WO1996013370A1 - Procede d'analyse des facteurs influençant la qualite de produit obtenu d'une machine de moulage par injection, et procede d'ajustement des conditions de moulage - Google Patents
Procede d'analyse des facteurs influençant la qualite de produit obtenu d'une machine de moulage par injection, et procede d'ajustement des conditions de moulage Download PDFInfo
- Publication number
- WO1996013370A1 WO1996013370A1 PCT/JP1995/002182 JP9502182W WO9613370A1 WO 1996013370 A1 WO1996013370 A1 WO 1996013370A1 JP 9502182 W JP9502182 W JP 9502182W WO 9613370 A1 WO9613370 A1 WO 9613370A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- conditions
- data
- condition
- injection molding
- molding
- Prior art date
Links
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B29—WORKING OF PLASTICS; WORKING OF SUBSTANCES IN A PLASTIC STATE IN GENERAL
- B29C—SHAPING OR JOINING OF PLASTICS; SHAPING OF MATERIAL IN A PLASTIC STATE, NOT OTHERWISE PROVIDED FOR; AFTER-TREATMENT OF THE SHAPED PRODUCTS, e.g. REPAIRING
- B29C45/00—Injection moulding, i.e. forcing the required volume of moulding material through a nozzle into a closed mould; Apparatus therefor
- B29C45/17—Component parts, details or accessories; Auxiliary operations
- B29C45/76—Measuring, controlling or regulating
- B29C45/7686—Measuring, controlling or regulating the ejected articles, e.g. weight control
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B29—WORKING OF PLASTICS; WORKING OF SUBSTANCES IN A PLASTIC STATE IN GENERAL
- B29C—SHAPING OR JOINING OF PLASTICS; SHAPING OF MATERIAL IN A PLASTIC STATE, NOT OTHERWISE PROVIDED FOR; AFTER-TREATMENT OF THE SHAPED PRODUCTS, e.g. REPAIRING
- B29C45/00—Injection moulding, i.e. forcing the required volume of moulding material through a nozzle into a closed mould; Apparatus therefor
- B29C45/17—Component parts, details or accessories; Auxiliary operations
- B29C45/76—Measuring, controlling or regulating
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B29—WORKING OF PLASTICS; WORKING OF SUBSTANCES IN A PLASTIC STATE IN GENERAL
- B29C—SHAPING OR JOINING OF PLASTICS; SHAPING OF MATERIAL IN A PLASTIC STATE, NOT OTHERWISE PROVIDED FOR; AFTER-TREATMENT OF THE SHAPED PRODUCTS, e.g. REPAIRING
- B29C45/00—Injection moulding, i.e. forcing the required volume of moulding material through a nozzle into a closed mould; Apparatus therefor
- B29C45/17—Component parts, details or accessories; Auxiliary operations
- B29C45/76—Measuring, controlling or regulating
- B29C45/766—Measuring, controlling or regulating the setting or resetting of moulding conditions, e.g. before starting a cycle
Definitions
- the present invention relates to a method for analyzing a factor affecting quality of a product molded by an injection molding machine, and a molding condition adjusting method using the analysis method.
- Data on various conditions related to the injection molding operation such as data on the injection molding machine for each molding cycle, including the injection time and weighing time, and data on peripheral equipment such as the drying temperature of the resin. — Used to determine the quality of products by detecting evening and room temperature, etc., and is used as a criterion for optimizing molding conditions.
- Molding cycle Preliminary work is carried out when detecting the causal relationship between data fluctuations of various conditions and the occurrence of defective products, for example, by detecting each data and graphing these data for each type. After that, it is necessary to select the data that is appropriate for the quality judgment of the product from these data, but this selection work itself contains a lot of operator's subjectivity. However, this is not always sufficient information. However, since the number of conditions to determine whether or not to employ the data as a data discrimination for pass / fail is several tens, only preliminary work is required to select appropriate conditions. However, considerable work time is required.
- the accuracy of pass / fail judgment is improved by selecting multiple types of data as judgment data for pass / fail judgment.However, two data sets with similar characteristics were selected. In such a case, the accuracy of the pass / fail judgment may not always be improved. For example, a change in one data may be dependent on a change in the other data, and in such a case, it is sufficient to detect only one of the data, and If you want to make a strict pass / fail judgment, you need to select data that is not dependent on these data and that has other conditions that are closely related to product quality. There is.
- the quality of products according to the conventional technology is determined only by the relationship between the individual fluctuations of data for each condition and the failure of the corresponding products.
- the data for pass / fail judgment is selected from the viewpoint, and the effects of various conditions on the product are not sufficiently considered. It was difficult to make a selection itself, and the reliability of the discrimination results was poor.
- the injection time becomes longer than the allowable range, the temperature of the injection cylinder decreases, etc., and the viscosity resistance of the resin increases, or if the weighing is excessive. And it is difficult to move the screw to the set position.
- the time has increased, etc., as described above, since the conditions related to the injection molding operation fluctuate in a complicated manner, it is necessary to accurately understand the cause. Is very difficult.
- the injection time is shortened to be within the allowable range by increasing the set value of the injection pressure to increase the moving speed of the screw. Although it may be possible, such a short-circuit operation is generally inadequate and does not eliminate the problem.
- An object of the present invention is to improve the disadvantages of the above-described conventional technology, to accurately select data useful for determining the quality of a product, and to easily set molding conditions for eliminating molding abnormalities.
- Judgment The purpose of the present invention is to provide a method for analyzing the influence factors of product quality in injection molding, which can perform appropriate adjustment work.
- data of various conditions related to the injection molding operation for example, data of injection time, measuring time, room temperature, etc. are detected for each molding cycle, and are used as data strings of each condition.
