CN116461066A - Intelligent setting method for injection molding process parameters based on scientific test mold - Google Patents
Intelligent setting method for injection molding process parameters based on scientific test mold Download PDFInfo
- Publication number
- CN116461066A CN116461066A CN202310536421.5A CN202310536421A CN116461066A CN 116461066 A CN116461066 A CN 116461066A CN 202310536421 A CN202310536421 A CN 202310536421A CN 116461066 A CN116461066 A CN 116461066A
- Authority
- CN
- China
- Prior art keywords
- pressure
- injection
- product
- setting
- mold
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 74
- 238000012360 testing method Methods 0.000 title claims abstract description 56
- 238000001746 injection moulding Methods 0.000 title claims abstract description 55
- 230000008569 process Effects 0.000 claims abstract description 45
- 238000013528 artificial neural network Methods 0.000 claims abstract description 30
- 238000012549 training Methods 0.000 claims abstract description 29
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 17
- 230000002068 genetic effect Effects 0.000 claims abstract description 14
- 238000002474 experimental method Methods 0.000 claims abstract description 11
- 238000005206 flow analysis Methods 0.000 claims abstract description 7
- 238000012795 verification Methods 0.000 claims abstract description 5
- 238000002347 injection Methods 0.000 claims description 104
- 239000007924 injection Substances 0.000 claims description 104
- 239000010410 layer Substances 0.000 claims description 26
- 239000000155 melt Substances 0.000 claims description 16
- 238000004519 manufacturing process Methods 0.000 claims description 15
- 239000000463 material Substances 0.000 claims description 14
- 238000000465 moulding Methods 0.000 claims description 10
- 238000001816 cooling Methods 0.000 claims description 9
- 238000004088 simulation Methods 0.000 claims description 5
- 238000010276 construction Methods 0.000 claims description 4
- 238000012546 transfer Methods 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 210000002569 neuron Anatomy 0.000 claims description 2
- 239000002356 single layer Substances 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 abstract description 5
- 238000012937 correction Methods 0.000 abstract description 5
- 238000005457 optimization Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 4
- 238000013461 design Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 239000012768 molten material Substances 0.000 description 3
- 238000003062 neural network model Methods 0.000 description 3
- 238000013526 transfer learning Methods 0.000 description 3
- 238000010200 validation analysis Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000013178 mathematical model Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000010008 shearing Methods 0.000 description 2
- 241001155433 Centrarchus macropterus Species 0.000 description 1
- 229930183217 Genin Natural products 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000010924 continuous production Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 229920000642 polymer Polymers 0.000 description 1
- 239000002861 polymer material Substances 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
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
- B29C45/766—Measuring, controlling or regulating the setting or resetting of moulding conditions, e.g. before starting a cycle
-
- 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
- B29C2945/00—Indexing scheme relating to injection moulding, i.e. forcing the required volume of moulding material through a nozzle into a closed mould
- B29C2945/76—Measuring, controlling or regulating
- B29C2945/76822—Phase or stage of control
- B29C2945/76913—Parameter setting
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Mechanical Engineering (AREA)
- Injection Moulding Of Plastics Or The Like (AREA)
Abstract
The invention discloses an intelligent setting method of injection molding process parameters based on a scientific test mold, which comprises the following steps of 11 steps, firstly, obtaining mold and product information; scientific test is carried out by using a die flow software to obtain a product forming process window; setting an orthogonal experiment table in a process window, and carrying out product model flow analysis according to an orthogonal experiment scheme; obtaining a model flow result data set, and dividing the model flow result data set into a training set, a verification set and a test set; constructing a BP neural network; training and verifying a neural network; constructing a genetic algorithm; recommending initial technological parameters; producing products by an injection molding machine; correcting the prediction model; and finally recommending the correction process parameters. According to the invention, the process window is determined by utilizing the scientific test model, the training data range of the neural network prediction model is determined through the process window, the problem that a large amount of training data is required by the neural network is solved, the calculation of invalid parameters by the model flow analysis is reduced, the training of the neural network on the invalid data is prevented, and the effectiveness of the training data and the reliability and the interpretability of the prediction result are improved.
Description
Technical Field
The invention belongs to the technical field of injection molding, and particularly relates to an intelligent setting method for injection molding process parameters based on a scientific test mold.
Background
Injection molding is one of the most important processing modes of polymer materials, has the advantages of high production efficiency, high product precision, good molding consistency and the like, and has the advantage of high degree of automation. However, in the actual production process, operators are required to have abundant production experience during the test production of the product, and the setting of the injection molding process parameters can be completed through repeated test. The method has the advantages that the time cost and the material cost of the test mould of the large-scale product are high, in addition, the quality of the product can be influenced by defects such as internal stress non-detection of the naked eyes of the precise product, and the subsequent continuous production process is finally influenced.
