JP7512870B2 - Viscosity estimation device, viscosity estimation method, viscosity estimation program, recording medium, and method for manufacturing molded product - Google Patents

Description

The present invention relates to a technology that is effective when applied to viscosity estimation technology for estimating the viscosity of resin.

Non-Patent Document 1 describes technology related to a rheometer that measures the viscosity of resins.

Ueda, Rheological property measurement method and its device, Journal of the Society of Rubber Science and Technology of Japan, Vol. 88, No. 8, (2015)

In general, in resin molding technology, the viscosity of the raw resin material depends on temperature and shear rate, and the quality of the molded product changes depending on the viscosity. Furthermore, depending on the type of resin, even resin materials of the same grade may vary in quality due to differences in product lots, etc. Therefore, in order to maintain the quality of molded products, it is necessary to optimize molding conditions such as temperature and shear rate each time according to changes in the viscosity of the resin material. In other words, in the manufacturing process of molded products, it is necessary to understand the viscosity of the resin in order to adjust the molding conditions to absorb the viscosity changes.

In this regard, currently there is no technology to grasp the viscosity of resin other than actually measuring the viscosity using a rheometer. However, introducing and operating technology that uses a rheometer is often difficult both financially and in terms of lead time. For this reason, molding conditions are often adjusted based on the experience of the worker. In other words, with current technology, measuring viscosity is difficult and time-consuming, so it is difficult to adjust molding conditions based on viscosity measurements. Therefore, there is a demand for technology that can easily grasp viscosity.

In one embodiment, the viscosity estimation device includes a viscosity estimation unit that estimates viscosity based on the melt flow rate.

The viscosity estimation method in one embodiment includes a viscosity estimation step that estimates the viscosity based on the melt flow rate.

The program in one embodiment includes a viscosity estimation process that estimates viscosity based on the melt flow rate. The program is recorded on a computer-readable recording medium.

The manufacturing method for a molded product in one embodiment includes a viscosity estimation process for estimating the viscosity of the resin based on the melt flow rate, a molding condition adjustment process for adjusting the molding conditions based on the estimated viscosity, and a molding process for molding the resin under the adjusted molding conditions.

According to one embodiment, the viscosity of the resin can be easily grasped.

FIG. 1 is a diagram showing an outline of a method for measuring "MFR". FIG. 2 is a diagram illustrating an example of a hardware configuration of a viscosity estimation device. FIG. 2 is a functional block diagram showing functions of the viscosity estimation device. FIG. 11 is a diagram illustrating an example of a known data group. FIG. 1 is a diagram illustrating a neural network learning process. 4 is a flowchart illustrating a first operation of the viscosity estimation device. 10 is a flowchart illustrating a second operation of the viscosity estimation device. FIG. 2 is a functional block diagram of the viscosity estimation device. 4 is a flowchart illustrating the operation of the viscosity estimation device. 1 is a graph showing a comparison result between an actual viscosity value of a sample not used for determining a constant and an estimated viscosity value estimated using (Equation 2) taking into account "MFR". 1 is a table showing the results of comparing a technique for actually measuring viscosity (comparative example) with the technical idea of the present embodiment in which viscosity is estimated using "MFR." 1 is a flowchart showing the flow of a manufacturing method for a molded product incorporating the technical idea of the present embodiment, which estimates viscosity based on "MFR."

In all drawings used to explain the embodiments, the same components are generally given the same reference numerals, and repeated explanations will be omitted. In addition, hatching may be used even in plan views to make the drawings easier to understand.

<Basic Concept of the Embodiment>
The basic idea in this embodiment is to estimate the viscosity of a resin based on the melt flow rate (hereinafter referred to as "MFR").

First, let us explain "MFR".

"MFR" is a scale that indicates the fluidity of a resin in a solution state. The test machine used to measure this "MFR" is an extrusion-type plastometer specified in "JIS K6760." The method for measuring "MFR" is specified in "JIS K7210."

Figure 1 shows an overview of the measurement method for measuring "MFR".

