Control method for working machine, control device for working machine and control program

Abstract

To provide a control method for a working machine, the method enabling further reduction of a rate of the occurrence of a defective product.SOLUTION: A control method for a working machine includes: a material database acquisition step in which correlation between a mold clamping speed of a metallic mold and apparent viscosity of material is acquired beforehand as a material database; a master data acquisition step in which master data indicating change in the apparent viscosity of the material with respect to the mold clamping speed is acquired on the basis of the correlation of the material database; an apparent viscosity calculation step in which the apparent viscosity of the material during press molding is calculated on the basis of the mold clamping speed at a mold clamping position; a difference calculation step in which the apparent viscosity calculated in the apparent viscosity calculation step and the apparent viscosity acquired from the master data are compared and difference between the compared two apparent viscosity is calculated; and a speed control step in which when the difference equal to or larger than a predetermined level is calculated in the difference calculation step, mold clamping speed causing the difference to converge is set, and the working machine is given such an instruction that the working machine controls the mold at the set mold clamping speed.SELECTED DRAWING: Figure 9

Description

[Technical field]

[0001] The present invention relates to a control method for a processing machine, a control device for a processing machine, and a control program. [Background technology]

[0002] 2. Description of the Related Art Conventionally, for example, in production sites of automobile parts and the like, processing machines such as press molding machines suitable for processing metal materials, plastic materials, and the like have been used.

[0003] One of the challenges in operating such processing machines is reducing the rate of defective products. For example, in the press molding of carbon fiber reinforced plastics (CFRP), which will be described later, the rate of defective products is often as high as 30%, even when using press machines with servo motors that allow precise motion control. Therefore, improving the rate of defective products is a major challenge for reducing costs and improving productivity in many manufacturing industries that process materials.

[0004] In response to these problems, a technique for feedback control of the change in the load applied to the material by the press has been proposed (see, for example, Patent Document 1). Also, a technique for improving the efficiency of the technique related to Patent Document 1 has been proposed (see, for example, Patent Document 2). [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2013-237062 A [Patent Document 2] JP 2018-161657 A Summary of the Invention [Problem to be solved by the invention]

[0006] Carbon Fiber Taped Reinforced Thermoplastic (CTT) materials, which are made by randomly laminating tapes impregnated with thermoplastic resin, have high material fluidity and excellent moldability for complex shapes because they are discontinuous fibers. On the other hand, CTT materials have a large variation in mechanical properties because the material flows in units of tape pieces during press molding, which causes a large change in fiber orientation. In press molding using such materials, simply measuring the change in the load applied to the material by the press machine is not necessarily sufficient to control the material in response to the variation in mechanical properties that occurs during the processing process. In addition, with conventional feedback control, it is difficult to reflect control that corresponds to the variation in mechanical properties of the material that occurs during the processing process on the material. Therefore, there is room for further improvement in the issue of suppressing the rate of defective products in press molding using materials with large variations in mechanical properties.

[0007] An object of the present invention is to provide a control method for a processing machine, a control device for a processing machine, and a control program that can further reduce the rate of defective products. [Means for solving the problem]

[0008] A control method for a processing machine according to a first aspect of the present invention is a control method for a processing machine that processes a material placed in a mold by press molding, and includes a material database acquisition step of acquiring in advance a correlation between a material-specific mold clamping speed and the apparent viscosity of the material as a material database, a master data acquisition step of acquiring mold-specific master data indicating a change in apparent viscosity of the material with respect to the mold clamping speed based on the correlation between the mold clamping speed and the apparent viscosity of the material in the material database, an apparent viscosity calculation step of calculating the apparent viscosity of the material during press molding based on the mold clamping speed at the mold clamping position, a difference calculation step of comparing the apparent viscosity calculated in the apparent viscosity calculation step with the apparent viscosity acquired from the master data to calculate a difference, and a speed control step of setting a mold clamping speed at which the difference converges when a difference equal to or greater than a predetermined value is calculated in the difference calculation step, and instructing the processing machine to control the mold at the set mold clamping speed.

[0009] A control method for a processing machine according to a second invention is a control method for a processing machine that processes a material placed in a mold by press molding, and includes the following steps: a master data acquisition step of acquiring mold-specific master data indicating a change in apparent viscosity of the material with respect to the mold clamping speed based on a correlation between the material-specific mold clamping speed and the apparent viscosity of the material; an apparent viscosity calculation step of calculating the apparent viscosity of the material during press molding based on the mold clamping speed at the mold clamping position; a difference calculation step of comparing the apparent viscosity calculated in the apparent viscosity calculation step with the apparent viscosity acquired from the master data to calculate a difference; and a speed control step of setting a mold clamping speed at which the difference converges when a difference greater than a predetermined value is calculated in the difference calculation step, and instructing the processing machine to control the mold at the set mold clamping speed.