With the development and perfection of artificial intelligence technology in the manufacturing field, researchers construct an injection molding process knowledge base through an expert system, a case reasoning system and the like, and intelligent adjustment of process parameters in the injection molding process is gradually realized. For setting of initial technological parameters of injection molding, researchers construct a prediction system by adopting an artificial neural network, so that the prediction of the initial technological parameters is realized. However, the black box calculation approach of the neural network results in poor interpretability of the prediction process and the prediction result. The existing intelligent prediction method for the injection molding process parameters has the problems of large training data quantity requirement and large prediction result deviation, and has poor practicability in the practical application process.
In fact, when the design of the product mold is finished and the mold production stage is entered, a manufacturing period of more than 30 days generally exists, so that numerical analysis and initial technological parameter prediction can be performed by utilizing the product and the mold three-dimensional model in the mold production period, the mold test period and the mold test cost can be greatly shortened, the dependence of technological setting on manual experience is reduced, and important promotion effect is provided for improving the market competitiveness of manufacturers.
Disclosure of Invention
Aiming at the problems, the invention provides an intelligent setting method for injection molding process parameters based on scientific test molds, which aims at analyzing the digital mold structure of products in the mold manufacturing period and realizing accurate prediction for the injection molding process parameters.
In the method, firstly, a digital three-dimensional model of a mold is analyzed by using mold flow analysis software, and information such as a mold cavity, a runner, a gate, a cooling channel and the like is identified; performing scientific model testing flow by using model flow analysis software to determine an injection molding process window, and performing model flow analysis again in the process window to obtain training data for training the BP neural network; and carrying out iterative optimization on the BP neural network prediction model by utilizing a genetic algorithm, and carrying out prediction and recommendation of technological parameters by the genetic algorithm according to the target gram weight set by a user. Meanwhile, the transfer learning is adopted to correct the prediction model and expand the database. The method comprises the following steps:
s10: acquiring die and product information;
s20: scientific test is carried out by using a die flow software to obtain a product forming process window; compared with the traditional experience test, the scientific test is characterized in that mathematical models and physical models which are subjected to theoretical derivation and experimental verification are arranged in the model flow software, and recommended process windows are obtained for some algorithms, so the application is called scientific test;
s30: setting an orthogonal experiment table in a process window, and carrying out product model flow analysis according to an orthogonal experiment scheme;
s40: obtaining a model flow result data set, and dividing the model flow result data set into a training set, a verification set and a test set;
s50: constructing a BP neural network;
s60: training and verifying a neural network;
s70: constructing a genetic algorithm;
s80: recommending initial technological parameters;
s90: producing products by an injection molding machine;
s100: correcting the prediction model;
s110: and recommending correction process parameters.
Preferably, the mold and product information comprises: product volume, gate number, cavity number, projected area, gate and cavity definition, waterway structure distribution, etc.
Preferably, the method for acquiring the mold and product information comprises the following steps: identifying the mould and the product information by using mould flow software; and importing the three-dimensional information of the die into die flow analysis software, identifying and acquiring structural characteristics of a sprue, a runner and a cavity of a die cavity through the die flow software, and identifying waterway characteristics of the die.
Preferably, the step of scientifically testing the mold comprises: injection speed determination, cavity balance determination, injection pressure determination, dwell time determination and the like;
preferably, the method for performing scientific modeling comprises: using Moldex3D, moldflow and other modular software;
preferably, the step of determining the S21 injection speed in the step of S20 scientific test comprises:
setting a single-stage injection speed, selecting a maximum injection speed, and simulating and recording the peak pressure and injection time of a screw; sequentially reducing the injection speed by taking 10% of the maximum injection speed as an interval, repeating the steps and recording data; drawing a relation curve of the shear rate and the effective viscosity according to the recorded screw peak pressure and filling time, wherein a section with gentle melt viscosity is a numerical section determined by the injection speed; in injection molding, the shear rate is typically plotted against effective viscosity using the following formula:
μ＝P p ×t×γ
wherein μ represents melt viscosity, P p Representing the peak screw pressure during injection, t representing injection time, r representing shear rate, and gamma representing the boost ratio (the value of which is determined according to the type of injection machine selected).