As shown in FIG. 1, the "MFR" measuring device 10 has a heatable barrel 1. The barrel 1 has a cylindrical through-hole. A test sample, resin 2, is inserted into the through-hole. The device is configured so that pressure can be applied to the resin 2 from above the resin 2 by a piston 4 via a piston head 3. In particular, a weight 5 is placed on the top of the piston 4, and pressure is applied from the piston 4 to the resin 2 by the weight 5. Meanwhile, a die 6 having an opening is provided at the bottom of the through-hole. In the measuring device 10 configured in this way, the resin 2 inserted into the through-hole is heated at a constant temperature and pressurized at a constant pressure. Under these conditions, the amount of resin extruded from the opening of the die 6 provided at the bottom of the through-hole in 10 minutes is measured. This allows the measuring device 10 to measure the "MFR". In general, the larger the "MFR" value, the better the fluidity and processability of the resin when melted.

For example, a large "MFR" means that a larger amount of resin is extruded, which can be understood to correspond to a state of high fluidity and low viscosity. In other words, qualitatively, it can be considered that as the "MFR" increases, the viscosity decreases. In other words, it can be considered that as the "MFR" decreases, the viscosity increases.

In this regard, viscosity is generally expressed as follows (Equation 1):

here,
η: Viscosity η ₀ : Constant dγ/dt: Shear rate n: Constant α: Constant T: Resin temperature According to this (Formula 1), assuming that the constants are known, the viscosity can be estimated if the shear rate and the resin temperature are known, but (Formula 1) does not include "MFR". For this reason, even if "MFR" is known, the viscosity cannot be estimated from (Formula 1). In other words, even if the qualitative relationship between "MFR" and viscosity is known, the viscosity cannot be quantitatively estimated from "MFR" based on (Formula 1). Thus, at present, a relational equation relating "MFR" and viscosity has not been obtained.

If the viscosity can be estimated from the "MFR", the following excellent advantages can be obtained. That is, for example, the "MFR" is often provided by a resin material manufacturer for each lot. Therefore, if a relational equation relating the "MFR" to the viscosity can be obtained, the viscosity can be estimated from the "MFR" provided by the manufacturer without measuring the viscosity with a rheometer. This means that the viscosity can be grasped without measuring it. As a result, in the manufacturing process of molded products, the viscosity can be grasped from the "MFR" without measuring the viscosity. This has the advantage that it becomes easier to adjust the molding conditions to absorb viscosity changes.

The inventors have focused on this point and, as a result of extensive research, have come up with the following (Formula 2) based on (Formula 1).

here,
η: Viscosity C: Constant MFR: Melt flow rate dγ/dt: Shear rate n: Constant α: Constant T: Resin temperature According to this (Formula 2), it is consistent with the qualitative understanding that the viscosity decreases as the "MFR" increases. And (MFR) ⁿ is determined by dimensional analysis. Specifically, attention is paid to the unit of "sec" included in (Formula 2). That is, it is considered that the unit of viscosity is "Pa·sec", the unit of shear rate is "1/sec", and the unit of "MFR" is "cc/10 min" = "cc/600 sec". At this time, the dimension of "sec" on the left side of (Formula 2) becomes "one-dimensional" from "Pa·sec". In contrast, the dimension of "sec" on the right side of (Formula 2) is 1/(MFR) ⁿ ×(dγ/dt) ^n-1 , so (sec) ⁿ ×(1/sec) ^n-1 =(sec), which is "one-dimensional." From such dimensional analysis, it makes sense that the dependency of "MFR" in (Formula 2) is introduced in the form of 1/(MFR) ⁿ .

In this way, according to (Formula 2) which was conceived based on the qualitative relationship between viscosity and "MFR" and dimensional analysis, it is possible to estimate viscosity if "MFR", shear rate, and resin temperature are known, assuming that the constants are known. In other words, it can be seen that (Formula 2) is a relational equation relating "MFR" and viscosity. Therefore, if the constants included in (Formula 2) can be appropriately determined, it is considered possible to accurately estimate viscosity from "MFR" based on (Formula 2). For this reason, it is important to determine the constants included in (Formula 2), and in this embodiment, the constants included in (Formula 2) are determined using machine learning.