[0010] The control device for a processing machine according to a third aspect of the present invention is a control device for a processing machine that processes a material placed in a mold by press molding, and comprises at least a calculation unit and a command unit, and the calculation unit executes a material database acquisition process that acquires a correlation between a material-specific mold clamping speed and the apparent viscosity of the material in advance as a material database, a master data acquisition process that acquires mold-specific master data that indicates a change in the apparent viscosity of the material with respect to the mold clamping speed based on the correlation between the mold clamping speed and the apparent viscosity of the material in the material database, an apparent viscosity calculation process that calculates the apparent viscosity of the material during press molding based on the mold clamping speed at the mold clamping position, a difference calculation process that compares the apparent viscosity calculated in the apparent viscosity calculation process with the apparent viscosity acquired from the master data to calculate a difference, and a speed control process that, when a difference greater than a predetermined value is calculated in the difference calculation process, sets a mold clamping speed at which the difference will converge, and instructs the processing machine to control the mold at the set mold clamping speed.

[0011] A control program according to a fourth aspect of the present invention is a control program for causing a computer to function as the control device for the processing machine. Effect of the Invention

[0012] According to the processing machine control method, processing machine control device, and control program of the present invention, the rate of defective products can be further reduced. [Brief description of the drawings]

[0013] [Figure 1] 1 is a conceptual diagram showing a configuration of a press molding system 1 in an embodiment. [Diagram 2] 2 is a block diagram for explaining the procedure of a control method in the press-molding system 1. FIG. [Diagram 3] 1 is a block diagram showing a hardware configuration of a press molding system 1. FIG. [Figure 4] 11 is a graph of first correlation data showing a change in stress with respect to a mold clamping position. [Diagram 5]11 is a graph of second correlation data showing the change in material strain rate with respect to the mold clamping position. [Figure 6] 1 is a graph of mold-specific master data showing the change in apparent viscosity of a material versus mold clamping position. [Figure 7] 1 is a graph from a material database showing the correlation between strain rate and apparent viscosity of a material. [Figure 8] 11 is a flowchart showing a procedure of a master data acquisition process in test molding. [Figure 9] 4 is a flowchart showing the procedure of an apparent viscosity calculation process, a difference calculation step, and a speed control process in main molding. [Figure 10A] FIG. [Figure 10B] This is a cross-sectional view of s1-s1 in FIG. 10A. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0014] Hereinafter, an embodiment of a press forming system to which a processing machine control method, a processing machine control device, and a control program according to the present invention are applied will be described. Note that the drawings attached to this specification are all schematic diagrams, and the shape, scale, aspect ratio, and the like of each part are modified or exaggerated from the actual ones in consideration of ease of understanding, etc.

[0015] 1 is a conceptual diagram showing the configuration of a press molding system 1 according to an embodiment. As shown in FIG 1, the press molding system 1 according to the embodiment includes a hydraulic press 2, a control computer 3, and an artificial intelligence computer 4.

[0016] The hydraulic press 2 is a processing machine that processes a material placed in a mold by press molding. In this embodiment, an example will be described in which a laminate of discontinuous carbon fiber composite material impregnated with a thermoplastic resin (hereinafter, also referred to as "CTT material") is used as the material. As shown in FIG. 1, the hydraulic press 2 includes molds 11 and 12, a press unit 13, a cylinder 14, an oil supply pipe 15, a hydraulic pump 16, a servo motor 17, and a housing 18. The hydraulic press 2 also includes a pressure sensor 21 and a displacement sensor 22. Note that the configuration of the hydraulic press 2 shown in FIG. 2 is an example, and the processing machine to which the present invention can be applied is not limited to the hydraulic press 2 shown in FIG. 2.

[0017] Mold 11 is a fixed mold having a molding surface of a predetermined shape (e.g., a concave shape). Mold 12 is a movable mold having a molding surface of a shape (e.g., a convex shape) corresponding to the molding surface of mold 11. Molds 11 and 12 are arranged separately in the vertical direction so that their respective molding surfaces face each other. Mold 12 is attached to a press unit 13 that is movable in the vertical direction. Molds 11 and 12, press unit 13, and cylinder 14 are arranged in a housing 18.