- Multivariate analysis such as principal component analysis, cluster analysis, and factor analysis, that classifies the above conditions into relevant groups using each data sequence, was performed. Analyze the factors that affect product quality by grouping similar conditions together.
- principal component analysis all correlation coefficients indicating the correlation of data fluctuations between different conditions are obtained, a correlation matrix is created, and principal component analysis is performed on this correlation matrix.
- the eigenvectors of the two principal components are determined for each condition, and the eigenvectors of the two principal components are plotted on each axis of the rectangular coordinate system. The points corresponding to the above-mentioned conditions are plotted.
- the conditions relating to the injection molding operation are adjusted to stabilize the data under one condition.
- the data of all other conditions in the same group is to be stabilized.
- one of the data is stabilized. If this can be done, it is safe to assume that all other data in the same group can be stabilized. Therefore, even if the condition for the pass / fail judgment is data of a type for which it is difficult to determine the cause of the abnormality, the cause of the abnormality can be found in the same group.
- adjusting the molding conditions to adjust the other conditions may make it easier to eliminate molding abnormalities. You. Also, if there is another group at the origin symmetric position on the orthogonal coordinate system for the group containing the data in which an abnormality was detected in the pass / fail determination, the group at the origin symmetric position The condition means that there is a negative correlation with the group containing the data in which the abnormality was detected. Therefore, if the molding conditions for adjusting at least one condition included in another group located at the position of symmetry with the origin are known, by adjusting the molding conditions, the same as above can be achieved. The problem of abnormal molding is eliminated.
- the work of detecting the data of the conditions and storing the data string and the work of blotting the points corresponding to each of the above conditions are performed. Adjust the molding conditions until each point is grouped around the origin of the Cartesian coordinate system, or calculate the standard deviation or standard deviation of all or specific conditions. Depending on whether the value divided by the average value is equal to or less than the set value, or if the contribution ratio of all principal components or a specific principal component converges to the set value or less, the condition is set. Adjust the molding conditions until the data converge.
- FIG. 1 is a block diagram schematically showing a main part of a control device that drives and controls an injection molding machine that carries out the method of the present invention, and an arithmetic device connected to the control device.
- FIG. 2 is a conceptual diagram showing an example of a data file provided in the arithmetic unit
- FIG. 3 is a flowchart showing the outline of the scatter diagram creation processing by the arithmetic unit.
- Figure 4 shows an example in which a series of large vibration data is displayed in a graph.
- Figure 5 shows an example of a graph of a data sequence from which vibration has been removed.
- Fig. 6 shows an example of the correlation matrix obtained from the data file.
- Fig. 7 shows the distance between conditions with strong correlation close, and the distance between conditions with weak correlation far.
- Fig. 8 is a conceptual diagram showing an example of the correlation variation
- Fig. 9 is the variation and its factors.
- FIG. 10 is a conceptual diagram showing an example of a first principal component axis and a second principal component axis.
- Fig. 11 is an example of a scatter plot showing the first principal component axis as the horizontal axis and the second principal component axis as the vertical axis.
- Figure 12 is an example of a scatter diagram divided into groups.
- FIG. 1 shows a main part of a control device 100 that drives and controls an injection molding machine that implements a method for analyzing a product quality influence factor and a method for adjusting a molding condition of the present invention, and is connected to the control device 100.
- FIG. 1 shows a main part of a control device 100 that drives and controls an injection molding machine that implements a method for analyzing a product quality influence factor and a method for adjusting a molding condition of the present invention, and is connected to the control device 100.
- the injection molding machine in the embodiment is a normal electric injection molding machine that drives each axis such as a mold clamping shaft, a screw shaft rotation shaft, a jewel shaft and an injection shaft by a servomotor.
- each axis such as a mold clamping shaft, a screw shaft rotation shaft, a jewel shaft and an injection shaft by a servomotor.
- the hardware configuration of this type of injection molding machine itself has already been known, the description thereof is omitted here.
- the control device 100 that drives and controls the injection molding machine is a CNC CPU 116 that is a micro-processor for numerical control, and a microphone for a programmable machine controller. It has a CPU for PMC as a low-speed processor, a servo CPU for microcontroller for servo control, and a CPU for pressure monitor. By selecting mutual input and output via 112, information can be transmitted between the microprocessors.
- the PMC CPU 110 has R0M104 that stores a sequence program that controls the sequence operation of the injection molding machine, and RAM that is used for temporary storage of operation data. 0 5 is connected, and the CPU 1 16 for CNC stores R 0 Ml 18 that stores programs and the like that control the entire injection molding machine and RAMI 1 that is used for temporary storage of calculation data, etc. 9 is connected.
- the servo CPU 111 and the pressure monitor CPU 109 each have a ROM 106 that stores a control program dedicated to servo control and a RAMI 07 that is used for temporary storage of data.
- a ROM 102 storing a control program relating to a sampling process and the like of the pressure overnight is connected to a RAMI 03 used for temporary storage of the data overnight.
- the servo CPU 11 1 has, based on instructions from the CPU 11 1, such as mold clamping, injection, screw rotation, and agitation of an injection molding machine.
- a servo chamber for driving the servo mode of each axis is connected.
- FIG. 1 shows only one servomotor 200 and a servomotor 108 for driving the servomotor.
- Each of the outputs from the servo motors 200 installed on the servo motors 200 of each axis is fed back to the servo CPU 111, and the feedback clocks of the panoramic cod
- the current position and moving speed of each axis calculated by the servo CPU 1111 are updated and stored in the current position storage register and current speed storage register of RAMI 09 sequentially. It is becoming.
- the CPU 109 for the pressure monitor passes through the AZD converter 101 and a pressure detector provided on the side of the injection molding machine (not shown) to maintain the injection holding pressure during the injection molding operation. Perform the back pressure sampling process.