The basic idea of this embodiment is to estimate viscosity based on "MFR", and more specifically, to estimate viscosity based on (Equation 2) that relates "MFR" to viscosity. In this basic idea, the constants included in (Equation 2) are determined by machine learning so that viscosity can be estimated from "MFR" with high accuracy.

<Configuration of Viscosity Estimation Device>
A viscosity estimation device that embodies the above-mentioned basic concept will be described below.

<<Hardware configuration>>
First, the hardware configuration of the viscosity estimation device in this embodiment will be described.

Figure 2 is a diagram showing an example of the hardware configuration of the viscosity estimation device 100 in this embodiment. Note that the configuration shown in Figure 2 is merely an example of the hardware configuration of the viscosity estimation device 100, and the hardware configuration of the viscosity estimation device 100 is not limited to the configuration shown in Figure 2, and may be other configurations.

In FIG. 2, the viscosity estimation device 100 includes a CPU (Central Processing Unit) 101 that executes a program. This CPU 101 is electrically connected to, for example, a ROM (Read Only Memory) 102, a RAM (Random Access Memory) 103, and a hard disk drive 112 via a bus 113, and is configured to control these hardware devices.

The CPU 101 is also connected to input devices and output devices via the bus 113. Examples of input devices include the keyboard 105, the mouse 106, the communication board 107, and the scanner 111. Examples of output devices include the display 104, the communication board 107, and the printer 110. The CPU 101 may also be connected to, for example, a removable disk device 108 and a CD/DVD-ROM device 109.

The viscosity estimation device 100 may be connected to a network, for example. For example, when the viscosity estimation device 100 is connected to other external devices via a network, the communication board 107 constituting a part of the viscosity estimation device 100 is connected to a LAN (local area network), a WAN (wide area network), or the Internet.

RAM 103 is an example of a volatile memory, and the recording media ROM 102, removable disk device 108, CD/DVD-ROM device 109, and hard disk device 112 are examples of non-volatile memory. These volatile and non-volatile memories constitute the storage device of the viscosity estimation device 100.

The hard disk device 112 stores, for example, an operating system (OS) 201, a group of programs 202, and a group of files 203. The programs included in the group of programs 202 are executed by the CPU 101 using the operating system 201. The RAM 103 also temporarily stores at least some of the programs of the operating system 201 and application programs executed by the CPU 101, as well as various data required for processing by the CPU 101.

The ROM 102 stores a BIOS (Basic Input Output System) program, and the hard disk drive 112 stores a boot program. When the viscosity estimation device 100 is started, the BIOS program stored in the ROM 102 and the boot program stored in the hard disk drive 112 are executed, and the operating system 201 is started by the BIOS program and the boot program.

The program group 202 stores programs that realize the functions of the viscosity estimation device 100, and these programs are read and executed by the CPU 101. In addition, the file group 203 stores information, data, signal values, variable values, and parameters that indicate the results of processing by the CPU 101 as each item of a file.

The files are recorded on a recording medium such as the hard disk drive 112 or memory. The information, data, signal values, variable values, and parameters recorded on the recording medium such as the hard disk drive 112 or memory are read by the CPU 101 into the main memory or cache memory, and are used for the operations of the CPU 101, such as extraction, search, reference, comparison, calculation, processing, editing, output, printing, and display. For example, during the above-mentioned operations of the CPU 101, the information, data, signal values, variable values, and parameters are primarily stored in the main memory, registers, cache memory, buffer memory, etc.

The functions of the viscosity estimation device 100 may be realized by firmware stored in ROM 102, or may be realized by software alone, by hardware alone such as elements, devices, boards, and wiring, a combination of software and hardware, or even a combination with firmware. The firmware and software are recorded as programs on a recording medium such as the hard disk drive 112, a removable disk, a CD-ROM, or a DVD-ROM. The programs are read and executed by the CPU 101. In other words, the programs cause the computer to function as the viscosity estimation device 100.

In this way, the viscosity estimation device 100 is a computer equipped with a CPU 101 as a processing device, a hard disk drive 112 and memory as storage devices, a keyboard 105, a mouse 106, and a communication board 107 as input devices, and a display 104, a printer 110, and a communication board 107 as output devices. The functions of the viscosity estimation device 100 are realized by using the processing device, the storage device, the input device, and the output device.