[0018] During molding, the press unit 13 moves downward by force applied from the cylinder 14. The cylinder 14 is connected to a hydraulic pump 16 via an oil supply pipe 15, and a downward force is applied by hydraulic oil delivered from the hydraulic pump 16. The hydraulic pump 16 delivers hydraulic oil whose flow rate and pressure are adjusted to the cylinder 14. The hydraulic pump 16 is driven by a servo motor 17. The servo motor 17 generates a rotational force for driving the hydraulic pump 16. Note that the power source for controlling the hydraulic pump 16 is not limited to the servo motor 17, and other mechanisms may be used. The rotation of the servo motor 17 is controlled by a control computer 3 (described later).

[0019] The mold (fixed mold) 11 is provided with a pressure sensor 21. The pressure sensor 21 is a sensor that detects the pressure inside the mold. In this embodiment, the pressure is detected by the pressure sensor 21 during test molding (described later). In main molding (described later) performed after the test molding, it is not necessary to detect the pressure by the pressure sensor 21. It is preferable to perform the main molding using a mold 11 that does not include a pressure sensor 21, but the main molding may be performed using a mold 11 that includes a pressure sensor 21. The pressure sensor 21 shown in FIG. 1 may be provided in the mold (movable mold) 12.

[0020] A displacement sensor 22 is provided near the die 12. The displacement sensor 22 detects the position of the die 12 relative to the die 11 (hereinafter also referred to as the "die clamping position"). The die clamping position is represented by a clearance (spacing) between the die 11 and the die 12. During press molding, the die clamping position displaces with time, so the die clamping speed can be calculated by differentiating the die clamping position with time.

[0021] The information detected by each of the above sensors is transmitted as information related to changes in the state of the material to the control computer 3 and the artificial intelligence computer 4 via an industrial high-speed network system. Such information communication between each of the sensors and the computer is performed by wired or wireless means, for example, through an industrial high-speed network system having a communication speed of 10 megabits per second or more. For example, "EtherCAT" (registered trademark) is widely known as an industrial high-speed network system having such a high level of communication performance.

[0022] Prior to test molding, which will be described later, the control computer (control device) 3 acquires in advance a correlation between the mold clamping speed specific to the material and the apparent viscosity of the material as a material database (material database acquisition process / material database acquisition step). In this specification, the "apparent viscosity" does not mean the viscosity of the resin alone, but the viscosity of a fluid made of a composite material of fibers and resin. As will be described later, the material database is acquired by molding using a material universal testing machine and a hydraulic press machine, and is stored, for example, in storage 35 (see FIG. 3). The control computer 3 acquires the material database from storage 35. The material database will be described later in detail with reference to FIG. 7.

[0023] In the test molding, the control computer 3 calculates data showing a change in stress relative to the mold clamping position (hereinafter also referred to as "first correlation data": see FIG. 4) and data showing a change in strain rate of the material relative to the mold clamping position (hereinafter also referred to as "second correlation data": see FIG. 5) based on information acquired from the pressure sensor 21 and the displacement sensor 22, and acquires data showing a change in apparent viscosity of the material relative to the mold moving position (hereinafter also referred to as "third correlation data") based on these correlation data. Although a graph of the third correlation data is not shown, it is data that is the basis of master data (see FIG. 7) described later. In this embodiment, the process of acquiring the third correlation data is included in the master data acquisition process (master data acquisition step) described later.

[0024] In this specification, "test molding" refers to molding for obtaining information necessary for calculating the above correlation data. "Main molding" refers to molding for actually manufacturing a product. Test molding is performed prior to main molding, but the same molds, materials, etc. are used. Molded products obtained in test molding include not only good products but also defective products. The control computer 3 calculates the first correlation data and the second correlation data based on the information obtained when a good product is obtained in test molding.

[0025] The control computer 3 acquires master data based on the third correlation data and a trained model (described later) generated by the artificial intelligence computer 4 (master data acquisition process / master data acquisition step). "Master data" is mold-specific data that indicates the change in apparent viscosity of the material relative to the mold fastening position that does not produce defective products. More specifically, the master data is data of an ideal form used in the main molding so that the hydraulic press machine 2 completes the processing step without producing defective products. Note that, since the flow time of the material changes depending on the shape of the mold, it is necessary to acquire master data for each mold.

[0026] In the present embodiment, in the master data acquisition process, a third correlation data is calculated based on the first correlation data and the second correlation data, and the learned model is applied to the third correlation data to calculate the master data, but the present invention is not limited to this. For example, in test molding, the third correlation data calculated based on the first correlation data and the second correlation data when a non-defective product is molded may be used as the master data.

[0027] During the main molding, the control computer 3 calculates the apparent viscosity of the material during press molding using the mold clamping speed calculated based on the mold clamping position detected by the displacement sensor 22 and formula (2) (apparent viscosity calculation process / apparent viscosity calculation step). The control computer 3 compares the apparent viscosity calculated in the apparent viscosity calculation process with the apparent viscosity acquired from the master data, and calculates the difference (difference calculation process / difference calculation step).

