KR20180116610A - The Optimization method for Finding Injection Foam Molding Input Condition and The computer Program of the Optimization method for Finding Injection Foam Molding Input Condition - Google Patents

Abstract

The present invention is capable of enabling even a beginner to find an injection condition by measuring a material property of an injected product and conducting a minimum injection test. According to the present invention, if input and output conditions for injection are set, the entire area of the input condition is divided at first in a proper size to make an injection test condition. Injection is performed in the test condition, and then, the porosity and standard deviation of an injected product are measured from a result of the conducted injection, and thus, a table is completed in the input condition. A secondary multiple regression analysis model is formed from the completed table for each input condition.

Description

TECHNICAL FIELD The present invention relates to a method of searching for an extraction molding condition using a second mathematical model for finding an injection condition of an injection molding having a predetermined shape, weight and strength, and a computer program using the method. computer Program of the Optimization method for Finding Injection Foam Molding Input Condition}

The present invention relates to a method for quickly and more scientifically finding a condition for producing a foam injection molded article having a cavity inside an injection mold by using a mathematical model and an experiment. Applicable technical fields are applicable to all extrusion injection fields such as general extrusion, injection molding, foam extrusion, foam injection molding, and plastic master batch production process.

The injection conditions in injection molding are set by CAE by molding simulation, and the actual moldings obtain the actual mold profile by measuring the load pressures measured at the molding step. The obtained actual mold profile is input to the CAE, and the CAE is injected along the input profile. However, there has been no suggestion in the prior art on how to find the actual mold profile.

  As another prior art, manufacturing parameters were found by optimizing the manufacturing parameters of plastic injection molding. However, it was impossible to adjust the injection molding parameters in this way because these values were different from actual injection molding results. However, And optimization of the process variables in a stepwise manner to reduce the errors caused by the actual and optimization methods.

U.S. Published Patent Application No. 2002/018375 Chinese Patent Publication CN103914581 A

 The present invention is a solid in which the inside is filled because general injection molding is injection without supplying gas. However, the injection molding according to the present invention relates to injection molding which increases the weight-to-weight ratio of an injection molding by forming a gap in the injection molding by feeding gas together with resin or plastic used for injection in injection molding. The injection molding experiment designing method of the present invention can be used even in the case of general injection molding because the process variable is large. However, the present invention is based on the fact that the output is determined by a complicated and various input such as foam injection molding, It is easier to find the optimum condition of the result.

  As in the case of the present invention, in the case of injection molding in which gas is supplied together with resin or plastic, the size and number of voids per unit volume of the injection molded article vary according to the gas supply amount, gas supply pressure, And strength to the designer's desired value is almost impossible for unexperienced people and there is a problem that can not be easily found even with a lot of experience.

In terms of the input conditions used for injection,

  A variety of plastics are used: polypropylene, polyethylen, HDPE, LDPE, polyvinyl chloride, polystyrene, acrylonitrile butadiene styrene, polylactic acid, polymethyl methacrylate, etc. Base agents (additives for improving the transparency of injection products) and additives: azodicarbonamide, calcium Nano clay, carbon nanofibers, and single walled carbon nanotubes are used. The types and amounts of foamable additives are: CO2, N2, Co2 (N2), n-butane, and water. In addition, processing conditions such as processing temp., Processing pressure and pressure drop rate, mold opening (The process control is performed by opening the mold slightly during injection molding), the shape of the mold die, the shape of the gate, the injection speed, etc. are influential on the injection result You can give.

 In addition, although the shape of the injection-molded product is of course important, the physical and chemical properties of the injection product are also important, so that the properties to be measured in the main properties of the injection product and to be managed in the production process are as follows: porosity, The output value can be a satisfactory physical property of a molded article formed by injection molding, such as non-rigidity, specific gravity, average pore size, product dimensional tolerance, shrinkage, and surface abrasion resistance (abrasion resistance).

Foam injection molding, which is mainly exemplified in the present invention, is supplied with gas and material at the same time. In the case of foam injection molding to form a molded product, trial and error method is mainly used by an experienced person. There is a problem in that many trial and error are required to obtain an injection molded article.

However, even with this empirical trial-and-error method, it is impossible to find the optimum value, and it has been satisfied with obtaining a similar degree of result. This is because a number of variables that have different effects on the physical properties of each other must be optimized at the same time.

