shrinkage behavior and optimization of injection molded parts studied by the taguchi method.

TAO C. CHANG [*]
The shrinkage behavior of plastic plays a vital role in determining the final size of the injectionmolded part.
As we all know, process conditions affect many properties of plastic parts, including shrinkage.
In this study, the influence of process conditions on shrinkage was systematically studied by Tian Kou method (along-and across-the-Direction of flow)
Three kinds of plastic: high
Density Polyethylene (HDPE), general-
Polystyrene for use (GPS)
, And acrylic-butadiene-styrene (ABS).
The results show that high density polyethylene as a crystalline plastic shrinks more than GPS and ABS.
Two amorphous materials
Degree of contraction along the path to the opposite sexthe-flow andacross-the-
The flow direction of HDPE is different from that of GPS and ABS.
More happened across the street. the-
Flow direction ratio of high density polyethylene in itsalong-the-flow direction.
The opposite is true for GPS and ABS.
Mold and melting temperature and holding pressure and holding time have the greatest impact on the shrinkage behavior of the three materials, although each material has different importance to each plastic.
The experiment verified the best condition for reducing shrinkage determined by Tian Kou method and passed t-statistictests.
The prediction is very consistent with the experimental value. the-
Traffic contracted by GPS.
Ernest Felson, III [*]
Introduction injection molding uses almost all thermoplastic plastics and some thermoforming materials to produce discrete parts with complex and changeable cross sections as well as various surface textures and properties.
There are more injection molding machines for plastic processing than any other type of molding equipment.
The product quality of the injection molded plastic parts is the result of the complex combination of materials, parts and mold designs used and the process conditions used to manufacture them.
Shrinkage is one of several important factors that determine the quality of injection molding products.
Many factors, including material, processing parameters, part and mold design, can affect the shrinkage behavior.
It is generally believed that semi-crystalline plastics shrink more than amorphous plastics because of the tight accumulation of crystal structures [1].
Of course, shrinkage is reduced when filler and reinforcement are added.
For the influence of processing parameters, Jansen et al. [2]
A systematic study of seven thermoplastic plastics has been carried out, and it is found that maintaining pressure is always the key parameter.
The higher holding pressure reduces the shrinkage of the part.
Other processing parameters such as melting temperature, Mold temperature and holding time also affect the shrinkage behavior.
It is found that increasing melting temperature can reduce shrinkage [2,3]
Because of better stress transmission.
However, it is reported that increasing the mold temperature will increase the shrinkage [3,4]
Although other people have different conclusions5,6].
Len gthening holding time always reduces shrinkage until flat when the gate freezesoff time [7].
Cavity pressure is a by-product of many process variables and has also been shown to be associated with shrinkage [3,7,8].
In addition, the shrinkage of the injection-molded parts may be heterosexual.
Factors that cause complex and uneven size changes include :(i)
Asymmetric shrinkage due to uneven cooling of parts ,(ii)
The volume of the plane is not uniform ,(iii)
Behavior of Materials of the opposite sex induced by flow orientation, and (iv)
Differential thermal strain caused by geometric effect [1,9].
This study is an extension of previous research [10]
The influence of process conditions on shrinkage is systematically studied by using Tian Kou method (along-and across-the-Direction of flow)
Three kinds of plastic: high
General density polyethylene
Purpose: polystyrene and acrylic-butadiene-styrene.
The parameters studied included melting and mold temperature, holding, injection and reverse pressure, and holding and cooling times.
Experiment, several designs of Experiment (DOE)
The method can be used to determine the most important factors affecting the quality of the product and to provide the best conditions for the best quality [11].
The Tian Kou method uses an orthogonal array (OA)
As the basis of the experiment.
The advantage of using orthogonal arrays is that a large amount of information from the effects of many processing parameters can be obtained from very few experiments [12].
Methods invented by doctors
Genichi Taguchi is a unique engineering process that unifies the optimization of cost, quality and product development cycle time.
This approach greatly simplifies the complex world of DOE and gives the power of the actual experimental design to the engineer. Ealey\'s book [13]
Gives the full history of the doctor.
Tian Kou and his works and their profound influence on the United StatesS. industries.
The purpose of this study is twofold: * determine the most important processing variables that affect the shrinkage behavior of each plastic, * experiment to verify the optimal conditions for the minimum shrinkage, the processing conditions identified with Taguchimethod.
The experimental materials affect their shrinkage behavior due to the molecular structure of the material, two amorphous plastics, GPS and ABS.
And a semi-crystalline protein (HDPE)
Tests were conducted in the study.
Table 1 lists the materials selected for the experiment and their melting indices.
Four-mold injection molding
Type T mud of cavity [R]
Change the insert mold (
Precision mould Co. , Ltd. )
Used to generate rectangular bars for shrinkage measurement as recommended in ASTM 955-89.
The doors and runners are closed.
The allowable part of the folding blade is formed chosencavities (Fig. 1).
A 22-ton injection molding machine (BOY 22M)
Completely closed-
Loop microprocessor control system with mold temperature controller (Model 4410-
C of pound)
Samples were used under different conditions.
Dry ABS in 88 [with Fisher constant temperature oven]degrees]
C. 4 hours before injection.
For the experimental design of each material, first by entering the resin, machine and part geometry information into the CABsoftware package (Dr. C-
Mould of advanced CAE Technology Co. , Ltd. ).
Before the experiment, several injection tests were run according to the best window to ensure a high quality product was obtained.
The window is then used as a reference for setting up the Taguchimethod orthogonal array. A modified [L. sub. 16]
Run the molding program using four factors, four levels, three factors, two levels of OA.
Factors used in the experiment included melting temperature, Mold temperature, injection pressure, hold pressure, cooling time, return pressure and hold time.
The levels and assigned values for each resin are shown as inTables 2 and 3.
In the experiment, each treatment condition was stable for at least half an hour.
After that, 15 parts were made before sampling for analysis.
Every fifth part is wasted as a sample.
Each run, 10 samples were taken out of 50 injections.
For each injection cycle, when the injection pressure reaches the set value, a switch from the injection stage to the hold stage occurs.
Since high density polyethylene is semi-crystalline, all samples can be relaxed for at least a week after the experiment before measurement.
Samples are stored and measured at 23 [degrees]C and 50% RH.
Measurement Figure 1 illustrates how to go along-andacross-the-
Measure the flow direction on a rectangular bar.
Two measurements along the waythe-
Flow and three measurementsthe-
Use a digital card clamp with precision [to collect traffic for each sample]+or -]0. 001 inch.
Relative shrinkage is calculated with the following equation: S = ([D. sub. m]-[D. sub. p])/([D. sub. m]
X 100%, where S = shrinks; [D. sub. m]
= Mold size and 【D. sub. p]
= Part size.
Specimens 10, 20, 30, 40 and 50 for each run were measured as each plastic.
Analyze a software program called WinRobust Lite [TM][12]
The results are analyzed.
A quadratic mass loss function was developed by Dr.
Tian Kou, in order to replace the traditional ladder function, and better estimate the currency loss of manufacturers and consumers when the product performance deviates from the target value.
There are four kinds of quality loss functions [12].
For shrinkagestudy, the mass loss function of the type is-smaller-the-better case.
The program can assign factors to orthogonal arrays and calculate signalsto-
Noise ratio, perform variance analysis (ANOVA)
, Graphical results affected by production factors, and predict the best results of quality objectives.
Below, the response map is shown and the relative factor effects analyzed by the ANOVA program are discussed.
Finally, the experiment verified the best control factors found from the data analysis.
Results and discussion mean contracting in two directions. Two directions of shrinkage of all samples in 16 runs of three plastics are calculated, as shown in Table 4.
As can be seen from the table, in any direction of the process window, the average size of the semi-crystalline high density polyethylene changes the most, followed by ABS.
The GPS contract is minimal in both directions.
The results also show that all plastics have uneven shrinkage, which is different in degree and direction for different plastics.
More shrinkage of semi-crystalline polyethylenethe-Than flow alongthe-flow.
The opposite is true for GPS and ABS.
For semi-crystalline thermoplastic, the same results were found by Mamat and others.
[Polypropylene]3].
However, Thomas\'s measurement of polypropylene near the gate shows that along-the-
Flow shrinkage greater-the-
The flow shrinks, away from the gate, and the opposite is true [14].
As mentioned earlier, many factors such as uneven cooling, flow-induced orientation, and complex geometric shapes affect the behavior of uneven shrinkage.
For high density polyethylene, the flow-induced orientation is along-the-
Flow direction, which leads to tighter cross packagingthe-flow direction.
For both amorphous plastics, GPS and ABS, due to its presence on the edge-the-
The flow direction causes them to shrink more than they do horizontally. the-flow.
Tian Kou method uses orthogonal matrix to establish an experimental matrix to study the influence of various factors.
\"Orthogonal\" refers to the balance of the combination of various factors, so no one factor in the experiment has more or less weight than other factors.
\"Orthogonal\" also means that the influence of each factor can be mathematically associated with the influence of other factors.
Analysis of means (ANOM)
It is the method that Tian Kou uses to quantify the average response of each individual control factor level.
Figures 2 to 4 illustrate the effect of seven processing factors on the shrinkage of three plastics in two directions.
These response graphs show a contraction change for each factor.
GPS Figure 2 shows that increasing hold pressure, hold time, and mold and melting temperatures reduce two-way shrinkage.
However, keeping the pressure and time seems to be on the Crossthe-flow direction.
Increase injection pressure-the-
The flow shrinks, but along-the-flowdirection.
The reverse pressure shows the opposite contraction effect in two directions;
The higher reverse pressure is reduced-the-
Contraction edge of flow-the-
Traffic changes, although the magnitude is not large.
In both flow directions, longer cooling results in more shrinkage.
ABS Figure 3 shows the effect of holding pressure, melting temperature and mold temperature similar to GPS on the two-way shrinkage of ABS.
However, the effect of injection pressure is different for ABS.
This suggests that a higher injection pressure can reduce two-way shrinkage.
In addition, the effect of the back pressure and cooling time on the lateralthe-
The traffic shrinkage of ABS and GPS is the opposite.
High Density Polyethylene response diagram shown in the high density polyethylene diagram
4 compared with the results of two kinds of deformed thermoplastic, the effects of different process parameters on the shrinkage behavior of crystalline plastic are given.
Compared with GPS and ABS, the effect of melting temperature and mold temperature on HDPE shrinkage is opposite.
The higher melting temperature and mold temperature will enhance the shrinkage in both directions.
Both of these will provide higher fluidity and longer cooling time for molecular chain formation of crystal structure, thus increasing shrinkage.
However, injection pressure and cooling time have a complex effect on the shrinkage of this semi-crystalline material.
For GPS and ABS, higher hold pressure and longer hold time reduce shrinkage in both directions.
Increasing the back pressure will also reduce the shrinkage in both directions.
The Tian Kou method can not only generate a response map to illustrate the quality changes caused by each factor change, but also conduct variance analysis (ANOVA)
Enables engineers to quantitatively estimate the relative contribution of each control factor to the overall measurement response.
In the analysis of variance results, there is a term called \"Fratio\", also known as \"variable \".
\"This ratio is used to test the meaning of the factor effect.
When F is far greater than 1, due to experimental errors and interaction effects, the influence of the control factors is significant compared to the variance.
This information can be very helpful because decisions must be made about which controls are worth further spending resources to enhance the robustness of the product.
After the analysis of variance, Table 5 lists the order of the significant factors, which greatly affect the shrinkage behavior of the three plastics in two directions.
It can be seen from the table that for all plastics, mold and melting temperature, holding pressure and holding time are the most important factors affecting shrinkage in either direction.
In the validation test of the Tian Kou method, the model was added to predict the impact of the control factors on the response.
The model refers to the sum of individual factors, no crossterms(interactions).
One of the main purposes of the validation experiment is to provide evidence that the addition equation is applicable and that the interaction is low.
The applicability of the additive model cannot be simply obtained from the factor effect diagram.
The general form of the prediction equation is as follows :[Y. sub. predicted]= [y. sub. exp]+ ([y. sub. A]-[y. sub. exp])+ ([y. sub. B]-[y. sub. exp]+ ([y. sub. C]-[y. sub. exp])+ where [y. sub. exp]
, Is the overall average response of orthogonal arrays and [y. sub. A], [y. sub. B], [y. sub. C]
Responsibility for factors A, B and C, respectively.
Factor Impact Score in response to the level of factors being modeled (
Usually the best level)
Used in the prediction equation.
Then, the experimental results are obtained and compared with the predicted results.
If the results of an experiment are similar to those predicted, the experiment is successful.
If the results of the validation test are significantly different from the predictions, it is clear that interaction is important.
This interaction must be revealed and eliminated, and then experimental procedures must be planned again.
Because the quality of the contraction is \"small-the-
Better yet, \"the best combination of study factors for minimal shrinkage can be determined from the response graph.
For example, based on a graph.
2, the level of the factor that will produce the lowest contraction in the transverse contractionthe-
The flow direction of the GPS is 188 bar of hold pressure and the hold time is 15 seconds.
58 [mold temperaturedegrees]
C. injection pressure of 376 bar and melting temperature of 272 [degrees]
C, back pressure 6.
9 bars, 30 seconds of cooling time.
Table 6 summarizes the best combination of factors for the three plastics in the present study in two flow directions.
By replacing the effects of the corresponding factors shown in the response graph with the prediction equation, the predicted optimal shrinkage percentage for each case was obtained.
The experimental verification test is the same as the previous running program.
Contraction data are measured and summarized in Table 6.
In practice, it is difficult to determine that the experimental number must be close to the predicted value in order to think that the protocol is good.
The experimental means in the range of the estimated onestandard deviation of the Table Month, except for the prediction of valuefor HDPE Shun-the-flow shrinkage.
In addition, for all plastics, all experimental results are lower than those of the first 16 times.
This shows that the Tian Kou method can predict the best conditions within the scope of the study.
It also shows that it may be due to the interaction of crystals, partial geometric shapes and chain directions that make predictions along-the-flow direction.
To provide statistical confirmation, t-
Test the following assumptions with a 95% confidence interval :[H. sub. 0]: [micro]= [[micro]. sub. o]versus [H. sub. a]: [micro][neq][[micro]. sub. o]where [micro]
Is the experimental value and [[micro]. sub. o]
Is the forecast value.
The calculated t value is listed in the last row of Table 6.
The results show that only GPS alongthe-T value of traffic (1. 081)
Smaller [t. sub. 19, 0. 025]= 2.
093 so t-
Test failedH. sub. o].
Other people say their t-
Value greater [t. sub. 19,0. 025]= 2. 093 or [t. sub. 29, 0. 025]= 2.
So I refused [H. sub. o]and accepting [H. sub. a].
Statistical analysis shows that the contraction prediction of GPS under optimal conditionsthe-
The flow direction of Tian Kou method was verified by experiments.
However, the prediction of other parameters by the Tian Kou method has not been verified by experiments, and the significant interaction between the parameters studied is the possible reason.
Understanding the interaction requires more research, and additional models may need to be modified.
The t value seems to indicate that the deviation between the interaction and the additional model is directional and that the material-dependent;
More across the streetthe-
Flow direction f or GPS andABS ratio along-the-flow direction;
The opposite is true for high density polyethylene.
Conclusions several conclusions can be drawn from this study: * the semi-crystalline plastic HDPE shrinks more than the GPS and ABS of the two amorphous materials in this study.
The author thanks toMr for the research assistant.
Faison is from the Ministry of Industrial Education and Technology, the School of Education and the Graduate School of Iowa State University.
* The degree of the contraction of the opposite sex in the direction-the-flow andacross-the-
The flow direction of HDPE is different from that of GPS and ABS.
More happened across the street. the-
Flow direction ratio along-the-flow direction.
The opposite is true for GPS and ABS.
* The Tian Kou method provides a very effective tool to study the effect of process parameters on the shrinkage of injection molding parts.
Mold and melting temperature and holding pressure and holding time are the most important factors in the shrinkage behavior of the three materials, although they are of different importance to each plastic.
* Additional models are very effective in predicting GPS shrinkage behavior along the linethe-flow direction.
However, more interactive research is needed in other cases.
Confirmation (*. )
Reference materials for corresponding authors (1. )R. A.
Malloy, plastic parts design for injection molding--
Introduction, p.
80, Hanser Press, New York (1994). (2. )K. M. B. Jansen, D. J.
Van Dick and M. H.
Pauline HusseinEng. Sci. 38, 838 (1998). (3. )A. Mamat, F. Trochu, and B.
Sanglin in boremeEng. Sci, 35,1511 (1995). (4. )P. Delbarre, J. Pabior, J. -F. Daurelle, V. Lamblin, and F.
Year 301 (1991). (5. )C. Liu and L. T.
Mancine of PaulineEng. Sci. , 36, 1 (1996). (6. )M. F. Bain, Jr. , S. L. Janicki, A. S. Ulmer, and L. S.
Year 977 (1992). (7. )V. Leo and Ch.
Cuvelliez, Polym. Eng. Sci. 36, 1961 (1996). (8. )P. M. Gipson, P. F. Grelle, and B. A. Salamon, J. Inject. Molding Tech. , 3, 117(1999). (9. )H.
242, year (1991). (10. )T. C. Chang and E. Faison, III, J. Inject. Molding Tech. , 3,61 (1999). (11. )S. R. Schmidt and R. G.
Section 4. understand the experiment of industrial design.
, Colorado Springs, Colorado (1994). (12. )W. Y. Fowlkes and C. M.
Innovation and engineering methods of robust product design: using the Tian Kou method in technology and product development
Redding Wesley, Massachusetts (1995). (13. )L.
Ealey, design quality: Tian Kou method and USAS.
Industry, ASI Publishing House (1988). (14. )R. Thomas and N.
371, year, spantec (1989). [Graph omitted][Graph omitted][Graph omitted]