[0028] Specifically, the control computer 3 refers to the master data, sequentially identifies points corresponding to the mold clamping positions detected by the displacement sensor 22, and at each identified point, compares the apparent viscosity calculated in the apparent viscosity calculation process with the apparent viscosity acquired from the master data. When a difference equal to or greater than a predetermined value is calculated in the difference calculation process, the control computer 3 sets a mold clamping speed at which the difference converges, and instructs the servo motor 17 (hydraulic press 2) to control the mold 12 at the set mold clamping speed (speed control process / speed control step).

[0029] Specifically, the control computer 3 transmits a rotation speed command (correction command d: see FIG. 2) to the servo motor 17 to correct the rotation speed so that the main molding process is performed at an appropriate processing speed. For this communication as well, the above-mentioned industrial high-speed network system is used to realize feedforward control. In the servo motor 17, the rotation speed is corrected according to the rotation speed command transmitted from the control computer 3. As a result, the hydraulic pump 16 is operated at the corrected rotation speed. The hydraulic pump 16 sends the corrected appropriate hydraulic pressure to the cylinder 14 via the oil supply pipe 15. The cylinder 14 applies the corrected appropriate force to the die 12 via the press unit 13.

[0030] The artificial intelligence computer 4 performs machine learning using the information acquired from the displacement sensor 22 and the pressure sensor 21 during the test molding and the result information on whether or not a defective product has occurred during the test molding as training data, and generates a trained model. That is, the artificial intelligence computer 4 uses information when a good product is molded as training data. The trained model is updated by sequentially executing machine learning in the artificial intelligence computer 4. The trained model generated by the artificial intelligence computer 4 is an algorithm for the control computer 3 to calculate master data. Note that the artificial intelligence computer 4 may acquire information from the displacement sensor 22 and the pressure sensor 21 via the control computer 3.

[0031] In machine learning, the more teacher data is referred to, the more accurate the generated algorithm is. In the artificial intelligence computer 4, the learning accuracy can be autonomously improved by sequentially updating the trained model (algorithm). Machine learning by the artificial intelligence computer 4 includes deep learning using a neural network. By performing such deep learning, for example, in test molding, even if unknown information not included in the teacher data is input, a trained model capable of calculating accurate and appropriate master data through inference can be generated. The artificial intelligence computer 4 transmits the trained model generated by machine learning to the control computer 3. Such transmission of the trained model is performed every time the trained model is updated.

[0032] 2 is a block diagram for explaining the procedure of the control method in the press molding system 1. The controller 19 outputs a rotation speed command b to the servo motor 17 based on a position command a input in advance by a programmable logic controller (PLC) or the like. In the test molding, the rotation speed of the servo motor 17 is controlled so that the die moves at a preset die fastening speed.

[0033] On the other hand, in the main molding, feedforward control is also used. In the feedforward control, when a difference equal to or greater than a predetermined value is calculated in the difference calculation step, the control computer 3 sets a mold clamping speed at which the difference converges, and outputs a correction command c to correct the rotation speed command b output from the controller 19. In this way, by implementing the feedforward control of this embodiment, the occurrence of defective products in the hydraulic press 2 can be further reduced.

[0034] 3 is a block diagram showing a hardware configuration of the press forming system 1. As shown in FIG.

[0035] The CPU 30 executes a program stored in the memory 34 or a program read from the storage 35 into the memory 34, thereby working with each piece of hardware to realize the functions of each part described below. The receiving unit 31 receives information relating to changes in the state of the material from each sensor disposed in the hydraulic press 2.

[0036] The calculation unit 32 executes a material database acquisition process, a master data acquisition process, an apparent viscosity calculation process, and a difference calculation process, which will be described later. In the material database acquisition process, the calculation unit 32 acquires a material database from the storage 35. In the master data acquisition process, the calculation unit 32 calculates master data indicating a change in the apparent viscosity of the material with respect to the mold position, based on a correlation (described later) between the mold clamping speed and the apparent viscosity of the material. In the apparent viscosity calculation process, the calculation unit 32 calculates the apparent viscosity of the material during main molding, based on the mold clamping speed at a predetermined mold clamping position. In the difference calculation process, the calculation unit 32 compares the apparent viscosity calculated in the apparent viscosity calculation process with the apparent viscosity acquired from the master data, and calculates the difference.