<<Function block configuration>>
Next, the functional block configuration of the viscosity estimation device 100 will be described.

Figure 3 is a functional block diagram showing the functions of the viscosity estimation device.

The viscosity estimation device 100 has an input unit 301, a constant determination unit 302, a viscosity estimation unit 303, an output unit 304, and a data storage unit 305.

The input unit 301 is configured to input a group of known data that associates "MFR", a combination of resin temperature and shear rate, and viscosity, where "MFR", temperature, shear rate, and viscosity are known.

Figure 4 shows an example of a known data group.

The known data group 400 includes a plurality of known data, each of which is composed of a combination of "MFR", temperature, and shear rate, as well as a viscosity associated with these combinations. For example, an example of known data may be data in which "MFR" is "17.6527", "temperature" is "190", "shear rate" is "60.8", and "viscosity" is "10307.03", as shown in FIG. 4.

This known data is obtained, for example, by measuring the viscosity of resin materials of the same grade but different "MFR" while changing the temperature and shear rate. In this case, the viscosity is measured using, for example, a rheometer.

In this embodiment, the constants included in (Equation 2) are determined based on the known data group. For this reason, from the viewpoint of determining the constants in (Equation 2) with high accuracy, it is desirable that the known data group contains a large number of pieces of known data. For example, it is desirable that the known data group contains a total of 20 or more pieces of different known data for "MFR", "temperature", and "shear rate".

Each of the multiple known data input to the input unit 301 is stored in the data storage unit 305. The data storage unit 305 also stores (Equation 2).

The constant determination unit 302 is configured to determine the constants ("C", "n", "α") included in (Equation 2) using the group of known data stored in the data storage unit 305. Specifically, as shown in FIG. 5, the constant determination unit 302 is configured to determine the constants included in (Equation 2) by using the group of known data as training data and training a neural network with inputs of "MFR", temperature, and shear rate and output of viscosity.

The viscosity estimation unit 303 is configured to estimate the viscosity based on the "MFR", and more specifically, is configured to estimate the viscosity based on the above-mentioned (Equation 2). In particular, the viscosity estimation unit 303 estimates the viscosity based on (Equation 2) into which the constants determined by the constant determination unit 302 are substituted. That is, the viscosity estimation unit 303 is configured to estimate the viscosity from a combination of "MFR", whose correspondence with viscosity is unknown, and temperature and shear rate, based on (Equation 2) into which the constants determined by the constant determination unit 302 are substituted. This makes it possible to determine the viscosity of a resin with unknown viscosity if the "MFR", temperature and shear rate are known.

The output unit 304 has the function of outputting the viscosity of the resin estimated by the viscosity estimation unit 303.

The viscosity estimation device 100 is configured as described above.

<Operation of Viscosity Estimation Device (Viscosity Estimation Method)>
Next, the operation of the viscosity estimation device 100 in this embodiment will be described.

The operation of the viscosity estimation device 100 includes a first operation of determining the constants included in (Equation 2) and determining the relational equation used to estimate the viscosity, and a second operation of using this relational equation to estimate the viscosity based on the "MFR", "temperature", and "shear rate" of a resin with unknown viscosity. Each operation will be explained below.

<<First Operation>>
FIG. 6 is a flowchart illustrating a first operation of the viscosity estimation device.

First, it is assumed that a set of known data has been acquired.

In FIG. 6, the input unit 301 inputs a group of known data (S101). Then, the group of known data input to the input unit 301 is stored in the data storage unit 305.

Next, the constant determination unit 302 determines the constants included in (Formula 2) by training a neural network with "MFR", temperature, and shear rate as inputs and viscosity as output, using the group of known data stored in the data storage unit 305 as training data (S102). This determines (Formula 2), which is a relational equation relating viscosity to "MFR". In this way, the relational equation between viscosity and "MFR" is obtained.

<<Second Operation>>
FIG. 7 is a flowchart illustrating a second operation of the viscosity estimation device.

Here, it is assumed that the first operation described above has already been performed and the constants in (Equation 2) have been determined.