Preferably, the step of determining the cavity balance in step S22 includes: setting the pressure maintaining time and the pressure maintaining pressure to be 0 by adopting a mode of pressure maintaining injection, simulating 10 times, calculating the weight difference of the cavities of the multi-cavity product, and considering that the cavities are filled and balanced when the weight difference is smaller than the specified difference of the products;
preferably, the step of determining the injection pressure in step S23 includes: setting dwell time and dwell pressure to be 0 by adopting a mode of non-dwell injection, setting VP (initial letters of velocity and pressure) conversion points (namely conversion from velocity control to pressure control) to fill 20%, 40%, 60%, 80% and 100% of a cavity respectively, simulating and recording screw peak pressure of each filling, judging whether the injection peak pressure exceeds a maximum injection pressure value of a machine, and considering that the injection machine meets the injection molding requirement of a product when the injection peak pressure does not exceed the maximum injection pressure value;
preferably, the step of determining the holding pressure in step S24 includes: setting the melt temperature as the lower limit of a material recommended interval, setting the VP transfer pressure point position as 98% of the cavity volume, setting the difference value between the minimum value and the maximum value by 5% or 10%, simulating and recording the pressure-maintaining pressure numerical interval of which the product quality meets the requirement; setting the melt temperature as the upper limit of a material recommended interval, setting the VP transfer pressure point position as 98% of the cavity volume, setting the difference value between the minimum value and the maximum value by 5% or 10%, simulating and recording the pressure-maintaining pressure numerical interval of which the product quality meets the requirement; thereby obtaining an effective numerical interval of the holding pressure;
preferably, the step wherein the dwell time determination of step S25 comprises:
after the setting of the melt temperature, the mold temperature, the injection speed and the dwell pressure is completed according to the steps, setting the dwell time from 0 to the maximum cooling time at a difference of 1s, simulating and recording the product quality, and determining the effective numerical interval of the dwell time;
the molding process window obtained by the scientific test flow comprises the following steps: melt temperature, mold temperature, injection speed, dwell pressure, dwell time, etc.; the melt temperature and the mold temperature are determined according to the processing temperature range recommended by a material provider;
preferably, the process window obtained by scientific experimentation contains a range of valid values for the following parameters: injection speed, melt temperature, mold temperature, dwell pressure, dwell time;
preferably, the orthogonal test is designed as follows: taking the injection speed, the melt temperature, the mold temperature, the pressure maintaining pressure and the pressure maintaining time of the technological parameters to be optimized as factors, dividing the effective value range of the parameters to be optimized obtained by scientific test mold into 5 levels, and adopting an orthogonal experiment table with 5 factors and 5 levels to determine 25 groups of technological parameter combinations for simulation calculation;
preferably, the modulo flow result dataset comprises: melt temperature, mold temperature, injection speed, dwell pressure, dwell time, product gram weight, product critical position warp curvature, etc.;
preferably, the model result dataset can be divided into a 70% training set, a 20% validation set and a 10% test set;
preferably, the expression of the dataset is:
D＝{(X 1 ,y 1 ),(X 2 ,y 2 )...,(X i ,y i )},
wherein X is i Representing the input of the ith set of sample data after construction of the hidden layer, y i Output X representing ith set of sample data i ＝(X i1 ,X i2 ,X i3 ,X i4 ,X i5 )∈R K ；
Preferably, the step of constructing the BP neural network comprises: setting the hidden layer number, the hidden layer node number, the input node number, the output node number, the iteration number and the like;
preferably, the step of constructing a genetic algorithm comprises: setting population size, evolution times, crossover probability and variation probability;
preferably, the expression of the objective function of the genetic algorithm is constructed as follows:
Obj＝-|y-y|
wherein y is the ideal product weight set by the user, and y is the product weight predicted by the neural network model according to the injection molding process parameters in the scientific test range;
compared with the prior art, the invention has the following technical advantages: the invention provides an intelligent setting method for injection molding process parameters based on a scientific test model, which utilizes the scientific test model to determine a process window, determines the training data range of a neural network prediction model through the process window, solves the problem that the neural network needs a large amount of training data, reduces the calculation of invalid parameters aiming at the model flow analysis, prevents the training of the neural network on the invalid data, improves the effectiveness of the training data, improves the reliability of a prediction result and improves the interpretability of the result. In addition, the invention also uses a prediction model correction mechanism, when the recommended initial technological parameter fails to meet the product quality requirement, the prediction initial technological parameter and the product production result are fed back to the prediction model to carry out model learning correction, and the corrected technological parameter is recommended, so that the efficient prediction of the injection molding technological parameter is realized. Finally, the invention introduces a transfer learning technology, after the technological parameter prediction is completed, the mold and material attribute of the predicted production period are recorded, and the prediction model is optimized by combining transfer learning data, so that the learning efficiency and the prediction precision of the prediction model are gradually improved.
According to the invention, in the production period of the die, analysis simulation and prediction can be performed by using computer data, no additional sensor or hardware system is needed, and a user can realize remote calculation and prediction by uploading the product and the die data to the cloud, so that the method has strong industrial popularization and application convenience.
Drawings
For more clearly expressing the implementation mode of the technical scheme of the invention, the drawings are described in detail.
FIG. 1 is a flow chart of an intelligent setting method for injection molding process parameters based on a scientific test mold provided by the embodiment of the invention;
FIG. 2 is a product diagram of an embodiment of the present invention;
FIG. 3 is a viscosity profile of an embodiment of the present invention;
FIG. 4 is a graph of injection pressure peaks for an embodiment of the present invention;
FIG. 5 is a dwell pressure window of an embodiment of the invention;
fig. 6 is a structural diagram of a BP neural network in an intelligent setting method for injection molding process parameters based on a scientific test model according to an embodiment of the present invention.
Detailed Description
The invention will be elucidated below in connection with specific embodiments. The detailed description is intended to be illustrative, but not limiting. The terms used herein are those commonly used in the art. Accordingly, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The raw materials, instruments, equipment and the like used in the invention can be obtained by market purchase or can be prepared by the existing method.