- the input / output circuit 113 receives signals from limit switches and operation panels provided in each part of the injection molding machine and transmits various commands to peripheral devices of the injection molding machine. Simplicity Input / output circuit.
- the manual data input device 120 with display is connected to the bus 112 via the CRT display circuit 117 to select the monitor display screen and function menu, and to select various data. It is capable of performing input operations and the like, and is provided with a numeric keypad for inputting numerical data and various function keys.
- Non-volatile memory 114 is used for various molding conditions related to injection molding operations, such as injection speed, injection holding pressure switching position, holding pressure, screw speed for metering and kneading, nozzles and injection.
- molding that stores the set values of various conditions related to the injection molding operation that can be actively controlled, such as the temperature of each part of the cylinder, as well as various parameters and macro variables.
- Memory for storing data.
- the PMC CPU 110 controls the sequence of the entire injection molding machine based on the sequence program in the ROM 104, while the CNC CPU 116 controls the sequence.
- pulse distribution is performed to the servomotors of each axis, and the servo CPU 11 1 is based on the movement command pulse-distributed to each $ ⁇ , the feedback signal of the position detected by a detector such as a pulse coder, and the feedback signal of the speed.
- servo control such as position loop control, speed loop control, and current loop control is performed in the same manner as in the past, and so-called digital servo control is performed as a whole.
- the temperature of the nozzle and the temperature of each part of the injection cylinder, the temperature of the lower part of the hopper, and the cycle time measured by the timing means provided in the controller 10 The data such as the required time and the time required for weighing are used as part of the detection data of the conditions related to the injection molding operation for each molding cycle during the continuous molding operation, and the communication interface -Transferred to the arithmetic unit 400 via the interface 115.
- the arithmetic unit 400 includes a microprocessor 405, a hard disk drive 402, a keyboard 403, a display device 406, and these. It is a normal personal computer consisting of a bus 404 etc. connecting the various types of data described above, which are transferred from the controller 100 to the communication interface. Received via the disk 410 and stored in the data storage of the hard disk drive 402 in the sequence channel. You. In addition, in synchronization with the reception of data from the control device 100, data from an external detector 300, such as a room temperature meter, is read and data corresponding to each molding cycle is read. The data is stored in the data file. The data from the external detector 300 is also part of the conditions associated with the injection molding operation ⁇ »
- Conditions related to the injection molding operation include the injection speed, the injection / holding pressure switching position, the holding pressure, the screw rotation speed for metering and kneading, the nozzle and the temperature of each part of the injection cylinder, etc. Some are detected on the injection molding machine side as described above, and others are detected by an external detector 300 such as room temperature. Injection molding such as injection speed, injection holding pressure switching position, holding pressure, screw rotation speed for metering and kneading, nozzle and temperature of each part of injection cylinder, etc. Some of the conditions related to the work can be adjusted by changing the setting of the molding conditions in the nonvolatile memory 114.
- Injection speed, injection holding pressure switching position, holding pressure, screw rotation for metering and kneading, nozzle and temperature of each part of injection cylinder are determined by molding conditions and various conditions. Although it is defined repeatedly, it indicates the setting value of the nonvolatile memory 114 when it is called a molding condition, and when it is called various conditions, it actually indicates It shall indicate the data detected by the pressure detector, thermocouple, etc.
- the conditions that can be actually detected are those detected by the detector on the injection molding machine side and those detected by the external detector.
- the file is being created.
- Figure 2 is a conceptual diagram showing an example of a data file.
- Each of the numbers 1 to 22 in the column direction is an address value indicating the name of various conditions. For example, condition "1" is cycle time, condition "2" is injection time, and condition "3".
- condition “4" is the minimum cushion amount
- condition "5" is the weighing complete position
- condition "13" is the temperature in the middle of the injection cylinder
- condition “22” is room temperature or the like.
- the numbers 1 to n in the row direction are address values indicating the number of shots from the start of sampling, and correspondingly, the detector on the injection molding machine side
- the external detector 300 detects the data of each condition (each shot) detected in each molding cycle (sequence). It is stored as a data string in the shal.
- the flowchart in FIG. 3 shows a scatter diagram creation process for creating a scatter diagram for easily displaying the correlation between the change characteristics of the detection data of various conditions and the like.
- the program is stored in the hard disk 402 in advance.
- the microprocessor 405 that started the scatter diagram creation process first read the data from the data file and temporarily stored it in the work area of the hard disk 402.
- step S1 the standard deviation and the average value are obtained for each condition, and the value Q obtained by dividing the standard deviation by the average value is obtained (step S2). Then, it is determined whether the value Q of each condition is less than or equal to the set evaluation value ⁇ and the variance of each condition is reduced and converged (step S3), and the setting is performed for all conditions. If the value is equal to or smaller than the evaluation value ⁇ , the analysis process by the scatter diagram creation process ends. Instead of obtaining the value Q obtained by dividing the above standard deviation by the average value for all the conditions, this value Q is obtained only for the preset conditions, and only the set conditions are determined in step S3. Processing may be performed.
- whether or not the variance of each condition has converged was determined by the value Q obtained by dividing the standard deviation by the average value. Only the standard deviation may be obtained. The standard deviation may be determined based on whether or not the standard deviation is smaller than ⁇ '(or not.
- step S 4 it is determined whether or not the vibration component removal command is set (step S 4), only when the vibration component elimination command is set, the data averaging process shown in step S 5 is performed, and the data file in the work area is processed. Update. Note that the vibration component removal command is issued by the operator via the keyboard 403 to the arithmetic unit. This is manually set to 400, and if it is before the start of the scatter diagram creation processing, the vibration component removal command can be arbitrarily set and canceled by the manual operation described above. You can force.