In FIG. 7, the input unit 301 inputs unknown data with unknown viscosity (S201). This unknown data includes data on "MFR", "temperature", and "shear rate" for a resin with unknown viscosity. This unknown data is stored in the data storage unit 305.

Next, the viscosity estimation unit 303 substitutes the unknown data stored in the data storage unit 305 into (Formula 2). That is, the known "MFR", "temperature", and "shear rate" contained in the unknown data with unknown viscosity are substituted into (Formula 2) (S202). This allows the viscosity to be estimated based on (Formula 2) (S203). The viscosity estimated by the viscosity estimation unit 303 is then output from the output unit 304 (S204). In this way, the viscosity estimation device 100 can estimate the viscosity based on the "MFR", "temperature", and "shear rate".

By operating the viscosity estimation device 100 in the above manner, the viscosity estimation method of this embodiment is realized.

<Viscosity estimation program>
The viscosity estimation method performed by the above-described viscosity estimation device 100 can be realized by a viscosity estimation program that causes a computer to execute a viscosity estimation process.

For example, in the viscosity estimation device 100 consisting of a computer shown in FIG. 2, the viscosity estimation program of this embodiment can be introduced as one of the programs 202 stored in the hard disk drive 112. Then, by having the computer that is the viscosity estimation device 100 execute this viscosity estimation program, the viscosity estimation method of this embodiment can be realized.

The viscosity estimation program that causes a computer to execute each process for creating data related to the viscosity estimation process can be recorded on a computer-readable recording medium and distributed. Recording media include, for example, magnetic storage media such as hard disks and flexible disks, optical storage media such as CD-ROMs and DVD-ROMs, and hardware devices such as non-volatile memory such as ROMs and EEPROMs.

<Further improvements to the constant determination section>
The constant determination unit 302 determines the constants included in (Equation 2) by training a neural network using a group of known data as training data, with inputs being “MFR”, temperature, and shear rate, and output being viscosity.

There are several methods for determining the constants included in (Equation 2) using machine learning with a neural network. However, in this embodiment, the inventor has newly discovered that the constants may not be determined stably using machine learning due to the unique nature of the known data group used as training data. In other words, the inventor has newly discovered that even if the number of data can be secured in a known data group, the small number of different "MFR" types means that it may not be possible to stably determine the constants using machine learning. Based on this discovery, the inventor has tried various methods and found the most stable method for determining constants using machine learning, which will be described below.

<<Configuration of Viscosity Estimation Device>>
FIG. 8 is a functional block diagram of the viscosity estimation device.

Since FIG. 8 and FIG. 3 have almost the same configuration, the following description will focus on the additional points added in FIG. 8. In FIG. 8, the viscosity estimation device 100 has a classification unit 306.

The classification unit 306 has a function of classifying the known data group into N different "MFRs" (N is a natural number). That is, the classification unit 306 is configured to classify the known data group into N types of classified data groups. For example, the classification unit 306 is configured to classify the known data group 400 shown in FIG. 4 into a first classified data group 400A, a second classified data group 400B, and a third classified data group 400C.

Here, the first classification data group 400A is composed of data groups with the same "MFR" of "17.6527" but different "temperature", "shear rate", and "viscosity". Similarly, the second classification data group 400B is a data group with the same "MFR" of "15.8989" but different "temperature", "shear rate", and "viscosity", and the third classification data group 400C is a data group with the same "MFR" of "3.4669" but different "temperature", "shear rate", and "viscosity".

The constant determination unit 302 includes a preliminary constant determination unit 307 and an average value calculation unit 308.

The preliminary constant determination unit 307 is configured to determine multiple preliminary constants included in the following (Formula 3) for each of the N different "MFRs" by training a neural network that uses each of the N types of classification data groups classified by the classification unit 306 as training data and has "MFR", temperature, and shear rate as inputs and viscosity as output.

here,
η: viscosity η _i : preliminary constant dγ/dt: shear rate n _i : preliminary constant α _i : preliminary constant T: temperature For example, in the preliminary constant determination unit 307, a neural network is trained so that the error between the preliminary constants included in (Formula 3) and the classification data group, which is the teacher data, is minimized for each "MFR." Here, an example of determining the preliminary constants will be described.

Taking the logarithm of (Equation 3) gives us the following (Equation 4).