Referring to fig. 1, the intelligent setting method of injection molding process parameters based on scientific test molds in the embodiment of the invention comprises the following steps:
s10: obtaining die and product information: the mold and product information to be acquired may include: product volume, gate number, cavity number, projected area, etc.; in the embodiment, mold and product identification is performed by using Moldex3D mold flow analysis software, and each part of a product structure body, a gate and a runner is determined in the software to determine the waterway distribution condition of the product mold. The product of this example is shown in FIG. 2, and the selected material is PP (YUNSSOX 1120).
S20: scientific test is carried out by using die flow software to obtain a forming process window: in the embodiment, the Moldex3D model flow analysis software is adopted for scientific model test, a machine is selected for analysis in the software, and a long-flight ZE1200 full-electric injection molding machine is selected for respectively carrying out injection speed determination, cavity balance determination, injection pressure determination, pressure maintaining pressure determination and pressure maintaining time determination.
S21: injection speed was determined. Due to the non-newtonian nature of the polymer melt, the injection rate affects the shear rate of the melt, which in turn causes a change in melt viscosity, and thus the plateau viscosity interval for melt filling is determined by the injection rate test. In the step, by adjusting the injection speed, the shearing rate of plastic melt and the peak pressure of a screw rod at different injection speeds are recorded, a relation curve of the shearing rate and the viscosity is drawn, and the optimal injection speed corresponding to the viscosity stable section is determined. According to the scientific molding theory accepted by the injection molding industry, the following general formula is adopted when the relation curve of the shear rate and the effective viscosity is drawn:
μ＝P p ×t×γ
wherein μ represents melt viscosity, P p The screw peak pressure representing the injection process, t representing the injection time, r representing the shear rate, and γ representing the enhancement ratio, the values of which are determined according to the type of injection molding machine, in this embodiment a long flier ZE1200 all-electric injection molding machine is used, all-electric injection molding machine γ=1; in the case of a hydraulically driven injection molding machine, the boost ratio is the ratio of the maximum injection pressure to the maximum hydraulic pressure.
The method comprises the following specific steps:
1. setting the pressure maintaining pressure and the pressure maintaining time to be 0, setting the position of the V/P switching point to be 95% of the position of the cavity, setting the position of the V/P switching point to be 35mm for the product, and setting the storage position to be 60mm;
2. setting a single-stage injection speed, wherein the injection speed is the highest injection speed allowed by the injection molding machine, starting simulation and recording the peak pressure and injection time of the screw;
3. sequentially reducing the injection speed by taking 10% of the maximum injection speed as an interval, repeating the steps and recording data;
4. drawing a shear rate versus effective viscosity curve from the recorded screw peak pressure and fill time, as shown in fig. 3;
according to the drawn relation curve, determining that the injection speed interval is 100-190 mm/s;
s22: and determining the balance of the cavity. In order to ensure the filling stability and balance of a multi-cavity product and the consistency of the product, the filling balance of the cavity needs to be determined. The cavity balance test steps are as follows:
1. randomly selecting the injection speed in the injection speed interval determined in the step S21, wherein the single-stage injection speed is set to be 120mm/S, and the position of the V/P switching point is set to be 35mm;
2. setting the dwell time and dwell pressure to 0, setting the cooling time to be sufficiently long, where the cooling time is set to 30s;
3. continuously injecting 10 molded articles, recording the weight of the articles in each mold cavity, and calculating the weight difference of the two mold cavities;
the test results are recorded in the following table, and the maximum deviation of the product weights of the two cavities is 0.003g, so that the balance requirement of the cavities is met. If the weight deviation of the product is too large, the mold is required to be modified when the unbalanced filling of the cavity exists.
| Group of | Cavity 1 (g) | Cavity 2 (g) | Weight difference (g) |
| 1 | 9.411 | 9.413 | 0.002 |
| 2 | 9.409 | 9.412 | 0.003 |
| 3 | 9.409 | 9.410 | 0.001 |
| 4 | 9.413 | 9.410 | 0.003 |
| 5 | 9.411 | 9.412 | 0.001 |
| 6 | 9.411 | 9.412 | 0.001 |
| 7 | 9.414 | 9.413 | 0.001 |
| 8 | 9.412 | 9.413 | 0.001 |
| 9 | 9.412 | 9.413 | 0.001 |
| 10 | 9.412 | 9.413 | 0.001 |
S23: injection pressure determination. Because the injection molding machine mainly takes the guaranteed speed as priority in the injection process, the pressure responds in real time according to the melt viscosity and the cavity structure, the set injection speed can not be reached due to the setting of the too small injection pressure, the cavity is worn due to the setting of the too large injection pressure, and the service life of the die is influenced. The injection pressure is determined as follows:
1. the injection speed was set to 120mm/S (selected within the interval determined in S21), and the dwell pressure and dwell time were set to 0;
2. setting the injection pressure to be the maximum pressure which can be provided by the injection molding machine;
3. adjusting the V/P switching position to enable the molten material to be filled to 20%, 40%, 60%, 80% and 100% of the cavity respectively, and recording the peak pressure of the screw rod filled each time;
4. drawing a graph of the flow position of the molten material and injection peak pressure, and judging whether the injection peak pressure exceeds a maximum injection pressure value of the machine;
5. after the machine is determined to provide enough injection pressure, determining an injection pressure set value according to the injection peak pressure;
the following table shows the recorded peak screw pressure in the mold test process, and the change curve of the screw pressure along with the injection volume in the injection process and the maximum pressure curve of the injection molding machine are shown in fig. 4, and the maximum injection pressure in the injection process is 135.521Mpa and is less than the peak injection pressure of 250Mpa, so that the equipment pressure in the injection process can meet the requirements of injection rate setting. In order to ensure stable operation of the injection molding machine, a safety factor of 1.2 was taken here, so the injection pressure was set to 162Mpa 20% higher than the peak injection pressure.