- the data averaging process shown in step S5 has a certain change pattern when the flow of the entire data sequence is viewed in chronological order, but when comparing individual data, Compared with the data trains under other conditions, the data trains under various conditions including many vibrational components, such as noticeable data variations between molding cycles, were compared. This is a data smoothing process that is effective when calculating numbers.
- Figure 4 shows the sampling data with the minimum cushion amount as an example of various conditions including the vibration component.
- the value on the horizontal axis of the graph in Fig. 4 indicates the number of shots, in short, the number of shots and address of the data file as shown in Fig. 2.
- the vertical axis indicates the value of data for each shot (molding cycle).
- the data sequence of various conditions with many vibration components and the data sequence of other conditions, or the data sequence of various conditions with many vibration components are directly connected to each other. If the correlation coefficient is determined by comparison, it may not be possible to obtain a correlation coefficient that correctly indicates the correlation between the data strings, so that it is necessary to perform data averaging processing. You.
- the simplest method of data averaging is to read and accumulate M data sequentially from the beginning of the shot number address of the data file as shown in Fig. 2. Divide the value by M to find the average of that section, and use that value as the representative value of the section Method. More specifically, the data of the representative value is written in the data finder by associating the value of the representative value with each of the addresses in the section. The same processing is applied to all of the other data rows, regardless of the amount of vibration components, even if the amount of vibration components is small.
- the number n of the first data is not always divisible by M, so if n ZM has a surplus, this surplus
- no, six of data to the pair described above of Yo you Do not handle the line Do not I was or Nodea Ru c to compress the data only remaining, and at the same time, was Ji raw I'm in the data compression
- the value of the total sampling number n is updated to INT (n / M) according to the fluctuation of the number of data (INT is converted to an integer by truncating the decimal part).
- Another method of averaging data is to generate a free-running curve that passes through the center of vibration of data containing fluctuations (see Fig. 4), for example (see Fig. 4).
- a free-running curve that passes through the center of vibration of data containing fluctuations (see Fig. 4), for example (see Fig. 4).
- all the data values corresponding to the positions of each shot are obtained for each shot, and the data file
- the micro processor 405 which has completed the discriminating process of step S4 or the averaging process of the data of step S5 described above, outputs the hard disk 40, 2 sets of data strings of various conditions are read from the data file stored at this point in the work area 2 and the correlation coefficient r is calculated for all combinations.
- the total number of combinations is (2 2 X 2 2 — 2 2) 2 and the correlation coefficient of the condition "y" for the condition "X” is the same as the correlation coefficient of the condition "X” for the condition "Y”.
- the number of correlation coefficients r to be calculated is 46 2/2, which is 231 above or below the diagonal line in Figure 6 which is required.
- the relative number r can be obtained by the operation of the following equation (1).
- n Number of data common to each data column
- FIG. 7 is based on the correlation matrix of FIG. 6, so that conditions with strong correlation are close to each other, and conditions with weak correlation are far from each other.
- This is a conceptual diagram in which the distance is inversely proportional to the relative number r, and the points indicating the names of the conditions are plotted and plotted.
- the point corresponding to the cycle time of the condition “1” is “1”
- the point corresponding to the injection time of the condition “2” is “2”. Therefore, the correlation between all the conditions can be easily understood from Fig. 7, but in reality, this question is composed of a two-dimensional array corresponding to the number of conditions. Therefore, if the number of conditions exceeds three, it is of course impossible for humans to visually recognize the condition, and even to display a blot.
- such multidimensional data can be reduced to three or less, and more desirably two-dimensional data, and easily visualized. You need to be able to recognize it.
- the variation in the point indicating the name of the condition is specified as shown in Fig. 8, If it has a direction, it can be specified that the point variation is caused by some factor that is not known per se. If the variation is divided into sections A, B, and C along the above-mentioned specific direction as shown in Fig. 9, one that is negatively affected by that factor ( It is predicted that there will be some that will be affected in the positive direction (section A), one that will be affected in the positive direction (section C), and one that will be hardly affected in either direction (section B).
- the following principal component analysis processing is performed.
- This contribution rate is calculated by summing the variance of each principal component as the sum of the variances of all principal components (first to 22nd principal components). Divide by to find. Then, it is determined whether the contribution ratio C 1 of the first principal component having the largest contribution ratio is smaller than the set value 9 (for example, about 0.1), and the contribution ratio is made uniform. It is determined whether or not the operation has been performed (step S9).
- the analysis process ends. It should be noted that the target value of the principal component and the contribution ratio of the principal component is set in advance and not at the stage where the adjustment of the molding conditions has been advanced, in particular, rather than the contribution ratio of the first principal component. This analysis processing may be terminated when the contribution ratio of the principal component becomes small. Next, if the contribution ratio of the first principal component is larger than the set value, the eigenvectors of the first principal component and the second principal component are obtained (step S10).
- the eigenvectors indicate the strength and direction of the principal component, their magnitude is normalized so that the force is “1”, and the new coordinate system unit represented by the principal component is used. Become a vector.
- This eigenvector indicates a numerical value indicating the influence of each condition (data) on the principal component, and each of the first principal component A to the second second principal component.
- the effect of all conditions on V is expressed by the following equation (2).
- a to V are independent of each other
- a scatter plot is obtained for the first principal component (first factor) A and the second principal component (second factor) B.
- unique vector door Le a of each of the conditions with respect to the first principal component :, b l C l, • ⁇ ⁇ , v, a unique base-click door of each condition with respect to the second principal component le a 2 , B 2 , c 2 , ⁇ , v 2 , and the orthogonal coordinate system with the first principal component axis as the horizontal axis and the second principal component axis as the vertical axis is supplied to the display device 406.