In order to solve the above problems, the present invention provides a search method for optimizing an injection molding condition for molding a plastic injection molded article having desired characteristics,

  (S1) setting an injection input condition and setting an objective function using an output condition and a desired characteristic; And

  Generating (S2) a minimum injection test condition for performing an injection test for finding an injection condition satisfying an output condition from the set input condition; The minimum test condition may be set by the user so that at least three test conditions can be set per one input condition in consideration of the minimum changeable set value and the maximum change range of the injection apparatus.

  Performing injection according to the injection test conditions (S3); And

  (S4) of collecting an injection sample of the injection molded product and measuring an output value corresponding to the output condition from the injection sample; And

  (S5) of inputting the measured output values for each of the injection test conditions together with the input conditions, and determining a correlation between the injection test conditions and the output values (S5); And

  A step (S6) of forming a second-order mathematical model expression between the input condition and the output condition through a multiple regression analysis of the correlation with the measurement result; And

  (S7) calculating an optimal condition satisfying the objective function using the second-order mathematical model equation generated in step S6; And

  (S8) of performing the injection by setting the injection conditions again using the calculated optimum conditions and performing the step S4 on the injected injection object; And

  Completing the injection experiment (S9) if the result at step S6 is satisfied with the objective function;

A search method for optimizing an injection molding condition for molding a plastic injection molded article having a desired output value characteristic

Further, an executable file generated from a code that can be executed in a computer as a search method for optimizing the above-mentioned forming conditions also provides a computer program.

According to the present invention, an input condition of a foam injection molding is divided into ranges, an appropriate number of test conditions are generated, and a second-order regression formula is provided for each input condition through multiple regression analysis using experimental results according to the generated test conditions. In the process of repeating the above steps, it is possible to obtain the second-order regression equation that is closest to the actual experimental result, and it is also possible to obtain the least experimental engineer who has little or no experience And a method of obtaining desired results through trial and error.

FIG. 1 is a block diagram showing the number setting screen
Fig. 2 is a view showing an input variable range setting screen
FIG. 3 is a view showing an output value range setting screen
4 is a graph showing the relationship between the porosity and porosity
FIG. 5 is a diagram illustrating an initial experiment condition creation screen
FIG. 6 is a graph showing an experimental condition table
Fig. 7 is a graph showing the results of experiments according to input conditions of the present invention
FIG. 8 is a graph showing an optimum condition screen

The present invention does not clearly reveal the correlation between the input conditions and the output conditions for injection molding, and the types and properties of plastics and resins used for injection are various. Therefore, even if the input conditions are not accurately known, There are many in the industrial field. In addition, even with the same injection molding apparatus, the injection conditions, such as the injection pressure, the difference between the set temperature at the injection and the temperature difference depending on the installation place, are uncertain due to the uncertainty of the input condition and the output condition, It is almost impossible to find out by using a simulation program or the like. There are several simulation programs, but in the end, it is necessary to confirm whether these results are true through actual injection. In most cases, it is rarely consistent with the simulation results. After the injection molding conditions shown in the simulation results, There is a problem to find. Before describing the present invention in detail, the terminology used in the present invention will be clarified.

I.S. (Injection Speed): Speed at which injection material is supplied at the time of injection

SD (Standard Deviation): The standard deviations of the distribution of porosity measured by cutting a sample of a certain size at the beginning, middle, and end of the injection-molded article. The larger the deviation, the greater the gap deviation. It can not be said. therefore. S.D. The value can be regarded as one of the quality indexes of the injection molding.

V. F. (Void Fraction)): porosity of the molding

N2 contents (w%): weight of nitrogen in total injection molding material%

GAS (w%):% by weight of gas in total mass including resin.

If you look at input conditions used in general injection,

Polyethylene, HDPE, LDPE, Polyvinyl chloride, Polystyrene, Acrylonitrile butadiene styrene, Polylactic acid, Polymethyl methacrylate) are different according to kinds of various plastics used as materials.

The difference in injection characteristics according to the addition amount of nucleating agent additives (azodicarbonamide, calcium carbonate, calcium stearate, magnesium silicate, sodium benzoate, stearic acid, silica products, talc, and zinc stearate, nano clay, carbon nanofibers and single walled carbon nanotubes) However,

There is a difference in injection characteristics depending on the type and addition amount of the foaming additive (mixture of CO 2, N 2, Co 2 and N 2, n-butane, water)

In addition to the above, conditions affecting the injection result include processing temp., Processing pressure and pressure drop rate, mold opening (process control by slightly opening the mold during injection molding) The shape of the mold die, the shape of the gate, the injection speed, etc., can affect the injection result as an input variable. Therefore, any one or more of the above conditions is an input variable that can be used as an input variable of the present invention.