| Sequence number | Flow-through region of the melt | Injection peak pressure (Mpa) |
| 1 | Nozzle | 14.736 |
| 2 | Main runner | 19.324 |
| 3 | Gate | 22.735 |
| 4 | Filling 1/3 | 50.756 |
| 5 | Filling 2/3 | 94.313 |
| 6 | Filling end | 135.521 |
S24: and (5) determining the pressure maintaining pressure. The pressure maintaining pressure can be determined through a pressure maintaining pressure window, the pressure maintaining window is established according to the pressure maintaining pressure and the temperature of the molten material, the size of the pressure maintaining window represents the stability of the injection molding process, the larger the pressure maintaining window is, the more stable the injection molding process is, the products produced in the pressure maintaining window are qualified products without appearance defects, and when the pressure maintaining pressure of the products produced by parameters outside the pressure maintaining window is larger than the maximum value of the pressure maintaining window, the problems such as flash and the like can occur; when the pressure is smaller than the minimum value of the pressure maintaining window, the problems of short shot, sinking and the like can occur. Therefore, in order to stabilize the injection molding process, parameters within the dwell window should be selected as actual molding parameters.
The specific steps of the pressure maintaining pressure determination are as follows:
1. randomly selecting an injection speed within the injection speed interval determined in the step S21, wherein the injection speed is set to be 120mm/S; the cooling time should be set long enough, where the cooling time is set to 30s; the V/P switching point is set to be 35mm;
2. setting the melt temperature to a lower limit value of the melt temperature recommended by the material manufacturer, and setting 220 ℃;
3. firstly, setting the pressure maintaining pressure to be a smaller value, then gradually increasing, and taking the pressure maintaining pressure at the moment as a low-temperature low-pressure point at the left lower corner of a pressure maintaining window when a first product with acceptable quality appears;
4. continuously increasing the pressure maintaining pressure with the same increment until the first product with flash appears, and taking the pressure maintaining pressure at the moment as a low-temperature high-pressure point of a pressure maintaining window;
5. setting the melt temperature to be the upper melt temperature limit recommended by the material manufacturer, and setting 240 ℃;
6. repeating the previous steps, gradually increasing the pressure maintaining pressure from a smaller value, and recording the pressure maintaining pressure at the moment when the first product with acceptable appearance appears as a high-temperature low-pressure point at the right lower corner of the pressure maintaining window;
7. continuously increasing the pressure maintaining pressure with the same increment until the first product with flash appears, and taking the pressure maintaining pressure at the moment as a high-temperature high-pressure point of a pressure maintaining window;
8. connecting the recorded four limit position points to generate a pressure maintaining pressure window, as shown in fig. 5;
the holding pressure and the material temperature should be selected within this window.
S25: and (5) determining the dwell time. The dwell time is determined depending on the gate freeze time, which may be determined by a change in the weight of the article. The dwell time was simulated from 1s to 10s increases and was determined by counting the product weight. The method comprises the following specific steps:
1. randomly selecting an injection speed within the injection speed interval determined in the step S21, wherein the injection speed is set to be 120mm/S; selecting a holding pressure within the holding pressure section determined in step S24, where the holding pressure is set to 90MPa; setting the V/P switching position to be 32mm; the cooling time should be set long enough, where the cooling time is set to 30s;
2. the dwell time is set from 0, the dwell time is gradually increased, one second is added each time, then injection is carried out, and the quality of the product after each molding is recorded;
the table below shows the change in product weight with dwell time, and when the dwell time is greater than 6s, the product weight no longer changes.
| Group of | Dwell time(s) | Product weight (g) |
| 1 | 1 | 26.048 |
| 2 | 2 | 26.050 |
| 3 | 3 | 26.054 |
| 4 | 4 | 26.061 |
| 5 | 5 | 26.066 |
| 6 | 6 | 26.071 |
| 7 | 7 | 26.071 |
| 8 | 8 | 26.071 |
| 9 | 9 | 26.071 |
| 10 | 10 | 26.071 |
S30: in the embodiment, the orthogonal experiment design is performed by using DOE genins in Moldex3D, and the field-opening orthogonal table is shown in the following table:
| control factor | Level 1 | Level 2 | Level 3 | Level 4 | Level 5 |
| Plastic temperature, DEG C | 220 | 225 | 230 | 235 | 240 |
| Die temperature, DEG C | 25 | 30 | 35 | 40 | 45 |
| Percent of maximum speed of injection% | 55 | 65 | 75 | 85 | 95 |
| Dwell time s | 2 | 3 | 4 | 5 | 6 |
| Percent of maximum holding pressure% | 55 | 65 | 75 | 85 | 95 |
Wherein, the maximum injection speed of the injection molding machine is 200mm/s, and the maximum holding pressure is 200MPa. And importing the orthogonal test table into a model flow analysis software for simulation.