- the points with the names corresponding to the respective conditions are displayed and plotted on this rectangular coordinate system (step S11).
- conditions "1" 1 and Bro will be about the re-Gu Ruta Lee-time values that value for the first principal component axis is against the a ,, the second principal component axis to a position that Do and a 2 of Tsu and Bok, in a similar manner, conditions "2" Seki to the that Do value value pair to the first principal component axis with respect to b ,, second principal component axis and b 2 located on the exit time
- a scatter chart as shown in Fig. 11 is displayed.
- the scatter plot of Fig. 11 created in this way has a one-dimensional variation direction as shown in Figs. 8 and 9.
- the display has been extended to two dimensions. Therefore, in Fig. 11, conditions located on the right side of the coordinate origin of the rectangular coordinate system, for example, conditions "15” and "16" are strongly influenced by the first principal component in the positive direction (see Fig. 11). 9), and the conditions located on the left side of the coordinate origin, such as conditions “4” and “2 2”, are strongly affected by the first principal component in the negative direction (A in FIG. 9). And the condition located near the coordinate origin when viewed from the direction of the first principal component, for example, the conditions “10”, “19”, “21”, etc.
- the eigenvectors a 3 , b 3 , c 3 ,..., And v 3 of the third principal component are expressed together in a three-dimensional orthogonal coordinate system up to the third principal component axis.
- the accuracy (contribution ratio) of the information indicating the variation is improved, but the operator grasps the three-dimensional information blotted on the display device 406. Since it is difficult to do so, such a display is not intentionally performed in this embodiment.
- FIG. It is technically easy to display the screen as shown in Fig.
- the first principal component axis which has the largest influence, is displayed. It is advisable to select the second principal component axis and the second principal component axis for display. Even if the other two axes are selected to be displayed, at least It must be configured so that the first principal component axis and the second principal component axis can be selected for display.
- the microprocessor 405 further groups the points that are plotted in close proximity to each other, for example, by circular segmentation as shown in FIG. Then, the distinction of the attribute is displayed on the display device 406 (step S12). For example, the sum of the condition "3" belonging to group B The amount time and conditions "5" metering completion position specific base-vector and e l and against the first major component of and the unique base click preparative Le c 2 and e 2 for the second principal component are both correct approximately equal, Since the data sequence of the weighing time of the condition "3" and the weighing completion position of the condition "5" can be regarded as changing with almost equal characteristics due to fluctuations of various conditions related to the injection molding operation. is there.
- the operator performing the injection molding operation can refer to the scatter diagram (shown in the screen in duplicate) as shown in FIG. 11 and FIG. Such considerations can be made.
- the conditions in each group as shown in Fig. 12 are based on the total variation of the conditions related to the injection molding operation, that is, the first condition, which is not known per se. Due to the influence of the factors of the principal component and the second principal component, they have data strings that change with almost the same characteristics, so select two or more conditions belonging to the same group. This means that it does not make much sense to detect the data as pass / fail judgment data. For example, the weighing time of condition "3" and the weighing completion position of condition "5" belonging to the same group B show the same fluctuation characteristics due to the influence of the factors of the first and second principal components. It is meaningless to detect both as data for pass / fail judgment.
- the conditions of the different groups have different data sequences with different characteristics due to the influence of the overall variation of the conditions related to the injection molding operation, so that the different groups are different. If conditions are selected one by one from each of the sensors and each of them is detected as a data item for quality judgment, the quality of the product can be accurately judged based on relatively independent conditions. This will be possible. Also, in FIG. 12, the glove located near the coordinate origin is hardly affected by the first principal component and the second principal component, in other words, is easily affected by other principal components. is there. As can be seen from equation (2), the principal components are independent of each other. Therefore, in order to determine the quality of the product more accurately, the first principal component and the second principal component are determined.
- First principal component and second principal component, with sensitive group conditions The condition is also selected from other groups that are not easily affected by, that is, groups near the coordinate origin that are susceptible to other principal components independent of the first and second principal components. It is a law to say that. Also, by selecting any two axes from the second principal component axis from the first principal component axis and forming a two-dimensional orthogonal coordinate system by the selected two axes. If the screen as shown in Fig. 11 is switched and displayed, groups that are susceptible to the first and second principal components and independent groups Since it is possible to obtain a group or the like that is easily affected by the third principal component and the fourth principal component, it is convenient to independently select the conditions for quality judgment. Of course, when selecting conditions from each group, as described above, the ones that have large fluctuations in the actual data sequence should be selected.
- the presence or absence of product defects is determined by the detection data of the various conditions described above, as in the prior art such as JP-A-6-155550 / JP-A-6-170907. Is detected for each molding cycle, and it is detected by determining whether the value deviates from the allowable value. Therefore, the presence or absence of a molding abnormality is eventually detected as an abnormality of the item such as the condition “1”, the condition “2”, the condition “3”,...
- an abnormal item does not mean that it is not always easy to know the molding conditions to be adjusted to eliminate the abnormality, and that the type of the abnormality is controlled. If this is not possible, it cannot be adjusted directly.
- the weighing time of condition "3" is a condition determined as a result of the weighing completion position of condition "5" and other molding conditions such as the screw rotation speed and weighing back pressure. Therefore, since the injection molding machine is not under the molding conditions that can be actively controlled, the measurement time itself is set to eliminate the measurement time abnormality because the measurement time abnormality is detected. You can't do that if you change.
- the scatter diagram creation process described above is executed according to the data sequence as shown in FIG. 2 that has been detected in the molding operation up to now, and FIG. 11 and FIG. By displaying a scatter plot as shown in Fig. 7, by specifying the group containing the condition in which the abnormality was detected, it is possible to obtain a similar change characteristic. Condition can be easily found.