 In addition, although the shape of the injection-molded product is of course important, the physical and chemical properties of the injection product are also important, so that the properties to be measured in the main properties of the injection product and to be managed in the production process are as follows: porosity, The output value can be a satisfactory physical property of a molded article formed by injection molding, such as non-rigidity, specific gravity, average pore size, product dimensional tolerance, shrinkage, and surface abrasion resistance (abrasion resistance). Therefore, any one or more of the porosity of the injection product, the uniform distribution of voids, the specific strength, the specific gravity, the average pore size, the dimensional tolerance of the product, the shrinkage, the surface wear characteristics (abrasion resistance) have.

The present invention shows an example of a foam injection molding method as an example, which is only an example, and the scope of use of the present invention and the rights of the present invention are not limited to the examples.

The error used as an example in the present invention is the difference between the predicted porosity and the measured porosity of the actual injection result. Therefore, if the error value is large, the mathematical modeling is wrong and the prediction and the result are different, which means that the mathematical model is wrong.

A process of finding a foam injection molding condition as an example in the present invention will be briefly described. The input condition to be used for injection is selected by selecting one or more of the input examples and selecting one or more of the output conditions Setting.

A minimum injection test condition is generated according to the range of the input condition and the number of the input condition according to the given condition. The generation of the injection test condition may be a method of dividing the range of the input condition inputted by the user by the number determined by the user or a predetermined value such as 3 or 4. Actual injection is performed in accordance with the injection test conditions thus generated, and the output value of the injection object to be measured from the result of the injection molding (FIG. 4) is actually measured. The present invention measures the porosity and standard deviation of an injection molding, for example, and completes the table (Fig. 7) together with the input condition table (Fig. 6).

From the completed table (FIG. 7), a second multiple regression analysis model is created for each input condition. Second order multiple regression analysis model is developed to find optimal injection condition satisfying user 's satisfactory output characteristics of injection molding by various input conditions.

The accuracy of the model is verified by comparing the output properties of the injection molding with the results of the second multiple regression analysis model after the actual injection molding with the optimal injection conditions for the various input conditions found, Performs the process of finding the condition.

However, in spite of the comparison with the above-mentioned experiment, if the injection molding optimization condition can not be found, the results of the physical property measurement of the molded product are additionally reflected in the table of FIG. 7 to correct the secondary multiple regression analysis model, And repeats the process of measuring and comparing the physical properties of the resultant through injection molding. This repetition allows a more accurate multiple regression model expression to be generated around the optimal value rather than the entire range, and more accurate prediction can be performed in the local range.

It is hard to predict the experimental results in the full range of the real world. It is difficult to predict the approximate conditions of the optimum values in the whole range through the first experiment. In the second and third iteration, more accurate input We want to predict the condition. In this process, it is possible to find the input condition group which affects the output value among the various input conditions, and to find the input condition which relatively and absolutely affects the output value.

We can apply the weight to the objective function so that these values can be reflected in the mathematical model expression. This weighting can be easily detected when only one input condition is changed when the other conditions are the same, so that even a technician with little experience can easily apply it. By doing so, The injection molding conditions can be found more easily than when the mathematical model alone is used.

The process of finding the mathematical model of the foam injection molding according to the present invention will be described in more detail as follows. Basically, in the present invention, a wide input condition is set and experimented, a mathematical model is created to reduce the setting range of the input condition, and the optimum input condition satisfying the output condition is searched. Repeating the injection results and the mathematical modeling, we set up more detailed injection conditions and perform the experiment and make the model again. In other words, we define a mathematical model that can predict the outcome of the results by defining the input / output related to the desired condition, and then use the mathematical model to confirm the target result as an objective function. Then, the mathematical model is compared with the mathematical model, and the mathematical model is corrected. Then, the mathematical model is checked again as an objective function and the experiment is repeated until the desired result is obtained.

The mathematical model for finding the injection conditions uses ANOVA (Analysis of Variance) and second-order polynomial regressions. The ANOVA model is a method of experimentally finding the influence of input variables on output values. The ANOVA itself is a method of verifying the statistical significance by means of differences between the groups by comparing the mean between two or more groups. In the present invention, by examining the statistical significance between the mathematical model and the actual result of injection by comparing the standard deviation of the average of the pores of the injection molding mathematically modeled with the standard deviation of the pores of the actual injection molded article, It was used as a verification method. Second-order polynomial regressions are a method of expressing input and output in a primary or secondary correlation. Using a number of mathematical models related to injection molding, the second-order multiple regression model in foam injection molding is the most optimal model .