S40: a modulo flow result dataset is obtained and split into a training set, a validation set and a test set. The obtained die flow result comprises the following steps: melt temperature, mold temperature, injection speed, dwell pressure, dwell time, product grammage, warp curvature, volume shrinkage; the resulting dataset obtained can be divided into a 70% training set, a 20% validation set and a 10% test set;
s50: and constructing a BP neural network.
The construction of the BP neural network comprises the following steps:
constructing an input layer;
constructing an output layer;
constructing a hidden layer;
wherein the input layer node is: melt temperature, mold temperature, injection speed, dwell pressure, dwell time;
wherein the output layer node is: product gram weight, warp rate and volume shrinkage rate;
wherein the hidden layer is a single layer, and the number of neurons of the hidden layer can be set to be 6;
wherein the input end of the hidden node layer is connected with the output end of the input layer, and the output end of the hidden node layer is connected with the input end of the output layer, so that the BP neural network shown in figure 6 can be obtained;
this embodiment uses python software to construct the neural network and optimization algorithm.
S60: training and validating the neural network. Importing the training set data obtained in the step S40 into a BP neural network for training, setting BP neural network parameters according to model prediction, setting training iteration times at 2000 rounds, setting batch size at 8 and setting learning rate at 0.05;
s70: a genetic algorithm is constructed. The optimization target of the genetic algorithm is that the deviation rate of the gram weight of the product and the design gram weight is smaller than 0.5%, the warp deformation is minimum, and the constraint conditions are the optimization range of the temperature of the die, the temperature of the melt, the injection speed, the pressure maintaining pressure and the pressure maintaining time;
the mathematical model of the genetic algorithm is as follows:
F min ＝f(x 1 ,x 2 ,x 3 ,x 4 ,x 5 ,x 6 )
wherein x is n The constraint range of (2) is determined according to the molding process window; setting the population scale of the genetic algorithm as 50, the evolution times as 100, the crossover probability as 0.5 and the variation probability as 0.4;
s80: initial process parameters are recommended. In the embodiment, the gram weight of the product is selected as a prediction object, the gram weight of the product is input into a prediction model to be 26.5g, and the prediction model predicts and recommends initial technological parameters according to the product requirement; and (2) automatically removing a prediction result outside the process window range by the prediction model according to the process parameter range determined in the step (S2), wherein the initial process parameters recommended by the prediction model are as follows:
the expression of the objective function of the construction genetic algorithm is as follows:
Obj＝-|26.5-y|
wherein, the ideal product weight set by the user is 26.5g, and y is the product weight predicted by the neural network model according to the injection molding process parameters in the scientific test range;
s90: the injection molding machine produces the product. Inputting recommended initial technological parameters on an injection molding machine, detecting whether the actual product quality meets the production requirement, detecting that the weight of the product is 26.492g, and the standard deviation of the product is 0.008g, so that the product meets the product requirement;
s100: and correcting the prediction model. If the gram weight of the produced product does not meet the requirement, inputting the recommended technological parameters of the step S8 and the gram weight of the actually produced product into a prediction model, and retraining the prediction model;
s110: and recommending correction process parameters. And (3) the retrained prediction model predicts again according to the product gram weight input in the step S8, and outputs new technological parameters.
The invention provides an intelligent setting method of injection molding process parameters based on a scientific test model, which utilizes the method of the scientific test model to determine a molding process window, performs small sample high-efficiency model flow analysis in the process window, trains a BP neural network by utilizing a small sample high-efficiency model flow result, performs target optimization on a prediction result by combining a genetic algorithm, and finally realizes the prediction of injection molding initial process parameters. The method can effectively solve the requirement of the BP neural network on the data size of the training sample, eliminates a large amount of invalid data by using a molding process window, and realizes the efficient training of the neural network model; and the genetic algorithm is combined to realize multi-objective optimization of the prediction result, so that the prediction precision can be effectively improved. In addition, the molding process window obtained based on scientific trial modeling can screen the prediction parameters output by the prediction model, and automatically reject the results outside the process window, thereby effectively improving the accuracy and scientificity of the prediction results.