- the injection molding operation is again performed continuously to detect and store the condition data and the above-mentioned conditions.
- each condition is the first component. Means that it is not affected by the main component and the second main component, and as long as the product is good, it is guaranteed that the most stable molding conditions are set .
- the factors affecting the product quality are analyzed by the principal component analysis.
- other multivariate analysis methods for example, similar analysis methods such as cluster analysis and factor analysis are used. It is also possible to analyze the product quality influencing factors by analyzing the quality of the molded product and to adjust the molding conditions in the same manner.
- a group including other conditions showing a change characteristic similar to the condition in which the abnormality is detected is included. And a condition that shows a completely opposite change characteristic. It is easy to find the group that contains the error, so even if it is difficult to directly adjust the condition itself in which the error was detected, the condition that can be adjusted for either group is difficult. If there is, by adjusting this, it becomes possible to stabilize the condition in which the abnormality is detected collectively.
Landscapes
- Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Mechanical Engineering (AREA)
- Injection Moulding Of Plastics Or The Like (AREA)
Description
Claims
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
DE69529149T DE69529149T2 (de) | 1994-10-27 | 1995-10-24 | Verfahren zum analysieren von beeinflussenden faktoren in der produktqualität von spritzgiessmaschinen und verfahren zum anpassen der formbedingungen |
EP95934878A EP0737560B1 (en) | 1994-10-27 | 1995-10-24 | Method of analyzing factors influential in product quality of injection molding machine and method of adjusting molding conditions |
US08/666,370 US5815397A (en) | 1994-10-27 | 1995-10-24 | Method of analyzing factors affecting product quality and adjusting molding conditions for injection molding machine |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP28614894A JP3161921B2 (ja) | 1994-10-27 | 1994-10-27 | 製品品質影響要因解析方法と装置及び成形条件調整方法、製品良否判別項目選択方法 |
JP6/286148 | 1994-10-27 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO1996013370A1 true WO1996013370A1 (fr) | 1996-05-09 |
Family
ID=17700565
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP1995/002182 WO1996013370A1 (fr) | 1994-10-27 | 1995-10-24 | Procede d'analyse des facteurs influençant la qualite de produit obtenu d'une machine de moulage par injection, et procede d'ajustement des conditions de moulage |
Country Status (6)
Country | Link |
---|---|
US (1) | US5815397A (ja) |
EP (1) | EP0737560B1 (ja) |
JP (1) | JP3161921B2 (ja) |
KR (1) | KR0169811B1 (ja) |
DE (1) | DE69529149T2 (ja) |
WO (1) | WO1996013370A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110920010A (zh) * | 2019-10-29 | 2020-03-27 | 上海澎睿智能科技有限公司 | 基于大数据分析的注塑工艺生产方法 |
Families Citing this family (42)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
FR2750918B1 (fr) * | 1996-07-09 | 1999-04-09 | Transvalor Sa | Procede de commande et de regulation d'une presse de moulage par injection |
JP3576335B2 (ja) * | 1996-10-31 | 2004-10-13 | 松下電器産業株式会社 | プロセス加工工程の異常抽出方法及び装置 |
DE19801881C1 (de) * | 1998-01-20 | 1999-05-27 | Karl Hehl | Verfahren zur Beurteilung von Spritzteilen |
US6301515B1 (en) * | 1998-04-07 | 2001-10-09 | Tmmm, Ltd. | System for establishing optimal critical characteristics of manufactured products |
AU2000280296A1 (en) * | 2000-02-15 | 2001-08-27 | General Electric Company | Method and system for providing design tools to multiple users |
US6592354B2 (en) | 2000-12-01 | 2003-07-15 | Avalon Vision Solutions, Llc | Part-forming machine having an infrared vision inspection system and method for verifying the presence, absence and quality of molded parts therein |
US20030211188A1 (en) * | 2000-06-19 | 2003-11-13 | Kachnic Edward F. | Wireless image processing method and device therefor |
US6592355B2 (en) | 2000-12-16 | 2003-07-15 | Ovalon Vision Solutions, Llc | Part-forming machine having an in-mold integrated vision system and method therefor |
WO2001098050A1 (en) * | 2000-06-19 | 2001-12-27 | Edward Kachnic | Part forming machine integrated controller |
US20040059452A1 (en) * | 2000-08-23 | 2004-03-25 | Edward Kachnic | Sensory inspection system and method thereof |
US6549864B1 (en) | 2001-08-13 | 2003-04-15 | General Electric Company | Multivariate statistical process analysis systems and methods for the production of melt polycarbonate |