[0037] The command unit 33 executes a speed control process. In the speed control process, when a difference equal to or greater than a predetermined value is calculated in the difference calculation process of the calculation unit 32, the command unit 33 sets a mold clamping speed at which the difference converges, and commands the hydraulic press 2 (servo motor 17) to control the mold at the set mold clamping speed. This command is a correction command c outputted to correct the rotation speed command b outputted from the controller 19 (see FIG. 2).

[0038] The CPU 30 is mutually connected to the memory 34, the storage 35, and the communication unit 36 ​​via a bus 38. The memory 34 is a storage device that temporarily stores working data used by the CPU 30 for calculations. The storage 35 is a storage device that stores various programs executed by the CPU 30, data files, and the like. The communication unit 36 ​​is connected to the pressure sensor 21, the displacement sensor 22, the servo motor 17, and the artificial intelligence computer 4 arranged in the hydraulic press 2 via a wired or wireless industrial high-speed network system 5.

[0039] The input / output unit 37 is composed of an input device, such as a touch panel or a keyboard, that can input various types of information, and an output device, such as a display, that can output various types of information as images or the like. The control computer 3 and the artificial intelligence computer 4 may be independent separate hardware, or may be constructed as separate functions integrated into a single industrial high-speed network system.

[0040] Next, the correlation between the mold clamping position during press molding and the apparent viscosity of the material will be described. Note that the graphs shown in Fig. 4 to Fig. 6, which will be described later, are graphs conceptually showing the change in each parameter with respect to the mold clamping position.

[0041] Fig. 4 is a graph of the first correlation data showing the change in stress with respect to the mold clamping position. In Fig. 4, the mold clamping position H on the horizontal axis indicates the clearance between the molds, and is measured by the displacement sensor 22 (see Fig. 1). The mold clamping position H decreases in value from right to left on the horizontal axis as the press molding progresses. The stress (pressure inside the mold) σ on the vertical axis is measured by the pressure sensor 21.

[0042] Fig. 5 is a graph of the second correlation data showing the change in strain rate of the material with respect to the mold clamping position. In Fig. 5, the horizontal axis indicates the mold clamping position H. The vertical axis, strain rate ε, indicates the rate of change over time of the strain occurring in the material. The strain rate ε is calculated by time-differentiating the mold clamping position H measured by the displacement sensor 22 to calculate the mold clamping speed (dt), and dividing this mold clamping speed by the original mold clamping position H. The strain rate ε is calculated by the following formula (1). Strain rate: ε[1 / sec]=1 / H dh(t) / dt (1) In this specification, the term "ε" representing strain rate means ε with a dot above it, which indicates time differentiation.

[0043] Figure 6 is a graph of master data specific to the mold, showing the change in apparent viscosity of the material with respect to the mold movement position. In Figure 6, the horizontal axis represents the mold clamping position H, and the vertical axis represents the apparent viscosity ηap of the material. As shown in Figure 6, the apparent viscosity ηap rises sharply once from the position where the mold contacts the material, then reaches an apparent viscosity at which the material flows, becomes an almost constant viscosity value, and then rises sharply again in the final stage of molding.

[0044] The apparent viscosity ηap is the ratio of the stress σ (first correlation data: see FIG. 4) to the strain rate ε (second correlation data: see FIG. 5), and is calculated by the following formula (2). Apparent viscosity: ηap[Pa·sec]=σ / ε···(2) In this way, by using the stress σ and strain rate ε obtained by test molding and equation (2), the change in apparent viscosity (estimated value) of the material with respect to the mold clamping position can be calculated.

[0045] FIG. 7 is a graph of a material database showing the correlation between the strain rate and the apparent viscosity of a material. The graph shown in FIG. 7 shows the test results of investigating the change in the apparent viscosity of a material with respect to the strain rate using a universal material testing machine and a hydraulic press machine equipped with dies of the same shape. In FIG. 7, the horizontal axis shows the strain rate ε. The strain rate ε can be calculated based on the die clamping speed (dt) as shown in the above-mentioned formula (1). Since the strain rate ε and the die clamping speed are in a proportional relationship, in FIG. 7, the strain rate ε can be essentially regarded as the die clamping speed. The vertical axis shows the apparent viscosity ηap of the material. As the material to be tested, multiple samples of the same shape and structure were prepared, and molding was performed by changing the press molding conditions (molding temperature, die clamping speed).