Therefore, the present invention uses a second-order multiple regression model as an example. The standard deviation of porosity and porosity is measured after each actual injection molding, and the mathematical model is found through the second multiple regression analysis of the correlation between the input condition value and the actual injection molding result value.

The second multiple regression model can have the following form.

In addition, the objective function used in the present invention is

Lt; / RTI >

That is, when the difference between the I-th physical property value and the I-th target physical property value of the injection-molded molded product is an I-th error, the sum of all the weights considering the weight Wi which is the importance of the properties of the injection- And setting the injection conditions so that this value is minimized is an injection molding input condition optimization. In the present invention, the target property value is the porosity of the injection molding. Therefore, the error is a difference in porosity of the injection molded article under the same input conditions as the porosity obtained by the mathematical model. The weights are related to the ratios that reflect the results in the mathematical model, and in general, the weights are increased in the case of characteristics that the manufacturer considers important in the produced output.

The input conditions used as an example of the present invention are the injection shape gate shape, the GAS addition amount, the expected void (VF), and the injection speed (IS) And the standard deviation (SD) of the porosity measured in the front, middle, and back portions of the sample.

In the present invention, for example, to verify the validity of the mathematical model, S.D. was set to 1 or less and the porosity error was set to 0.5 or less.

Other input conditions include the nitrogen content in the injection material and the desired porosity.

In addition, some of the injection conditions can be set to fixed values, which can be set differently depending on the end of the injection equipment or resin used. (HDPE, SCLAIR 2710) and a density of 0.954 g / cm 3, and the end N 2 (BOC gas) of the injection gas used for injection, etc. have.

The injection molding is performed according to the injection conditions of the present invention, and the porosity and standard deviation are measured to make the table of FIG. From this table, we make the objective function as follows.

FIG. 7 is a table in which SD and Error are measured in a table in which a result of an actual injection test is measured in an injection test condition designed by inputting a range of values of Gate, Gas, VF, and IS as an input condition of the present invention. This table is analyzed by the ANOVA method, and the regression equation can be obtained by judging the correlation and the interaction between the input and the result value. Here, S.D. and the above-mentioned conditions are calculated to be the minimum error,

The optimum condition can be found for each gate.

Gate 1, Gas 0.38%, VF 30%, IS 400mm / s

Objective function value = 1.58, SD = 1.75, Error = 1.33

Gate 2, Gas 0.35%, VF 22%, IS 400 mm / s

Objective function value = 0.82, SD = 0.18, Error = 1.32

Gate 3, Gas 0.63%, VF 30%, IS 400 mm / s

Objective function value = 1.26, SD = 0.53, Error = 1.76

In this case, the object function value of gate 2 is minimized and the overall object function value is smaller than that of gate 2, so that additional experiments are performed on gate 2 in the second stage. In the first experiment, we create a new experimental condition based on the value which shows the optimal result, so that more accurate optimization value can be found in a given range.

The objective function is obtained from the above conditions and table as follows.

By calculating the optimal GAS and VF conditions from this objective function,

Can be obtained.

Injection molding is carried out under these conditions and it is confirmed whether or not it matches with the desired result. Through the repetition of the second multiple regression analysis, an inexperienced or less experienced engineer can perform the injection molding method to find the optimum injection molding condition of the foam injection molding .

Claims (2)

A search method for optimizing an injection molding condition for molding a plastic injection molded article having desired characteristics,
(S1) setting an injection input condition and setting an objective function using an output condition and a desired characteristic; And
Generating (S2) a minimum injection test condition for performing an injection test for finding an injection condition satisfying an output condition from the set input condition; The minimum test condition may be set by the user so that at least three test conditions can be set per one input condition in consideration of the minimum changeable set value and the maximum change range of the injection apparatus.
Performing injection according to the injection test conditions (S3); And
(S4) of collecting an injection sample of the injection molded product and measuring an output value corresponding to the output condition from the injection sample; And
(S5) of inputting the measured output values for each of the injection test conditions together with the input conditions, and determining a correlation between the injection test conditions and the output values (S5); And
A step (S6) of forming a second-order mathematical model expression between the input condition and the output condition through a multiple regression analysis of the correlation with the measurement result; And
(S7) calculating an optimal condition satisfying the objective function using the second-order mathematical model equation generated in step S6; And
(S8) of performing the injection by setting the injection conditions again using the calculated optimum conditions and performing the step S4 on the injected injection object; And
Completing the injection experiment (S9) if the result at step S6 is satisfied with the objective function;
Wherein the injection molding condition is optimized for molding a plastic injection molded article having a desired output value characteristic. A computer program, characterized in that the method of claim 1 is implemented on a computer.

Images