US6845278B2 (en) | 2002-08-07 | 2005-01-18 | Kimberly-Clark Worldwide, Inc. | Product attribute data mining in connection with a web converting manufacturing process |
JP3756872B2 (ja) * | 2002-11-07 | 2006-03-15 | 日精樹脂工業株式会社 | 成形品の判別条件設定方法 |
DE60326391D1 (de) * | 2002-11-07 | 2009-04-09 | Honda Motor Co Ltd | Clustern von erscheinungsbildern von objekten unter verschiedenen beleuchtungsbedingungen |
US7117050B2 (en) | 2002-11-08 | 2006-10-03 | Toshiba Kikai Kabushiki Kaisha | Management supporting apparatus, management supporting system, management supporting method, management supporting program, and a recording medium with the program recorded therein |
JP3830453B2 (ja) * | 2003-01-15 | 2006-10-04 | ファナック株式会社 | 射出成形機のモニタ装置 |
WO2005052525A2 (en) * | 2003-11-20 | 2005-06-09 | Nicol William A | Sensory system and method thereof |
JP4438745B2 (ja) * | 2003-12-18 | 2010-03-24 | 三菱電機株式会社 | 射出成形機の制御装置 |
US7534378B2 (en) * | 2004-03-03 | 2009-05-19 | Rexam Prescription Products Inc. | Plastic forming process monitoring and control |
JP2006021470A (ja) | 2004-07-09 | 2006-01-26 | Fanuc Ltd | 射出成形機のモニタリング装置及びモニタリング方法 |
JP4490863B2 (ja) * | 2005-04-27 | 2010-06-30 | 株式会社 日立東日本ソリューションズ | 品目分類支援システムおよび品目分類支援プログラム |
CN100533657C (zh) * | 2006-06-30 | 2009-08-26 | 株式会社东芝 | 警报装置 |
US8271103B2 (en) * | 2007-05-02 | 2012-09-18 | Mks Instruments, Inc. | Automated model building and model updating |
US9098766B2 (en) | 2007-12-21 | 2015-08-04 | Honda Motor Co., Ltd. | Controlled human pose estimation from depth image streams |
US9165199B2 (en) * | 2007-12-21 | 2015-10-20 | Honda Motor Co., Ltd. | Controlled human pose estimation from depth image streams |
JP5232561B2 (ja) * | 2008-07-30 | 2013-07-10 | 本田技研工業株式会社 | 品質検査方法 |
JP5232560B2 (ja) * | 2008-07-30 | 2013-07-10 | 本田技研工業株式会社 | 品質予測方法 |
US8494798B2 (en) * | 2008-09-02 | 2013-07-23 | Mks Instruments, Inc. | Automated model building and batch model building for a manufacturing process, process monitoring, and fault detection |
US9069345B2 (en) * | 2009-01-23 | 2015-06-30 | Mks Instruments, Inc. | Controlling a manufacturing process with a multivariate model |
JP5647155B2 (ja) * | 2009-02-25 | 2014-12-24 | 本田技研工業株式会社 | 内側距離形状関係を使用する身体特徴検出及び人間姿勢推定 |
JP5460355B2 (ja) * | 2010-01-27 | 2014-04-02 | ファナック株式会社 | 相関係数算出機能を有する射出成形機の制御装置 |
DE202010007655U1 (de) | 2010-06-07 | 2011-09-08 | Ulrich Seuthe | Vorrichtung zur Überwachung und Optimierung von Spritzgießprozessen |
US8855804B2 (en) | 2010-11-16 | 2014-10-07 | Mks Instruments, Inc. | Controlling a discrete-type manufacturing process with a multivariate model |
JP5648157B2 (ja) * | 2011-12-28 | 2015-01-07 | 株式会社日立ハイテクノロジーズ | 半導体製造装置 |
US9541471B2 (en) | 2012-04-06 | 2017-01-10 | Mks Instruments, Inc. | Multivariate prediction of a batch manufacturing process |
US9429939B2 (en) | 2012-04-06 | 2016-08-30 | Mks Instruments, Inc. | Multivariate monitoring of a batch manufacturing process |
US10899099B2 (en) | 2014-10-23 | 2021-01-26 | Furukawa Industrial Machinery Systems Co., Ltd. | Device and method for evaluating operating conditions of briquetting machine, briquetting machine, method for manufacturing briquette, control device of briquetting machine, control method of briquetting machine, and program |
US9475221B2 (en) | 2014-12-12 | 2016-10-25 | The Procter & Gamble Company | Method of predicing injection molding cycle time |
JP7143713B2 (ja) * | 2018-10-02 | 2022-09-29 | 東ソー株式会社 | 金属部材-ポリアリーレンスルフィド樹脂部材複合体の製造方法 |
KR20210017577A (ko) * | 2019-08-09 | 2021-02-17 | 주식회사 엘지화학 | 제조 설비 품질에 대한 정량화 진단법 |
TWI783427B (zh) * | 2020-03-31 | 2022-11-11 | 日商旭化成股份有限公司 | 支援樹脂成形之裝置、方法及程式 |
JP2021192137A (ja) * | 2020-06-05 | 2021-12-16 | アズビル株式会社 | 品質影響因子特定支援装置および方法 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS5320081A (en) * | 1976-08-04 | 1978-02-23 | Martin Sanchez Juan | Control system improvement in |
JPS57212042A (en) * | 1981-06-24 | 1982-12-27 | Toshiba Mach Co Ltd | Molding data collector for injection molding machine |
JPH02241718A (ja) * | 1989-03-15 | 1990-09-26 | Minolta Camera Co Ltd | 成形機の成形条件解析装置 |
JPH03116301A (ja) * | 1989-09-29 | 1991-05-17 | Canon Inc | 適応装置 |
Family Cites Families (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS59211103A (ja) * | 1983-05-16 | 1984-11-29 | Japan Steel Works Ltd:The | 合成樹脂用押出成形機における多変量制御方法および制御装置 |
JPS61137662A (ja) * | 1984-12-07 | 1986-06-25 | Toshiba Mach Co Ltd | ダイカスト製品の不良要因自動集録システム |
JPS6220008A (ja) * | 1985-07-19 | 1987-01-28 | Toshiba Corp | 原因推定装置 |
JPS6424719A (en) * | 1987-07-20 | 1989-01-26 | Komatsu Mfg Co Ltd | Controlling apparatus for injection molding machine |