[0046] In FIG. 7, the black circle marks and black square marks indicate the apparent viscosity of the material measured by the universal material testing machine. The black circle marks indicate the measured value at a molding temperature of 200°C. The black square marks indicate the measured value at a molding temperature of 150°C. The white circle marks and white square marks indicate the apparent viscosity of the material calculated using the values ​​measured by the displacement sensor and pressure sensor installed in the hydraulic press machine and equations (1) and (2). The white circle marks indicate the measured value at a molding temperature of 200°C. The white square marks indicate the measured value at a molding temperature of 150°C. Each mark (plot point) indicates the apparent viscosity extracted at the mold clamping position (the position where the material is assumed to flow at a nearly constant speed) where the apparent viscosity shows a nearly constant value among the values ​​calculated by the universal material testing machine and the hydraulic press machine.

[0047] As shown in Figure 7, the change in apparent viscosity measured by the hydraulic press machine is almost the same as that measured by the universal testing machine. The apparent viscosity characteristics of the material measured in this experiment are not dependent on the device, but are considered to be inherent characteristics of the material, namely, CTT material. CTT material has excellent formability due to its high material fluidity, but has a large variation in mechanical properties, so it is necessary to appropriately control the forming conditions according to the material flow state during press forming. However, with conventional sensor measurement technology, it is difficult to quantitatively measure the material flow state of CTT material during press forming, which makes it difficult to appropriately control the forming conditions. In response to this, the applicants of the present application have clarified the correlation between the strain rate and apparent viscosity during press forming (see Figure 7), and have found a method to estimate the material fluidity during press forming from the mold clamping position (displacement sensor) based on a material database showing this correlation.

[0048] Next, a method for controlling the hydraulic press 2 (processing machine) in this embodiment will be described. First, the master data acquisition process in test molding will be described. FIG. 8 is a flowchart showing the procedure of the master data acquisition process in test molding. In test molding, the die 12 is controlled at a preset die clamping speed. Test molding is performed a preset number of times. From the start to the end of one molding, the process shown in FIG. 8 is executed in conjunction with the movement of the die, for example, at intervals of tens to hundreds of ms.

[0049] Although not shown, the control computer 3 executes a process (material database acquisition process) for acquiring a correlation between the mold clamping speed and the apparent viscosity of a material specific to the material in advance as a material database prior to the test molding. If the correlation between the mold clamping speed and the apparent viscosity of a material specific to the material is known in advance, the material database acquisition process may be omitted in the control computer 3.

[0050] In step S101 of FIG. 8, the control computer obtains information on the mold clamping position and stress during test molding from the pressure sensor 21 and the displacement sensor 22.

[0051] In step S102, the control computer 3 calculates data (first correlation data) indicating the change in stress relative to the mold clamping position and data (second correlation data) indicating the change in strain rate of the material relative to the mold clamping position based on the acquired information.

[0052] In step S103, the control computer 3 calculates data (third correlation data) indicating a change in apparent viscosity of the material with respect to the mold clamping position, based on the calculated first correlation data and second correlation data.

[0053] In step S104, the control computer 3 calculates master data based on the calculated third correlation data and the trained model received from the artificial intelligence computer 4. The master data is stored in, for example, the storage 35 (see FIG. 3). After the process of step S104 is performed, the process of this flowchart ends.

[0054] On the other hand, in step S201 of FIG. 8, the artificial intelligence computer 4 obtains information on the mold clamping position and stress during test molding from the pressure sensor 21 and the displacement sensor 22. In step S202, the artificial intelligence computer 4 performs machine learning using the information acquired from the sensor and the result information on whether or not a defective product was generated in the test molding as training data, and generates a trained model. The trained model is stored in, for example, the storage 35 (see FIG. 3).

[0055] In step S203, the artificial intelligence computer 4 transmits the trained model generated by machine learning to the control computer 3.

[0056] In step S204, the artificial intelligence computer 4 determines whether or not the trained model (algorithm) needs to be updated. In step S204, if the artificial intelligence computer 4 determines that the trained model needs to be updated (YES), the process proceeds to step S205. In addition, in step S204, if the artificial intelligence computer 4 determines that the trained model does not need to be updated (NO), the process returns to step S201.

[0057] In step S205 (step S204: YES), the artificial intelligence computer 4 updates the trained model and returns to step S201.

[0058] Next, the apparent viscosity calculation process, the difference calculation step, and the speed control process in the main molding will be described. Fig. 9 is a flow chart showing the procedure of each process in the main molding. In the main molding, the mold clamping speed is controlled by feedforward control using a position command a and a correction command c (see Fig. 2). The main molding is performed a predetermined number of times for the production of products. The process shown in Fig. 9 is executed in conjunction with the movement of the mold, for example, at intervals of tens to hundreds of milliseconds.

[0059] In step S301 shown in FIG. 9, the control computer 3 calculates the apparent viscosity of the material during press molding using the mold clamping speed calculated based on the mold clamping position detected by the displacement sensor 22 and equation (2) (apparent viscosity calculation process).