In this (Formula 4), multiple regression analysis (Lasso regression or Ridge regression) is performed using machine learning with "ln(dγ/dt)" and "T" as variables. This allows the preliminary constant determination unit 307 to determine preliminary constants (η _i , n _i , α _i ) "i=1...N" for each of N different "MFRs."

The average value calculation unit 308 is configured to calculate an average value for some of the preliminary constants among the plurality of preliminary constants determined for each of the N different "MFRs." Specifically, the average value calculation unit 308 is configured to calculate an average value "<α>" for the preliminary constants "α _i " based on the following (Formula 5), and to calculate an average value "<n>" for the preliminary constants "n _i " based on the following (Formula 6).

The constant determination unit 302 is configured to determine the constants included in (Equation 2) by training a neural network that uses the above-mentioned average value as an initial value and the group of known data as training data, with the inputs being "MFR", temperature, and shear rate, and the output being viscosity.

For example, if we take the logarithm of (Formula 2), we get the following (Formula 7).

In this (Formula 7), the above-mentioned average values are substituted as the initial values for "n" and "α", and then a neural network is trained using a group of known data as training data, with "MFR", temperature, and shear rate as inputs and viscosity as output, to determine the constants contained in (Formula 7).

<<Operation of Viscosity Estimation Device>>
FIG. 9 is a flowchart illustrating the operation of the viscosity estimation device.

In FIG. 9, first, the input unit 301 inputs a known data group (S301). The known data group input to the input unit 301 is then stored in the data storage unit 305. Next, the classification unit 306 classifies the known data group stored in the data storage unit 305 into classified data groups for each of the same "MFR" (S302). For example, if there are N types of "MFR" in the known data group, the classification unit 306 classifies the known data group into N types of classified data groups. The classified data groups classified by the classification unit 306 are then also stored in the data storage unit 305.

Next, the viscosity estimation device 100 sets "N=1" (S303), and then the preliminary constant determination unit 307 determines the preliminary constants included in (Equation 3) based on the Nth classification data (S304). Specifically, the preliminary constant determination unit 307 determines the preliminary constants included in (Equation 3) by training a neural network with "temperature" and "shear rate" as inputs and "viscosity" as output, using the Nth classification data as teacher data.

Then, the viscosity estimation device 100 determines whether "N = Nmax" (S305).

Here, if the viscosity estimation device 100 determines that "N = Nmax" is not true, it sets "N = N + 1" (S306) and repeats S204. This repetition continues until "N = Nmax". After that, if the viscosity estimation device 100 determines that "N = Nmax" is true, the average value calculation unit 308 calculates the average value for some of the multiple preliminary constants determined for each of the N different "MFRs" (S307).

Next, the constant determination unit 302 determines the constants included in (Formula 2) by training a neural network with the above-mentioned average value as the initial value and the group of known data as training data, with the input being "MFR", temperature, and shear rate, and the output being viscosity (S308). This allows the determined constants to be substituted into (Formula 2) to ultimately determine the relational equation (Formula 2) relating viscosity to "MFR" (S309).

<Verification of effectiveness>
According to this embodiment, the viscosity can be estimated based on the "MFR".

Below, we will explain the verification results that show that the viscosity estimated in this embodiment matches the actually measured viscosity with high accuracy. Specifically, we used thermoplastic polyurethane as the resin material to verify the usefulness of the technical idea in this embodiment, and we will explain the verification results.

Four types of thermoplastic polyurethanes were prepared, all of the same grade but from different production lots. The "MFR" of each type of thermoplastic polyurethane was then measured at a temperature of 200°C and a load of 5 kg using a melt indexer (manufactured by Toyo Seiki Co., Ltd.).

The viscosity was measured using a Capillograph (manufactured by Toyo Seiki) at temperatures ranging from 185°C to 210°C and shear rates ranging from 6.08 to 6080 (1/s). A total of 60 measurements were taken.

Of the four types of samples from different production lots, the constants in (Formula 2) were determined by incorporating "MFR" using three types of sample data as a known data group. In other words, the constants in (Formula 2) were determined by machine learning using the three types of sample data as training data.