Aiming at a longer blank window period in the mould manufacturing process, the intelligent prediction of injection molding process parameters is realized by combining computer numerical analysis with artificial intelligence, and the mould test period and the material cost can be effectively reduced. It should be noted that the above-described embodiments are merely illustrative of the principles of the present invention and not in limitation thereof. The mode of realizing the injection molding process parameter setting function by utilizing various types of neural networks, optimization algorithms and the like and relying on the process is within the protection scope of the invention. Any modification, equivalent replacement, improvement, etc. made without departing from the spirit and scope of the present invention should be included in the scope of the present invention. Furthermore, the appended claims are intended to cover all such changes and modifications that fall within the scope and boundary of the appended claims, or equivalents of such scope and boundary.
Claims (6)
1. An intelligent setting method for injection molding process parameters based on scientific test molds is characterized by comprising the following steps:
step S10: acquiring die and product information;
step S20: scientific test is carried out by using a die flow software to obtain a product forming process window;
step S30: setting an orthogonal experiment table in a process window, and carrying out product model flow analysis;
step S40: obtaining a model flow result data set, and dividing the model flow result data set into a training set, a verification set and a test set;
step S50: constructing a BP neural network;
step S60: training and validating a neural network: importing the training set data obtained in the step S40 into a BP neural network for training, and setting BP neural network parameters according to model prediction;
step S70: constructing a genetic algorithm;
step S80: the initial process parameters are recommended: the prediction model predicts and recommends initial technological parameters according to the product requirements;
step S90: and (3) producing products by an injection molding machine: inputting recommended initial technological parameters on an injection molding machine, and detecting whether the actual product quality meets the production requirement;
step S100: correcting the prediction model:
step S110: recommending and correcting technological parameters;
the step S10 includes: the three-dimensional information of the mold is imported into mold flow analysis software, the structural characteristics of a sprue, a runner and a cavity of a mold cavity are identified and acquired through the mold flow software, and the waterway characteristics of the mold are identified;
the step S20 includes:
defining material information of a product in a model flow software; and carrying out a scientific model testing process in model flow software, wherein the scientific model testing process comprises the following steps of: injection speed determination, cavity balance determination, injection pressure determination, dwell time determination and the like, wherein the molding process window obtained by the scientific mold testing flow comprises the following steps: melt temperature, mold temperature, injection speed, dwell pressure, dwell time, etc.; wherein the melt temperature and the mold temperature are determined according to recommended temperatures of a material provider.
2. The intelligent setting method of injection molding process parameters based on scientific test mold according to claim 1, wherein step S20 comprises the following steps:
a) The step of determining the injection speed in step S21 includes:
setting a single-stage injection speed, selecting a maximum injection speed, and simulating and recording the peak pressure and injection time of a screw; sequentially reducing the injection speed by taking 10% of the maximum injection speed as an interval, repeating the steps and recording data; drawing a relation curve of the shear rate and the effective viscosity according to the recorded screw peak pressure and filling time, wherein a section with gentle melt viscosity is a numerical section determined by the injection speed;
b) The step of determining the cavity balance in step S22 includes:
setting the pressure maintaining time and the pressure maintaining pressure to be 0 by adopting a mode of pressure maintaining injection, simulating 10 times, calculating the weight difference of the cavities of the multi-cavity product, and considering that the cavities are filled and balanced when the weight difference is smaller than the specified difference of the products;
c) The step of determining the injection pressure in step S23 includes:
setting the pressure maintaining time and the pressure maintaining pressure to be 0 by adopting a mode of pressure-maintaining-free injection, setting the V/P conversion points to be respectively filled to 20%, 40%, 60%, 80% and 100% of the cavity, simulating and recording the peak pressure of a screw rod filled each time, judging whether the injection peak pressure exceeds the maximum injection pressure value of the machine, and considering that the injection machine meets the injection molding requirement of a product when the injection peak pressure does not exceed the maximum injection pressure value of the machine;
d) The step of determining the holding pressure in step S24 includes:
setting the melt temperature as the lower limit of a material recommended interval, setting the V/P transfer pressure point position as 98% of the cavity volume, setting the difference value between the minimum value and the maximum value by 5% or 10%, simulating and recording the pressure-maintaining pressure numerical interval of which the product quality meets the requirement; setting the melt temperature as the upper limit of a material recommended interval, setting the V/P transfer pressure point position as 98% of the cavity volume, setting the difference value between the minimum value and the maximum value by 5% or 10%, simulating and recording the pressure-maintaining pressure numerical interval of which the product quality meets the requirement; thereby obtaining an effective numerical interval of the holding pressure;
e) The step of determining the dwell time in step S25 includes:
after the setting of the melt temperature, the mold temperature, the injection speed and the dwell pressure is completed according to the steps, the dwell time is set from 0 to the maximum cooling time at the interval of 1s, the simulation is carried out, the product quality is recorded, and the effective numerical interval of the dwell time is determined.
3. The intelligent setting method of injection molding process parameters based on scientific test mold according to claim 1, wherein the step S30 comprises:
designing an orthogonal experiment table through an orthogonal experiment method, and carrying out an orthogonal experiment through a model flow software;
the orthogonal experimental table contains the following data: melt temperature, mold temperature, injection speed, dwell pressure, dwell time, product grammage, or/and product critical position warp curvature, etc.