[0060] In step S302, the control computer 3 compares the apparent viscosity calculated in the apparent viscosity calculation process with the apparent viscosity acquired from the master data to calculate the difference (difference calculation process). In this difference calculation process, the control computer 3 refers to the master data, sequentially identifies points corresponding to the mold clamping position detected by the displacement sensor 22, and at each identified point, compares the apparent viscosity calculated in the apparent viscosity calculation process with the apparent viscosity acquired from the master data.

[0061] In step S303, the control computer 3 judges whether or not a difference equal to or greater than a predetermined value has been calculated in the difference calculation process. If it is judged in step S303 that a difference equal to or greater than a predetermined value has been calculated (YES), the process proceeds to step S304. If it is judged in step S303 that a difference equal to or greater than a predetermined value has not been calculated (NO), the process of this flowchart ends.

[0062] In step S304 (step S303: YES), the control computer 3 sets a mold clamping speed at which the difference calculated in the difference calculation process converges, and instructs the servo motor 17 to control the mold 12 at the set mold clamping speed (speed control process). In step S304, feedforward control is performed using the position command a and the correction command c (see FIG. 2). After the process of step S304 is performed, the process of this flowchart ends.

[0063] (Example) Next, experimental results of manufacturing a molded product by the press molding system 1 of the embodiment will be described. FIG. 10A is a plan view of the molded product. FIG. 10B is a cross-sectional view taken along the line s1-s1 of FIG. 10A. As shown in FIGS. 10A and 10B, the molded product 100 is rectangular in plan view as a whole, and has a box-shaped recessed portion 101 and a cross-shaped reinforcing portion 102 near the center. The reinforcing portion 102 is a portion that protrudes in the height direction from the bottom surface of the recessed portion 101. The molded product 100 also has a brim-shaped rib 103 on the outer periphery of the recessed portion 101. The shape of this molded product 100 is designed to enable evaluation of the material fluidity of the rib-shaped portion (rib 103) in the out-of-plane direction by replacing the die insert. The dimensions of each portion of the molded product 100 are as follows.

[0064] L1: 230mm L2: 166 mm (length of the reinforcing part 102 in the X direction) L3: 140mm L4: 48 mm (length of the reinforcing portion 102 in the Y direction) L5: 2.5 mm (thickness of reinforcing portion 102) H: 15mm (height of reinforcement part 102) D: 28mm (depth of molded part 100) t:3mm 10A and 10B show mutually orthogonal coordinate systems of XYZ. In this coordinate system, the longitudinal direction of the molded article 100 when viewed from the position of FIG. 10A is defined as the X direction, and the lateral direction is defined as the Y direction. As shown in FIG. 10B, the depth direction of the molded article 100 is defined as the Z direction.

[0065] As the molding material, a tape with a width of 13 mm, which had been previously impregnated with a thermoplastic resin, was cut to a tape length of 26 mm, and the material was laminated in a random direction and molded into a flat plate by press molding. As the thermoplastic resin, a thermoplastic epoxy resin was used. Then, the molding material was heated to 200°C by an infrared heater, and then transferred to a mold that had also been heated to 200°C, and press molding was performed. In the comparative example, press molding was performed with the mold clamping speed set to a set value of 10 mm / sec. In the example, press molding was performed while changing the mold clamping speed between 10 and 14 mm / sec by the feedforward control (see FIG. 9) of the embodiment.

[0066] The quality of the molded products was judged as "defective" when the molding material could not reach the end of the mold and stopped midway, and "good" otherwise. However, the defect in the clear shape that occurred on the rib of the molded product was a structure that occurred at the end of the tape of the tape molding material, so the molded product with this defect was not judged as defective.

[0067] As a result of checking the appearance of the molded products, in the comparative example where the mold clamping speed was 10 mm / sec, 10 out of 18 molded products were found to have defective appearances. This is thought to be because in the comparative example, the apparent viscosity of the molding material increased during press molding, preventing the molding material from reaching the edge of the mold, resulting in a short molding in which the tips of the ribs of the molded products were missing.

[0068] On the other hand, in the example in which the mold clamping speed was changed between 10 and 14 mm / sec, the appearance of all four molded products was good. From these experimental results, it is considered that in the example in which the mold clamping speed was changed during press molding, the increase in the apparent viscosity of the molding material during press molding was suppressed, and the molding material reached the end of the mold, so that short molding did not occur. From the above results, it became clear that the press molding system 1 of the embodiment can further reduce the rate of defective products.

[0069] In addition, by analyzing the results of the speed control process (see step 3304 in Figure 9) performed in this molding in real time and performing test molding again using the analysis results, the accuracy of the first correlation data and the second correlation data obtained in the test molding can be further improved.