Figure 10 is a graph showing the results of a comparison between the measured viscosity values of samples not used to determine the constants and the estimated viscosity values estimated using (Equation 2) that takes into account "MFR." In Figure 10, the vertical axis shows the measured viscosity values, while the horizontal axis shows the estimated viscosity values.

The correlation coefficient between the measured viscosity value and the estimated viscosity value calculated from FIG. 10 is 0.9893, which is almost 1. Considering that the closer the correlation coefficient is to 1, the smaller the error between the measured viscosity value and the estimated viscosity value, the verification results in FIG. 10 support the idea that the viscosity estimated in this embodiment matches the measured viscosity with high accuracy.

Next, Figure 11 is a table showing the results of a comparison between a technology that actually measures viscosity (comparative example) and the technical idea of this embodiment, which estimates viscosity using "MFR."

As shown in Figure 11, the comparative example shows the measurement time when a capillary rheometer is used to collect data for a total of 15 points, with three temperature levels and five shear rate levels. As a result, the viscosity measurement time required for the comparative example is approximately 3 to 6 hours, and approximately 150 g of resin material is used. Based on the results shown in Figure 11, it can be seen that the comparative example would be difficult to use on a manufacturing process line for molded products, as the viscosity measurement time is long.

In contrast, the column for this embodiment in FIG. 11 shows the time required to estimate the viscosity based on "MFR" and the amount of resin material used. In FIG. 11, according to this embodiment, when "MFR" is measured at one temperature level and N=3 (N is the number of types), it takes about one hour. Also, if "MFR" is known, there is no need to measure it, and it only takes a few seconds to calculate it. Also, when "MFR" is measured, the amount of resin material used is 150 g, which is the same as the comparative example. Based on the results shown in FIG. 11, this embodiment can estimate the viscosity in a short time, and since the viscosity estimation accuracy is high (see FIG. 10), it can be used in the manufacturing process line of molded products.

Therefore, this embodiment overturns the conventional understanding that it is difficult and time-consuming to measure viscosity, and therefore difficult to adjust molding conditions based on viscosity measurements. Furthermore, the technical idea of making it possible to easily grasp viscosity by making good use of "MFR" can be said to have important technical significance in that it makes it easier to adjust molding conditions to absorb viscosity changes in the manufacturing process line of molded products.

The manufacturing process for a molded product that utilizes the technical concept of this embodiment is described below.

<Manufacturing method of molded products>
FIG. 12 is a flow chart showing the flow of a manufacturing method for a molded product incorporating the technical idea of this embodiment in which viscosity is estimated based on "MFR."

In FIG. 12, first, the viscosity is estimated based on the "MFR" using the technical idea of this embodiment (S401). As a result, the viscosity can be estimated. Then, based on the estimated viscosity, the molding conditions are adjusted (S402). Specifically, by estimating the viscosity, the viscosity change can be grasped, and the molding conditions such as temperature and shear rate are adjusted to absorb this viscosity change. In particular, it is possible to grasp from experience how to adjust the molding conditions in response to the viscosity change. Therefore, based on this experience, the adjustment of the molding conditions in response to the viscosity change can be automated by a program. If this automation can be realized, a configuration can be realized in which the molding conditions are automatically output when the estimated viscosity is input. Then, the resin is molded under the adjusted molding conditions (S403). As a result, according to the manufacturing method of the molded product in this embodiment, the molding conditions such as the temperature and shear rate can be optimized each time according to the viscosity change of the resin material. In other words, according to the manufacturing method of the molded product in this embodiment, the molding conditions can be adjusted to absorb the viscosity change, and therefore, a remarkable effect can be obtained in that the quality of the molded product can be maintained.

The invention made by the inventor has been specifically described above based on the embodiment thereof, but it goes without saying that the invention is not limited to the above embodiment and can be modified in various ways without departing from the gist of the invention.

Resin materials include plastic materials and rubber materials, and there is no limit to the type of resin material to which the technical ideas of this embodiment can be applied as long as the "MFR" can be obtained. In other words, the technical ideas of this embodiment can be widely applied to resin materials whose quality is affected by viscosity changes and for which an "MFR" can be obtained.