JPH0720651B2 (ja) * | 1988-11-09 | 1995-03-08 | 東芝機械株式会社 | 射出成形機の最適成形条件設定システム |
GB8828713D0 (en) * | 1988-12-08 | 1989-01-11 | Vickers Systems Ltd | Apparatus & method for simultaneous control of two/more related variables |
JPH0435923A (ja) * | 1990-05-31 | 1992-02-06 | Komatsu Ltd | エキスパートシステムを用いた成形条件探索方法 |
JP2543793B2 (ja) * | 1991-07-23 | 1996-10-16 | 日精樹脂工業株式会社 | 射出成形機群のデ―タ収集方法及び装置 |
JP2593374B2 (ja) * | 1991-08-30 | 1997-03-26 | 日精樹脂工業株式会社 | 射出成形機の成形条件設定方法 |
EP0603244B1 (de) * | 1991-09-12 | 1997-03-19 | Engel Maschinenbau Gesellschaft m.b.H. | Verfahren zur steuerung einer maschine für die herstellung von produkten, insbesondere zur steuerung einer spritzgiessmaschine |
US5442562A (en) * | 1993-12-10 | 1995-08-15 | Eastman Kodak Company | Method of controlling a manufacturing process using multivariate analysis |
-
1994
- 1994-10-27 JP JP28614894A patent/JP3161921B2/ja not_active Expired - Fee Related
-
1995
- 1995-10-24 DE DE69529149T patent/DE69529149T2/de not_active Expired - Lifetime
- 1995-10-24 WO PCT/JP1995/002182 patent/WO1996013370A1/ja active IP Right Grant
- 1995-10-24 EP EP95934878A patent/EP0737560B1/en not_active Expired - Lifetime
- 1995-10-24 US US08/666,370 patent/US5815397A/en not_active Expired - Lifetime
- 1995-10-26 KR KR1019950037268A patent/KR0169811B1/ko not_active IP Right Cessation
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS5320081A (en) * | 1976-08-04 | 1978-02-23 | Martin Sanchez Juan | Control system improvement in |
JPS57212042A (en) * | 1981-06-24 | 1982-12-27 | Toshiba Mach Co Ltd | Molding data collector for injection molding machine |
JPH02241718A (ja) * | 1989-03-15 | 1990-09-26 | Minolta Camera Co Ltd | 成形機の成形条件解析装置 |
JPH03116301A (ja) * | 1989-09-29 | 1991-05-17 | Canon Inc | 適応装置 |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110920010A (zh) * | 2019-10-29 | 2020-03-27 | 上海澎睿智能科技有限公司 | 基于大数据分析的注塑工艺生产方法 |
CN110920010B (zh) * | 2019-10-29 | 2021-10-19 | 上海澎睿智能科技有限公司 | 基于大数据分析的注塑工艺生产方法 |
Also Published As
Publication number | Publication date |
---|---|
KR960013628A (ko) | 1996-05-22 |
DE69529149T2 (de) | 2003-04-17 |
JP3161921B2 (ja) | 2001-04-25 |
EP0737560A4 (en) | 1997-07-09 |
EP0737560B1 (en) | 2002-12-11 |
US5815397A (en) | 1998-09-29 |
EP0737560A1 (en) | 1996-10-16 |
KR0169811B1 (ko) | 1999-03-20 |
JPH08118443A (ja) | 1996-05-14 |
DE69529149D1 (de) | 2003-01-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO1996013370A1 (fr) | Procede d'analyse des facteurs influençant la qualite de produit obtenu d'une machine de moulage par injection, et procede d'ajustement des conditions de moulage | |
US20180264613A1 (en) | Abnormality detection apparatus and machine learning apparatus | |
US20150276557A1 (en) | State monitoring system, state monitoring method and medium | |
JP2767320B2 (ja) | 射出成形機の成形条件出し方法 | |
US6658319B2 (en) | Molding-condition setting method for an injection molding machine | |
JP2002525757A (ja) | 統計的プロセス管理のための多次元的方法及びシステム | |
JP2019135638A (ja) | 生産プロセス中に異常なワークピースを自動検出するためのデータ駆動方法 | |
KR20190081844A (ko) | 가상의 센서 데이터를 생성하는 방법 및 장치 | |
Wiklund | Bayesian and regression approaches to on‐line prediction of residual tool life | |
JP2016197432A (ja) | 部品の製造工程の過程を制御するための装置および方法 | |
JP6772119B2 (ja) | 演算処理装置、演算処理装置の演算方法及びプログラム | |
US11422542B2 (en) | Workpiece surface quality issues detection | |
CN113665079B (zh) | 一种塑料注塑成型工艺控制方法及系统 | |
JP6739502B2 (ja) | データ解析装置、データ解析方法、制御プログラム、および記録媒体 | |
JP3677627B2 (ja) | 製品の良否判別方法 | |
JP4571044B2 (ja) | モニタリング表示装置および表示方法 | |
Novikov et al. | Optimization of the parameters of the sewing thread cutting process | |
US10882236B2 (en) | Molding system, molding apparatus, inspection apparatus, inspection method, and program | |
JP2002079560A (ja) | 射出成形機の製品良否判別方法 | |
JP6117841B2 (ja) | 工具オフセットデータの誤入力防止機能を備えた数値制御装置 | |
JP6633403B2 (ja) | 解析対象決定装置及び解析対象決定方法 | |
CN108250342A (zh) | 一种聚全氟乙丙烯产品生产方法及系统 | |
US20230039170A1 (en) | Molding management apparatus | |
WO2022085580A9 (ja) | 成形条件設定装置及び成形条件設定方法 | |
Quellmalz et al. | Influence analysis on the model comparison performance index for servo drive control |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): US |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): AT BE CH DE DK ES FR GB GR IE IT LU MC NL PT SE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 08666370 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1995934878 Country of ref document: EP |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWP | Wipo information: published in national office |
Ref document number: 1995934878 Country of ref document: EP |
|
WWG | Wipo information: grant in national office |
Ref document number: 1995934878 Country of ref document: EP |