer portion.

Images