REFERENCE SIGNS LIST 1 Barrel 2 Resin 3 Piston head 4 Piston 5 Weight 6 Die 10 Measuring device 100 Viscosity estimation device 101 CPU
102 ROM
103 RAM
104 Display 105 Keyboard 106 Mouse 107 Communication board 108 Removable disk device 109 CD/DVD-ROM device 110 Printer 111 Scanner 112 Hard disk device 113 Bus 201 Operating system 202 Program group 203 File group 301 Input section 302 Constant determination section 303 Viscosity estimation section 304 Output section 305 Data storage section 306 Classification section 307 Preliminary constant determination section 308 Average value calculation section 400 Known data group 400A First classified data group 400B Second classified data group 400C Third classified data group

Claims (7)

A viscosity estimation device configured to estimate a viscosity of a resin, comprising:
The viscosity estimation device includes a viscosity estimation unit that estimates the viscosity based on a melt flow rate,
The viscosity estimation unit estimates the viscosity based on Equation 2 shown below .

here,
η: Viscosity
C: constant
MFR: Melt flow rate
dγ/dt: shear rate
n: constant
α: constant
T: Temperature 2. The viscosity estimation device according to claim 1 ,
The viscosity estimation device comprises:
a known data group input unit for inputting a known data group that associates a combination of a melt flow rate, a resin temperature, a shear rate, and a viscosity, the known data group including the melt flow rate, the temperature, the shear rate, and the viscosity;
a constant determination unit that determines a constant included in the formula 2 by learning a neural network using the known data group as teacher data, the input being the melt flow rate, the temperature, and the shear rate, and the output being the viscosity;
having
The viscosity estimation unit estimates the viscosity from a combination of the melt flow rate, the temperature, and the shear rate, the correspondence between which is unknown and the viscosity, based on the formula 2 into which the constant determined by the constant determination unit is substituted. 3. The viscosity estimation device according to claim 2 ,
The known data group is composed of N types of classified data groups classified according to N different melt flow rates (N is a natural number),
The constant determination unit is
a preliminary constant determination unit that determines a plurality of preliminary constants included in the following formula 3 for each of the N different melt flow rates by learning a neural network that uses each of the N types of classification data groups as teacher data and has the melt flow rate, the temperature, and the shear rate as inputs and the viscosity as output;
An average value calculation unit that calculates an average value for some of the plurality of preliminary constants determined for each of the N different melt flow rates;
having
The constant determination unit determines the constants included in Equation 2 by training a neural network using the average value as an initial value and the group of known data as training data, the neural network having the melt flow rate, the temperature, and the shear rate as inputs and the viscosity as an output.

here,
η: viscosity η _i : preliminary constant dγ/dt: shear rate n _i : preliminary constant α _i : preliminary constant T: temperature A viscosity estimation method for estimating a viscosity of a resin, comprising the steps of:
The viscosity estimation method includes a viscosity estimation step of estimating the viscosity based on a melt flow rate,
The viscosity estimation method, in which the viscosity estimation step estimates the viscosity based on Equation 2 shown below .

here,
η: Viscosity
C: constant
MFR: Melt flow rate
dγ/dt: shear rate
n: constant
α: constant
T: Temperature A viscosity estimation program for causing a computer to execute a process for estimating a viscosity of a resin,
the viscosity estimation program includes a viscosity estimation process for estimating the viscosity based on a melt flow rate,
The viscosity estimation process estimates the viscosity based on the following formula 2 .

here,
η: Viscosity
C: constant
MFR: Melt flow rate
dγ/dt: shear rate
n: constant
α: constant
T: Temperature A computer-readable recording medium having the program according to claim 5 recorded thereon. A viscosity estimation step of estimating a viscosity of the resin based on a melt flow rate;
a molding condition adjusting step of adjusting molding conditions based on the estimated viscosity;
a molding step of molding the resin under the adjusted molding conditions;
A method for manufacturing a molded product, comprising:
The method for manufacturing a molded product, wherein the viscosity estimation step estimates the viscosity based on Equation 2 shown below.

here,
η: Viscosity
C: constant
MFR: Melt flow rate
dγ/dt: shear rate
n: constant
α: constant
T: Temperature

Images