[0070] (Variations) Although the embodiment of the present invention has been described above, the present invention is not limited to the above-mentioned embodiment, and various modifications and changes are possible, such as the modified forms described below, which are also included in the technical scope of the present invention. Furthermore, the effects described in the embodiment are merely a list of the most preferable effects resulting from the present invention, and are not limited to those described in the embodiment. The above-mentioned embodiment and the modified forms described below can be used in appropriate combination, but detailed description will be omitted.

[0071] The material used for press molding is not limited to a laminate of discontinuous carbon fiber composite material impregnated with a thermoplastic resin (CTT material), but may be other materials, such as a laminate of discontinuous carbon fiber composite material impregnated with a thermosetting resin, composite fiber, glass fiber, aramid fiber, or naturally derived fiber. [Explanation of symbols]

[0072] 1: Press forming system, 2: Hydraulic press machine, 3: Control computer, 4: Artificial intelligence computer, 22: Displacement sensor, 30: CPU, 31: Receiving unit, 32: Calculation unit, 33: Command unit, 34: Memory, 35: Storage, 36: Communication unit, 37: Input / output unit

Claims

1. A method for controlling a processing machine that processes a material placed in a die by press molding, comprising the steps of: A material database acquisition step of acquiring a correlation between a material-specific mold clamping speed and an apparent viscosity of the material as a material database in advance; a master data acquisition step of acquiring mold-specific master data indicating a change in the apparent viscosity of the material relative to the mold clamping speed based on a correlation between the mold clamping speed and the apparent viscosity of the material in the material database; An apparent viscosity calculation step of calculating an apparent viscosity of the material during press molding based on a mold clamping speed at a mold clamping position; a difference calculation step of comparing the apparent viscosity calculated in the apparent viscosity calculation step with the apparent viscosity acquired from the master data to calculate a difference; a speed control step of setting a mold clamping speed at which the difference is converged when a difference equal to or greater than a predetermined value is calculated in the difference calculation step, and instructing the processing machine to control the mold at the set mold clamping speed; A method for controlling a processing machine having the above structure.

2. A method for controlling a processing machine that processes a material placed in a die by press molding, comprising the steps of: A master data acquisition step of acquiring mold-specific master data indicating a change in the apparent viscosity of the material relative to the mold clamping speed based on a correlation between the mold clamping speed and the apparent viscosity of the material specific to the material; An apparent viscosity calculation step of calculating an apparent viscosity of the material during press molding based on a mold clamping speed at a mold clamping position; a difference calculation step of comparing the apparent viscosity calculated in the apparent viscosity calculation step with the apparent viscosity acquired from the master data to calculate a difference; a speed control step of setting a mold clamping speed at which the difference is converged when a difference equal to or greater than a predetermined value is calculated in the difference calculation step, and instructing the processing machine to control the mold at the set mold clamping speed; A method for controlling a processing machine having the above structure.

3. The master data is calculated based on first correlation data indicating a change in stress inside the mold relative to the mold clamping position, and second correlation data indicating a change in strain rate of the material relative to the mold clamping position. A method for controlling a processing machine according to claim 1 or 2.

4. The master data is sequentially learned by machine learning using an artificial intelligence computer that uses information on the molding of a non-defective product as training data. A method for controlling a processing machine according to claim 1 or 2.

5. The material is a laminate of discontinuous carbon fiber composite material impregnated with a thermoplastic or thermosetting resin; A method for controlling a processing machine according to claim 1 or 2.

6. A control device for a processing machine that processes a material placed in a die by press molding, The apparatus includes at least a calculation unit and a command unit, The calculation unit is A material database acquisition process for acquiring a correlation between a material-specific mold clamping speed and an apparent viscosity of the material as a material database in advance; a master data acquisition process for acquiring master data specific to a mold, which indicates a change in the apparent viscosity of the material with respect to the mold clamping speed, based on a correlation between the mold clamping speed and the apparent viscosity of the material in the material database; An apparent viscosity calculation process for calculating the apparent viscosity of the material during press molding based on a mold clamping speed at a mold clamping position; a difference calculation process for comparing the apparent viscosity calculated in the apparent viscosity calculation process with the apparent viscosity acquired from the master data to calculate a difference; When a difference equal to or greater than a predetermined value is calculated in the difference calculation process, a mold clamping speed at which the difference is converged is set, and a speed control process is executed to instruct the processing machine to control the mold at the set mold clamping speed. Control device for processing machinery.

7. A control program for causing a computer to function as the control device for the processing machine according to claim 6.

Images