4. The intelligent setting method of injection molding process parameters based on scientific test according to claim 1, wherein the step S40 comprises: dividing the orthogonal experiment result data obtained in the step S30 into a training set, a verification set and a test set; wherein the expression of the dataset is:
D＝{(X 1 ,y 1 ),(X 2 ,y 2 )...,(X i ,y i )},
wherein X is i Representing the input of the ith set of sample data after construction of the hidden layer, y i Output X representing ith set of sample data i ＝(X i1 ,X i2 ,X i3 ,X i4 ,X i5 )∈R K 。
5. The intelligent setting method of injection molding process parameters based on scientific test according to claim 2, wherein the step S50 comprises: constructing an input layer, an output layer and a hidden node layer, wherein the input layer node is as follows: melt temperature, mold temperature, injection speed, dwell pressure, dwell time; the output layer node is: product gram weight, warp rate and volume shrinkage rate; the hidden layer is a single layer, the number of neurons of the hidden layer is set to be 6, and the output end of the input node layer is connected with the input end of the hidden layer; and connecting the output end of the hidden node layer with the input end of the output layer.
6. The intelligent setting method for injection molding process parameters based on scientific test molds according to claim 1, wherein when the produced product target does not meet the requirements, the process parameters recommended in the step S80 and the gram weights of the actually produced products are input into a prediction model, and the prediction model is retrained.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310536421.5A CN116461066A (en) | 2023-05-12 | 2023-05-12 | Intelligent setting method for injection molding process parameters based on scientific test mold |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310536421.5A CN116461066A (en) | 2023-05-12 | 2023-05-12 | Intelligent setting method for injection molding process parameters based on scientific test mold |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116461066A true CN116461066A (en) | 2023-07-21 |
Family
ID=87177145
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310536421.5A Pending CN116461066A (en) | 2023-05-12 | 2023-05-12 | Intelligent setting method for injection molding process parameters based on scientific test mold |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116461066A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117245873A (en) * | 2023-11-14 | 2023-12-19 | 广东美的制冷设备有限公司 | Method and device for determining initial technological parameters of injection molding machine and injection molding machine |
-
2023
- 2023-05-12 CN CN202310536421.5A patent/CN116461066A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117245873A (en) * | 2023-11-14 | 2023-12-19 | 广东美的制冷设备有限公司 | Method and device for determining initial technological parameters of injection molding machine and injection molding machine |
| CN117245873B (en) * | 2023-11-14 | 2024-04-05 | 广东美的制冷设备有限公司 | Method and device for determining initial technological parameters of injection molding machine and injection molding machine |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110377948A (en) | A kind of injection parameters Multipurpose Optimal Method | |
| CN111079338A (en) | Method for optimizing injection molding process of automobile rearview mirror shell | |
| CN113771289B (en) | Method and system for optimizing injection molding process parameters | |
| CN111231251B (en) | Product detection method, equipment and system of injection molding machine | |
| CN100405376C (en) | Determination method of plastic injection technological parameter and injection moulding machine | |
| CN102773981B (en) | Implementation method of energy-saving and optimizing system of injection molding machine | |
| CN110531722B (en) | Process parameter recommendation system and method based on data acquisition | |
| CN1851715A (en) | Intelligent repair method of injection molding during plastic injection process and injection molding machine | |
| US11860590B2 (en) | Model-free optimization method of process parameters of injection molding | |
| CN116461066A (en) | Intelligent setting method for injection molding process parameters based on scientific test mold | |
| Karasu et al. | FIS-SMED: a fuzzy inference system application for plastic injection mold changeover | |
| CN110502779A (en) | A kind of Intelligentized design method of the injection forming mold based on fuzzy reasoning | |
| Moayyedian et al. | Multi-objective optimization of injection molding process for determination of feasible moldability index | |
| CN109100941B (en) | Multi-stage intermittent process two-dimensional anti-interference prediction controller design method | |
| CN109783898A (en) | A kind of intelligent optimization method of injection molding manufacturing technique parameter | |
| CN109614651A (en) | A kind of high-precision evaluation method of moulding machined parameters and deformation relationship | |
| US20210213662A1 (en) | Method for controlling a machine for processing plastics | |
| CN111428329A (en) | Model-based machine learning system | |
| CN116252447A (en) | Gear injection molding regulation and control method | |
| CN206633366U (en) | A kind of injection-molded item defect self-cure regulation and control device | |
| Studer et al. | Numerical shape optimization as an approach to reduce material waste in injection molding | |
| Yang et al. | Research on Optimization of Injection Molding Process Parameters of Automobile Plastic Front‐End Frame | |
| Zhou et al. | An integrated intelligent system for injection molding process determination | |
| CN109932908B (en) | Multi-directional principal component analysis process monitoring method based on alarm reliability fusion | |
| CN109101988A (en) | A kind of analysis method of the unqualified injection molding product based on Bayesian network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |