- Email: [email protected]
A novel clamping force searching method based on sensing tie-bar elongation for injection molding
International Journal of Heat and Mass Transfer 109 (2017) 223–230
Contents lists available at ScienceDirect
International Journal of Heat and Mass Transfer journal homepage: www.elsevier.com/locate/ijhmt
A novel clamping force searching method based on sensing tie-bar elongation for injection molding Ming-Shyan Huang ⇑, Cheng-You Lin Department of Mechanical and Automation Engineering, National Kaohsiung First University of Science and Technology, 2 Jhuoyue Road, Nanzih, Kaohsiung City 811, Taiwan, ROC
a r t i c l e
i n f o
Article history: Received 28 September 2016 Received in revised form 1 February 2017 Accepted 2 February 2017
Keywords: Clamping force Flash Injection molding Mold separation Short shot Tie bar elongation
a b s t r a c t Clamping force greatly influences the injection molding quality, particularly in molding thin-walled plastic parts. Low clamping on mold halves can easily cause flash defects in the part geometry, whereas high clamping can cause poor air venting that in turn causes a short shot. Therefore, using an optimal clamping force setting is crucial. However, traditional methods for estimating the clamping force for injection molding mainly use the total projected area of the cavity, sprue, and runner along the clamping direction multiplied with the predictive cavity pressure of a molten polymer. Because this prediction is rough, a maximal machine specification is commonly applied during practical operations. Thus, heavy loading on the machine and mold may generate defects on molded parts, cause extra energy consumption, and shorten the tool life. A strain sensor mounted on the tie bar can reveal the dynamics of the clamping force during injection molding. For example, tie-bar elongation increases during mold filling and packing when the high-pressure molten polymer acts on the mold halves. This study developed a novel searching algorithm based on information about tie-bar elongation with various clamping force settings to identify the proper clamping force value to set. An experimental verification shows that the clamping force determined using the proposed method feasibly improves the injection molding quality. Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction Injection molding is a cyclic process that consists of four phases: filling, melt compressing (or packing), holding, and cooling. As shown by the typical cavity pressure profile in Fig. 1, the pattern from Points A to C represents the filling stage, in which a molten polymer enters the mold cavities, following which the cavity pressure is gradually increased according to the applied injection pressure. The filling phase is completed at Point C, where the cavity is only volumetrically filled by the molten polymer without being compressed. The packing process then starts, and the pressure increases rapidly to its peak value (Pmax) at Point D, which has the greatest effect on the resisting mold clamping force introduced by the clamping mechanism in the whole injection molding process. The molten polymer within the cavity is then maintained at a set pressure during the holding phase. Additional molten polymer can be packed into the cavity to compensate for plastic shrinkage caused by cooling, thus ensuring that the mold is completely filled. This process continues until the gate is frozen, as marked at Point E. This is followed by the final cooling phase,
⇑ Corresponding author. E-mail address: [email protected] (M.-S. Huang). http://dx.doi.org/10.1016/j.ijheatmasstransfer.2017.02.004 0017-9310/Ó 2017 Elsevier Ltd. All rights reserved.
and it continues until the end of the cycle. During this phase, the melt solidifies gradually as the coolant that circulates within the cooling channels in the mold removes heat. The cooling and solidification rates determine the rate at which the cavity pressure decreases [1]. The main parameters influencing the quality of injectionmolded parts include the injection speed, melt and mold temperatures, filling–packing switchover, packing pressure and time, and extent of mold separation at various clamping force values [2–4]. In particular, mold separation occurs as excessive force is applied on the mold walls because of the high cavity pressure at the end of the filling, which momentarily exceeds the operating clamping force. This situation is often serious when high injection pressure is required in a thin-walled molding [5–8]. The invisible mold separation that elongates the tie bars of the injection molding machine may cause flash defects, resulting in inconsistent part weight and thickness. Additionally, asymmetric mold separation may dramatically increase the damage caused to the mold and reduce its lifetime. Previous studies have revealed that cavity pressure is strongly associated with the degree of mold separation in injection molding, and these factors determine the part quality. For instance, Chen et al. [9] installed linear variable differential transformer sensors on each corner of mold halves to detect their displacement and
224
M.-S. Huang, C.-Y. Lin / International Journal of Heat and Mass Transfer 109 (2017) 223–230
Fig. 1. Typical cavity pressure profile.
found that mold separation is positively correlated with part weight. To isolate the injection molding quality from the change in mold and plastic materials, they built an adaptive controller to detect adequate velocity/pressure switchover timing on the basis of mold separation. The injection molding quality can be predicted by monitoring the mold separation, which is strongly correlated with cavity pressure [10]. Fei et al. [11] studied the optimization of process parameters by using a neural network and genetic algorithm and concluded that the warpage of injection-molded parts depends on the clamping force. Satoshi et al. [12] found that the volumetric shrinkage of parts is associated with the clamping force and suggested that high injection and holding pressures combined with high clamping force are beneficial for achieving a minimal shrinkage rate. To prevent mold separation during mold filling and packing, mold clamping is typically performed at the maximum specification of the mold-clamping mechanism that achieves secure clamping without the occurrence of defects such as flash. However, with the application of excessive mold clamping force to the mold, mold deterioration is accelerated and energy consumption is increased unnecessarily. Moreover, such excessive clamping force settings stain and damage the surfaces of mold cavities, and insufficient gas venting leads to the formation of weld lines, burns, and black streaks. By contrast, clamping a mold with the required minimum clamping force, that is, the proper clamping force, can extend the mold life, reduce energy consumption, and prevent the occurrence of defects in injection-molded parts [13–15]. The conventional prediction of the minimal clamping force for preventing the mold from opening during injection is based on the estimation of the applied injection pressure for injecting a specific molten polymer into a mold cavity multiplied with the projected area of the part and sprue-runner-gate system along the clamping direction. This calculation is rough, and it neglects the increase in the clamping force required resulting from the asymmetric layout of the cavity and gates. Most people setting the magnitude of the clamping force on the controller choose the maximum specification permitted by the injection molding machine. This habit causes a venting problem that may generate short shots and increase the damage to mechanical components of the toggle clamping unit that, in particular, causes tie-bar breakage. Using a high clamping force in an injection mold shortens its lifetime.
Currently, press-on strain sensors on tie bars are used for measuring the surface strain directly at the mounting location, in a manner similar to bonded strain gauges such as tie-bar strain sensors can be used to measure the clamping force. The strain gauges are pressed under a stainless steel protective foil wrapped tightly on the cylindrical surface of the tie bar to be measured.
ei ¼
Fi EA
ð1Þ
Fi ¼
EAei 109 9:81
ð2Þ
F¼
n X Fi
ð3Þ
i¼1
where ei is the stress of the ith tie bar in micrometers, E is Young’s modulus of the tie bar (=210,000 kgf/cm2), A is the cross-sectional area of a single tie bar in squared millimeters, Fi and F respectively represent the ith tie bar and the total clamping force in tons, and n is the number of tie bars. On the basis of an accurate measurement of the clamping force acting on mold halves during injection molding with tie-bar strain sensors, an injection molding machine can detect the variation of tie-bar elongation online. The minimal clamping force for achieving high injection quality with low energy consumption and machine and mold damage can be precisely estimated easily. This study proposes an intelligent clamping force searching method that quickly and precisely suggests a proper clamping force setting value. Various experimental case studies verify the feasibility of this method. 2. Status of mold separation Fig. 2 shows three mold conditions with different clamping force settings during injection molding: S1, S2, and S3. S1 is a state in which even upon a reduction in the clamping force setting, the largest increment in the applied clamping force does not change during the interval. In this state, because the clamping force with respect to the injection pressure is sufficiently high, the mold is compressed; that is, the mold height between the movable and
M.-S. Huang, C.-Y. Lin / International Journal of Heat and Mass Transfer 109 (2017) 223–230
225
pressure until switchover because high injection pressure is still applied after volumetric filling. Over-packing adds weight and stress to the part and makes demolding more difficult. An alternative approach is to reduce the injection pressure; nonetheless, overly low pressure can cause defects such as sink marks. A more effective solution is to switch over earlier. However, switching over too early may produce an under-packed cavity characterized by a pressure drop in the compression phase. Part of the filling then occurs at lower holding pressure, and screw advancement subsequently increases the pressure. As a result, the injection part can be easily rejected because its dimensions are reduced, it is underweight, or it shows sink marks. 2.3. Cavity layout and gate location Fig. 2. Status of mold separation (S1: top; S2: middle; S3: bottom).
the stationary platens decreases. In S2, the clamping force is set at a relatively lower value compared with that in S1; therefore, the injection mold opens gradually and the tie-bar elongation during mold filling and packing increases. In this state, the mold height between the movable and the stationary platens increases slightly. S3 is a state in which the clamping force setting is reduced to a certain level such that the mold halves are opened significantly to generate a flash defect. In this state, tie-bar elongation appears significantly. In particular, increasing the clamping force applied on the injection mold results in an increment in the tie-bar elongation during mold filling and packing. The clamping force setting should ideally be between S1 and S2, as shown in Fig. 3. The mold temperature, filling–packing switchover, cavity layout, and gate location can affect the magnitude of the clamping force acting on the injection mold. 2.1. Mold temperature The variation of the mold temperature greatly influences the tie-bar elongation, resulting in variation in the clamping force. Moreover, the commonly used approach for isolating the influence of the mold temperature on the clamping force setting is to perform mold adjustment in practical injection molding operations. 2.2. Filling–packing switchover Inadequate switchover-to-holding conditions such as (1) switchover occurring too late and (2) switchover occurring too early may affect the cavity pressure. The former causes an overpacked cavity characterized by a pressure peak in the compression phase. The peak pressure does not decrease to the lower holding
Fig. 3. Maximum increment of clamping force and mold conditions with respect to clamping force settings.
The cavity layout and gate location may affect the magnitude and distribution of the cavity pressure during mold filling and packing and then cause different mold separation and tie-bar elongation at each corner of the mold. For example, in symmetric layouts with two cavities, mold separation and tie-bar elongation at the upper portion are more highly correlated with the part thickness compared with those at the lower portion, because for an injection molding machine, the top of the stationary platen is less rigid than the bottom. Furthermore, those located near the gates are more highly correlated with the part thickness compared with those located far from the gates. In addition, in asymmetric cavity layouts, the part thickness is highly correlated with mold separation and tie-bar elongation on the filling side. Nevertheless, the correlation coefficients of the other side are small. This nonuniform mold separation may be due to two factors: an asymmetric cavity layout and nonuniform strength of the clamping structure. On the basis of our previous investigation on detecting the influence of the clamping force on the injection-molding quality, a minimum of two tie-bar sensors should be installed on the injection molding machine, and the location is suggested to be mounted at the top portion related to the stationary platen. This finding was applied in developing the clamping force searching method based on sensing the tie-bar elongation. 3. Clamping force searching method based on sensing tie-bar elongation The proposed method aims to quickly and systematically identify a proper clamping force setting that is free of defects and minimizes energy consumption. This approach mainly consists of four steps: Step 1: Acquire tie bar signals Ta and Tb with the clamping force set at 50% of the machine specification. Choose one as a reference signal that satisfies MAX {DTa, DTb}, where Ta and Tb are the voltage signals of the strain sensors mounted on the upper two tie bars (Fig. 4). Step 2: On the basis of Step 1, compute DTa (or DTb) for various clamping force settings from 20% to 100% of the machine specification in increments of 10%. Step 3: Define two threshold values a and b (a < b) related to the slope of the maximal clamping force increment with respect to the clamping force setting (Fig. 5). Perform linear regression of the data set with y = ax + b. If the slope a is greater than b, choose 100% of the machine specification as the predicted value. If the slope a is smaller than a, choose 50% of the machine specification as the predicted value. If the slope a is between a and b, proceed to Step 4. Step 4: Find two lines that best the fit data and use their intersection as the predicted value (Fig. 6).
226
M.-S. Huang, C.-Y. Lin / International Journal of Heat and Mass Transfer 109 (2017) 223–230
In Step 3, the two threshold values a and b are used to identify whether the collected data are in condition S1 or S3. For example, if a is smaller than a, meaning that the mold is in compression status (S1) corresponding to a clamping force setting of 20–100%, the strategy is to suggest 50% of the machine specification as the proper clamping force setting; in this case, the machine specification may be excessive for clamping the mold. When a is larger than b, meaning that the mold is in significant separation (S3) corresponding to clamping force setting of 20–100%, the strategy is to suggest 100% of the machine specification as a proper clamping force setting; in this case, the machine specification may be insufficient for clamping the mold.
4. Case studies
Fig. 4. Locations of mold cavities, gates, and sensors with respect to stationary platens.
Fig. 5. Case of required clamping force beyond machine specifications.
Fig. 6. Case of required clamping force within machine specifications.
In this study, we used a dog-bone-shaped specimen mold that allows a combination of various cavity layouts and gate locations to examine the feasibility of our approach for identifying a proper clamping force setting. The mold was tested with different plastic materials (polycarbonate (PC) and Acrylonitrile Butadiene Styrene (ABS)) and other machines. To investigate the effect of the clamping force setting on the quality of parts fabricated using thin-walled injection molding, a measuring system that detects mold separation and tie-bar elongation during injection molding and a two-cavity injection mold for molding a dog-bone-shaped specimen were used in this study. Fig. 4 shows the employed injection mold having dimensions of 300 mm 350 mm 270 mm, in which four pairs of sensors for detecting tie-bar elongation (Ta–Td) and mold separation (Ma–Md) are installed on each tie bar and in each corner of the mold; here, Ta–Td are tie-bar strain sensors (GE1029, Gefran Corp., Germany) and Ma–Md are inductive sensors (IF6029, IFM Corp., US). A dog-bone-shaped specimen with a 1.5-mm thickness and 125-mm length was designed to analyze injection molding quality by referring to the test specimens in ASTM D638 (Fig. 7). In addition, PC (PC-110, Chi-Mei Corp., Taiwan) was used. To monitor the mold temperature profile during injection-molding experiments, two thermocouples (T1 and T2) were mounted on the surface of the male mold (Fig. 8). In addition, the specimen thicknesses were measured at Pa–Pd. To investigate the influence of filling balance on the mold separation, tie elongation, and part thickness, each cavity was equipped with two edge gates at each end of the specimen to allow or block the passage of the polymer melt from the runner. Therefore, the mold could be adjusted as a single cavity or as two cavities, and each cavity could be filled from a single gate or two gates. The experiments were conducted on an all-electric injection molding machine (ROBOSHOT S-2000i100B, Fanuc, Japan) with a clamping force of 100 tonf. As mentioned before, mold separation typically occurs during the period from the end of the filling stage to the end of the packing and holding stage, and its extent depends on the cavity size and materials employed because these factors influence the mold separation force. Notably, mold separation is often tolerated with an acceptable range (e.g., 75 lm) without causing flash defects. Such mold separation may cause added tie-bar extension, which in turn increases the predetermined clamping force during injection molding. Fig. 9 shows the variation in clamping force with respect to the period of injection molding in which the mold is filled with molten polymer from the injection unit. The tie-bar elongation is further developed rapidly as the molten polymer is filled, and the increase is affected by the polymer pressure. In particular, the effect of the polymer pressure becomes relatively large when the clamping force is set to a low value. On the basis of the calculation of the clamping force given in Eq. (3), additional tie-bar elongations can be treated as increments of the clamping force during injection molding.
227
M.-S. Huang, C.-Y. Lin / International Journal of Heat and Mass Transfer 109 (2017) 223–230
Fig. 7. Geometry of tested specimens.
Table 1 Experimental combinations. Two-cavity filling:
Single-cavity filling:
C1C2GaGbGcGd
C1GaGc
C1C2GaGb
C1Ga
Fig. 8. Positions for measuring thickness of injection-molded parts.
Fig. 9. Typical clamping force profile during one cycle of injection molding.
This study first investigated the effects of mold-filling patterns on the cavity pressure distribution that may generate nonuniform tie-bar elongation and mold separation at each corner. Specifically, the filling pattern of the molten polymer into a mold cavity affects the pressure distribution within the cavity, resulting in the generation of an anticlamping force that acts on the mold halves. Consequently, mold separation and tie-bar elongation occur in the filling and packing stage. Notably, the pressure distribution within a mold can be symmetric or asymmetric, depending mainly on the cavity layout and gate locations. In this study, four filling patterns for the dog-bone-shaped specimen were examined to determine their influence on the elongation of each tie bar and on the mold separation at each corner. On the basis of the symmetric and asymmetric layouts of the mold cavities, the filling patterns are categorized as (1) balanced two-cavity layouts and (2) unbalanced one-cavity layouts. Additionally, the cavities can be filled from a single gate
or from multiple gates. The types of filling patterns investigated in this study are denoted by CxCyGxGy, where Cx and Gx represent Cavity x and Gate x, respectively. For example, C1Ga indicates Cavity 1 with Gate a employed, and C1C2GaGb indicates Cavities 1 and 2 with Gates a and b employed. Table 1 lists the combinations employed in this study for experimentally evaluating the effect of flow balance, categorized as two-cavity filling and singlecavity filling, on the quality of injection-molded parts. Twocavity filling involves two types of gate locations used commonly in practice, namely a four-gate system (C1C2GaGbGcGd) and a two-gate system (C1C2GaGb). Single-cavity filling also involves two types of gate locations (C1GaGb and C1Ga). The effects of the five filling patterns mentioned previously on mold separation and tie-bar elongation were examined and compared according to the injection molding quality. Notably, C1C2GaGbGcGd is widely considered the most balanced sprue-runner-gate system, and it is expected to achieve high injection molding quality. By contrast, C1Ga is a relatively unbalanced mold-filling pattern, and it is expected to result in poor injection molding quality. Table 2 lists the material and processing parameters. 4.1. Computer simulations This study used Moldex3D R13 commercial simulation software to simulate the cavity pressure distribution immediately after mold filling and packing with respect to Cases C1C2GaGbGcGd, C1C2GaGb, C1GaGc, and C1Ga. Consequently, the corresponding mold separation patterns could be predicted and then compared with the experimental results.
228
M.-S. Huang, C.-Y. Lin / International Journal of Heat and Mass Transfer 109 (2017) 223–230
Table 2 Material and processing parameters (PC). Material
Specification
Manufacturer
Polycarbonate
PC-110
Chi-Mei Corp.
Temperature parameters
Unit
Values
Melt temperature Mold temperature Ejection temperature Solidification temperature Room temperature
°C °C °C °C °C
295 80 145 165 25
Filling parameters
Unit
Values
Injection speed Injection pressure
mm/s MPa
100 240
Holding parameters
Unit
Values
Holding time Holding pressure
s MPa
5 150
Cooling parameters
Unit
Values
Cooling time
s
20
Clamping parameters
Unit
Values
Initial clamping force
tonf
80
Fig. 10(a) depicts the simulation result of Case C1C2GaGbGcGd regarding the cavity pressure distribution after mold filling and packing. Because of the symmetry of the cavity layouts and gate locations, the central gravity of the total force acting on the mold is located at the central point of the mold. Fig. 10(b) depicts the simulation result of Case C1C2GaGb regarding the cavity pressure distribution after mold filling and packing. The right and left cavity pressures are symmetric because of the symmetry of the cavity layouts. However, the cavity pressures near and far from the gates are quite different. Compared with the former case, in this case, the central gravity of the total force acting on the mold is shifted upward to a point 24.9% higher than the central point of the mold. Fig. 10(c) depicts the simulation result of Case C1GaGc regarding the cavity pressure distribution after mold filling and packing. Because the single-cavity layout in the mold is asymmetric, in this case, the central gravity of the total force acting on the mold is shifted to a point 6.3% leftward from the central point of the mold. Fig. 10(d) depicts the simulation result of Case C1Ga regarding the cavity pressure distribution after mold filling and packing. Because the single-cavity layout in the mold is asymmetric and
(a) C1C2GaGbGcGd
(b) C1C2GaGb
(c) C1GaGc
(d) C1Ga
Fig. 10. Cavity pressure distribution after mold filling and packing.
229
M.-S. Huang, C.-Y. Lin / International Journal of Heat and Mass Transfer 109 (2017) 223–230 Table 5 Material and processing parameters (ABS).
Fig. 11. Maximum clamping force increment and part thickness with respect to various clamping force set values in Case C1C2GaGbGcGd.
Table 3 Predicted clamping forces (PC materials; Fanuc ROBOSHOT S-2000i100B). Cavity layout and gate location
Reference set value (tonf)
Actual acting value (tonf)
Predicted set value (tonf)
C1C2GaGbGcGd C1C2GaGb C1GaGc C1Ga
60 60 60 70
60.5 60.3 59.4 67.7
60 60 59 64
Table 4 Predicted clamping forces (PC materials; FCS AF-110). Cavity layout and gate location
Reference set value (tonf)
Actual acting value (tonf)
Predicted set value (tonf)
C1C2GaGbGcGd C1C2GaGb C1GaGc C1Ga
60 60 60 70
64.5 64.5 64.5 73.3
72 72 54 72
Material
Specification
Manufacturer
Polycarbonate
PA756
Chi-Mei Corp.
Temperature parameters
Unit
Values
Melt temperature Mold temperature Ejection temperature Solidification temperature Room temperature
°C °C °C °C °C
205 60 100 120 25
Filling parameters
Unit
Values
Injection speed Injection pressure
mm/s MPa
100 230
Holding parameters
Unit
Values
Holding time Holding pressure
s MPa
5 150
Cooling parameters
Unit
Values
Cooling time
s
20
Clamping parameters
Unit
Values
Initial clamping force
tonf
100
Table 6 Predicted clamping forces (ABS materials; Fanuc ROBOSHOT S-2000i100B). Cavity layout and gate location
Reference set value (tonf)
Actual acting value (tonf)
Predicted set value (tonf)
C1C2GaGbGcGd C1C2GaGb C1GaGc C1Ga
60 50 50 50
61.3 52.4 52.0 51.9
50 50 58 43
has one gate, in this case, the central gravity of the total force acting on the mold is shifted upward to a point 23.6% higher and 7.1% leftward from the central point of the mold.
prediction of the proper clamping force is close to the observed clamping force settings that result in high-quality injection molding. Moreover, the more symmetric the cavity layout and gate location, the more accurately the proper clamping force can be predicted. This study also tested ABS polymer materials (Table 5), which have low viscosity and are beneficial to thin-walled molding, to evaluate the feasibility of the proposed method. The experimental results shown in Table 6 indicate that the prediction of the proper clamping force is also feasible for these materials.
4.2. Experimental verification
5. Conclusions
This study initially used Case C1C2GaGbGcGd as an example to illustrate the experimental testing and results. We collected the maximal increment in tie-bar elongation corresponding to various clamping force settings of 20%–100% of the machine specification in increments of 10%. These observations were then transformed into the maximal increment of the clamping force. Furthermore, the part thicknesses at Pa–Pd were measured using a micrometer. Fig. 11 shows the relationship of the maximal increment in the clamping force measured at Ta and the part thickness corresponding to various clamping force settings. When the setting is lower than 30 tonf, flash defects occur in the plastic part. By contrast, the increment and part thickness tend to be stably low when the setting is higher than 60 tonf. On the basis of the experimental observations, we conclude that the proper clamping force is 60 tonf. This is used as a reference to evaluate the predicted value obtained using the proposed method (see Table 3). Tables 4 and 5 show the clamping force settings predicted using PC and injection molding with different machines (Fanuc ROBOSHOT S-2000i100B and FCS AF-110). The results indicate that the
On the basis of a previous series of experiments that was conducted to systematically examine the effect of the clamping force setting on tie-bar elongation, mold separation, and injection molding quality, this study proposes a novel searching method for identifying the proper clamping force. The method is based on information about tie-bar elongation sensed during mold filling and packing, and it is reliable. The conclusions of this study are summarized as follows: (1) The thin-walled dog-shaped specimen was molded under various clamping force settings, and the experimental observation shows that the part thickness tends to be stably uniform under high clamping force. (2) For asymmetric cavity layouts or gate locations, the elongations of the four tie bars are not uniform and cause asymmetric mold separation that may dramatically increase damage to the mold and reduce its lifetime. Additionally, the minimal clamping force required to ensure part quality is higher than that expected.
230
M.-S. Huang, C.-Y. Lin / International Journal of Heat and Mass Transfer 109 (2017) 223–230
(3) The proposed method was experimentally verified using a special mold that allows the combination of various cavity layouts and gate locations, different polymer materials, and injection molding machines. The results indicate that the predicted values of the proper clamping force setting are consistent with the ideal reference values.
References [1] M.S. Huang, Cavity pressure based grey prediction of the filling-to-packing switchover point for injection molding, J. Mater. Process. Technol. 183 (2007) 419–424. [2] M.S. Huang, T.Y. Lin, An innovative regression-model based searching method for setting the robust injection molding parameters, J. Mater. Process. Technol. 198 (2008) 436–444. [3] M.S. Huang, T.Y. Lin, Simulation of a regression-model & PCA based searching method developed for setting the robust injection molding parameters of multi-quality characteristics, Int. J. Heat Mass Transf. 51 (2008) 5828–5837. [4] S.C. Nian, M.H. Lee, M.S. Huang, Warpage control of headlight lampshades fabricated using external gas-assisted injection molding, Int. J. Heat Mass Transf. 86 (2015) 358–368. [5] S.C. Chen, E.L. Liaw, P.L. Su, M.H. Chung, Investigation of mold plate separation in thin-wall injection molding, Adv. Polym. Technol. 22 (2003) 306–319. [6] M.S. Huang, J.C. Yu, Y.Z. Lin, Effect of rapid mold surface induction heating on the replication ability of microinjection molding light-guided plates with v-grooved microfeatures, J. Appl. Polym. Sci. 118 (2010) 3058–3065.
[7] M.S. Huang, C.F. Chung, Injection molding and injection compression molding of thin-walled light-guided plates with v-grooved microfeatures, J. Appl. Polym. Sci. 121 (2011) 1151–1159. [8] S.C. Nian, C.Y. Wu, M.S. Huang, Warpage control of thin-walled injection molding using local mold temperatures, Int. Commun. Heat Mass 61 (2015) 102–110. [9] Z. Chen, L.S. Turng, K.K. Wang, Adaptive online quality control for injectionmolding by monitoring and controlling mold separation, Polym. Eng. Sci. 46 (2006) 569–580. [10] H.Y. Wong, K.T. Fung, F. Gao, Development of a transducer for in-line and through cycle monitoring of key process and quality variables in injection molding, Polym. Eng. Sci. 141 (2008) 712–722. [11] Y. Fei, M. Huajie, H. Lin, A hybrid of back propagation neural network and genetic algorithm for optimization of injection molding process parameters, Mater. Des. 32 (2011) 3457–3464. [12] S. Kitayama, S. Natsume, Multi-objective optimization of volume shrinkage and clamping force for plastic injection molding via sequential approximate optimization, Simul. Model Pract. Theory 48 (2014) 35–44. [13] M.H. Chiang, Y.P. Yeh, F.L. Yang, Y.N. Chen, Integrated control of clamping force and energy-saving in hydraulic injection moulding machines using decoupling fuzzy sliding-mode control, Int. J. Adv. Manuf. Technol. 27 (2005) 53–62. [14] Y. Kudo, T. Hakoda, N.K. Nakamura, S. Morozumi, Method of Setting Mold Clamping Force of Injection Molding Machine, US Patent US 2010/0109182 A1, 2010. [15] M.S. Huang, K.Y. Chen, R.F. Fung, Comparison between mathematical modeling and experimental identification of a spatial slider-crank mechanism, Appl. Math. Model 34 (2010) 2059–2073.
Contents lists available at ScienceDirect
International Journal of Heat and Mass Transfer journal homepage: www.elsevier.com/locate/ijhmt
A novel clamping force searching method based on sensing tie-bar elongation for injection molding Ming-Shyan Huang ⇑, Cheng-You Lin Department of Mechanical and Automation Engineering, National Kaohsiung First University of Science and Technology, 2 Jhuoyue Road, Nanzih, Kaohsiung City 811, Taiwan, ROC
a r t i c l e
i n f o
Article history: Received 28 September 2016 Received in revised form 1 February 2017 Accepted 2 February 2017
Keywords: Clamping force Flash Injection molding Mold separation Short shot Tie bar elongation
a b s t r a c t Clamping force greatly influences the injection molding quality, particularly in molding thin-walled plastic parts. Low clamping on mold halves can easily cause flash defects in the part geometry, whereas high clamping can cause poor air venting that in turn causes a short shot. Therefore, using an optimal clamping force setting is crucial. However, traditional methods for estimating the clamping force for injection molding mainly use the total projected area of the cavity, sprue, and runner along the clamping direction multiplied with the predictive cavity pressure of a molten polymer. Because this prediction is rough, a maximal machine specification is commonly applied during practical operations. Thus, heavy loading on the machine and mold may generate defects on molded parts, cause extra energy consumption, and shorten the tool life. A strain sensor mounted on the tie bar can reveal the dynamics of the clamping force during injection molding. For example, tie-bar elongation increases during mold filling and packing when the high-pressure molten polymer acts on the mold halves. This study developed a novel searching algorithm based on information about tie-bar elongation with various clamping force settings to identify the proper clamping force value to set. An experimental verification shows that the clamping force determined using the proposed method feasibly improves the injection molding quality. Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction Injection molding is a cyclic process that consists of four phases: filling, melt compressing (or packing), holding, and cooling. As shown by the typical cavity pressure profile in Fig. 1, the pattern from Points A to C represents the filling stage, in which a molten polymer enters the mold cavities, following which the cavity pressure is gradually increased according to the applied injection pressure. The filling phase is completed at Point C, where the cavity is only volumetrically filled by the molten polymer without being compressed. The packing process then starts, and the pressure increases rapidly to its peak value (Pmax) at Point D, which has the greatest effect on the resisting mold clamping force introduced by the clamping mechanism in the whole injection molding process. The molten polymer within the cavity is then maintained at a set pressure during the holding phase. Additional molten polymer can be packed into the cavity to compensate for plastic shrinkage caused by cooling, thus ensuring that the mold is completely filled. This process continues until the gate is frozen, as marked at Point E. This is followed by the final cooling phase,
⇑ Corresponding author. E-mail address: [email protected] (M.-S. Huang). http://dx.doi.org/10.1016/j.ijheatmasstransfer.2017.02.004 0017-9310/Ó 2017 Elsevier Ltd. All rights reserved.
and it continues until the end of the cycle. During this phase, the melt solidifies gradually as the coolant that circulates within the cooling channels in the mold removes heat. The cooling and solidification rates determine the rate at which the cavity pressure decreases [1]. The main parameters influencing the quality of injectionmolded parts include the injection speed, melt and mold temperatures, filling–packing switchover, packing pressure and time, and extent of mold separation at various clamping force values [2–4]. In particular, mold separation occurs as excessive force is applied on the mold walls because of the high cavity pressure at the end of the filling, which momentarily exceeds the operating clamping force. This situation is often serious when high injection pressure is required in a thin-walled molding [5–8]. The invisible mold separation that elongates the tie bars of the injection molding machine may cause flash defects, resulting in inconsistent part weight and thickness. Additionally, asymmetric mold separation may dramatically increase the damage caused to the mold and reduce its lifetime. Previous studies have revealed that cavity pressure is strongly associated with the degree of mold separation in injection molding, and these factors determine the part quality. For instance, Chen et al. [9] installed linear variable differential transformer sensors on each corner of mold halves to detect their displacement and
224
M.-S. Huang, C.-Y. Lin / International Journal of Heat and Mass Transfer 109 (2017) 223–230
Fig. 1. Typical cavity pressure profile.
found that mold separation is positively correlated with part weight. To isolate the injection molding quality from the change in mold and plastic materials, they built an adaptive controller to detect adequate velocity/pressure switchover timing on the basis of mold separation. The injection molding quality can be predicted by monitoring the mold separation, which is strongly correlated with cavity pressure [10]. Fei et al. [11] studied the optimization of process parameters by using a neural network and genetic algorithm and concluded that the warpage of injection-molded parts depends on the clamping force. Satoshi et al. [12] found that the volumetric shrinkage of parts is associated with the clamping force and suggested that high injection and holding pressures combined with high clamping force are beneficial for achieving a minimal shrinkage rate. To prevent mold separation during mold filling and packing, mold clamping is typically performed at the maximum specification of the mold-clamping mechanism that achieves secure clamping without the occurrence of defects such as flash. However, with the application of excessive mold clamping force to the mold, mold deterioration is accelerated and energy consumption is increased unnecessarily. Moreover, such excessive clamping force settings stain and damage the surfaces of mold cavities, and insufficient gas venting leads to the formation of weld lines, burns, and black streaks. By contrast, clamping a mold with the required minimum clamping force, that is, the proper clamping force, can extend the mold life, reduce energy consumption, and prevent the occurrence of defects in injection-molded parts [13–15]. The conventional prediction of the minimal clamping force for preventing the mold from opening during injection is based on the estimation of the applied injection pressure for injecting a specific molten polymer into a mold cavity multiplied with the projected area of the part and sprue-runner-gate system along the clamping direction. This calculation is rough, and it neglects the increase in the clamping force required resulting from the asymmetric layout of the cavity and gates. Most people setting the magnitude of the clamping force on the controller choose the maximum specification permitted by the injection molding machine. This habit causes a venting problem that may generate short shots and increase the damage to mechanical components of the toggle clamping unit that, in particular, causes tie-bar breakage. Using a high clamping force in an injection mold shortens its lifetime.
Currently, press-on strain sensors on tie bars are used for measuring the surface strain directly at the mounting location, in a manner similar to bonded strain gauges such as tie-bar strain sensors can be used to measure the clamping force. The strain gauges are pressed under a stainless steel protective foil wrapped tightly on the cylindrical surface of the tie bar to be measured.
ei ¼
Fi EA
ð1Þ
Fi ¼
EAei 109 9:81
ð2Þ
F¼
n X Fi
ð3Þ
i¼1
where ei is the stress of the ith tie bar in micrometers, E is Young’s modulus of the tie bar (=210,000 kgf/cm2), A is the cross-sectional area of a single tie bar in squared millimeters, Fi and F respectively represent the ith tie bar and the total clamping force in tons, and n is the number of tie bars. On the basis of an accurate measurement of the clamping force acting on mold halves during injection molding with tie-bar strain sensors, an injection molding machine can detect the variation of tie-bar elongation online. The minimal clamping force for achieving high injection quality with low energy consumption and machine and mold damage can be precisely estimated easily. This study proposes an intelligent clamping force searching method that quickly and precisely suggests a proper clamping force setting value. Various experimental case studies verify the feasibility of this method. 2. Status of mold separation Fig. 2 shows three mold conditions with different clamping force settings during injection molding: S1, S2, and S3. S1 is a state in which even upon a reduction in the clamping force setting, the largest increment in the applied clamping force does not change during the interval. In this state, because the clamping force with respect to the injection pressure is sufficiently high, the mold is compressed; that is, the mold height between the movable and
M.-S. Huang, C.-Y. Lin / International Journal of Heat and Mass Transfer 109 (2017) 223–230
225
pressure until switchover because high injection pressure is still applied after volumetric filling. Over-packing adds weight and stress to the part and makes demolding more difficult. An alternative approach is to reduce the injection pressure; nonetheless, overly low pressure can cause defects such as sink marks. A more effective solution is to switch over earlier. However, switching over too early may produce an under-packed cavity characterized by a pressure drop in the compression phase. Part of the filling then occurs at lower holding pressure, and screw advancement subsequently increases the pressure. As a result, the injection part can be easily rejected because its dimensions are reduced, it is underweight, or it shows sink marks. 2.3. Cavity layout and gate location Fig. 2. Status of mold separation (S1: top; S2: middle; S3: bottom).
the stationary platens decreases. In S2, the clamping force is set at a relatively lower value compared with that in S1; therefore, the injection mold opens gradually and the tie-bar elongation during mold filling and packing increases. In this state, the mold height between the movable and the stationary platens increases slightly. S3 is a state in which the clamping force setting is reduced to a certain level such that the mold halves are opened significantly to generate a flash defect. In this state, tie-bar elongation appears significantly. In particular, increasing the clamping force applied on the injection mold results in an increment in the tie-bar elongation during mold filling and packing. The clamping force setting should ideally be between S1 and S2, as shown in Fig. 3. The mold temperature, filling–packing switchover, cavity layout, and gate location can affect the magnitude of the clamping force acting on the injection mold. 2.1. Mold temperature The variation of the mold temperature greatly influences the tie-bar elongation, resulting in variation in the clamping force. Moreover, the commonly used approach for isolating the influence of the mold temperature on the clamping force setting is to perform mold adjustment in practical injection molding operations. 2.2. Filling–packing switchover Inadequate switchover-to-holding conditions such as (1) switchover occurring too late and (2) switchover occurring too early may affect the cavity pressure. The former causes an overpacked cavity characterized by a pressure peak in the compression phase. The peak pressure does not decrease to the lower holding
Fig. 3. Maximum increment of clamping force and mold conditions with respect to clamping force settings.
The cavity layout and gate location may affect the magnitude and distribution of the cavity pressure during mold filling and packing and then cause different mold separation and tie-bar elongation at each corner of the mold. For example, in symmetric layouts with two cavities, mold separation and tie-bar elongation at the upper portion are more highly correlated with the part thickness compared with those at the lower portion, because for an injection molding machine, the top of the stationary platen is less rigid than the bottom. Furthermore, those located near the gates are more highly correlated with the part thickness compared with those located far from the gates. In addition, in asymmetric cavity layouts, the part thickness is highly correlated with mold separation and tie-bar elongation on the filling side. Nevertheless, the correlation coefficients of the other side are small. This nonuniform mold separation may be due to two factors: an asymmetric cavity layout and nonuniform strength of the clamping structure. On the basis of our previous investigation on detecting the influence of the clamping force on the injection-molding quality, a minimum of two tie-bar sensors should be installed on the injection molding machine, and the location is suggested to be mounted at the top portion related to the stationary platen. This finding was applied in developing the clamping force searching method based on sensing the tie-bar elongation. 3. Clamping force searching method based on sensing tie-bar elongation The proposed method aims to quickly and systematically identify a proper clamping force setting that is free of defects and minimizes energy consumption. This approach mainly consists of four steps: Step 1: Acquire tie bar signals Ta and Tb with the clamping force set at 50% of the machine specification. Choose one as a reference signal that satisfies MAX {DTa, DTb}, where Ta and Tb are the voltage signals of the strain sensors mounted on the upper two tie bars (Fig. 4). Step 2: On the basis of Step 1, compute DTa (or DTb) for various clamping force settings from 20% to 100% of the machine specification in increments of 10%. Step 3: Define two threshold values a and b (a < b) related to the slope of the maximal clamping force increment with respect to the clamping force setting (Fig. 5). Perform linear regression of the data set with y = ax + b. If the slope a is greater than b, choose 100% of the machine specification as the predicted value. If the slope a is smaller than a, choose 50% of the machine specification as the predicted value. If the slope a is between a and b, proceed to Step 4. Step 4: Find two lines that best the fit data and use their intersection as the predicted value (Fig. 6).
226
M.-S. Huang, C.-Y. Lin / International Journal of Heat and Mass Transfer 109 (2017) 223–230
In Step 3, the two threshold values a and b are used to identify whether the collected data are in condition S1 or S3. For example, if a is smaller than a, meaning that the mold is in compression status (S1) corresponding to a clamping force setting of 20–100%, the strategy is to suggest 50% of the machine specification as the proper clamping force setting; in this case, the machine specification may be excessive for clamping the mold. When a is larger than b, meaning that the mold is in significant separation (S3) corresponding to clamping force setting of 20–100%, the strategy is to suggest 100% of the machine specification as a proper clamping force setting; in this case, the machine specification may be insufficient for clamping the mold.
4. Case studies
Fig. 4. Locations of mold cavities, gates, and sensors with respect to stationary platens.
Fig. 5. Case of required clamping force beyond machine specifications.
Fig. 6. Case of required clamping force within machine specifications.
In this study, we used a dog-bone-shaped specimen mold that allows a combination of various cavity layouts and gate locations to examine the feasibility of our approach for identifying a proper clamping force setting. The mold was tested with different plastic materials (polycarbonate (PC) and Acrylonitrile Butadiene Styrene (ABS)) and other machines. To investigate the effect of the clamping force setting on the quality of parts fabricated using thin-walled injection molding, a measuring system that detects mold separation and tie-bar elongation during injection molding and a two-cavity injection mold for molding a dog-bone-shaped specimen were used in this study. Fig. 4 shows the employed injection mold having dimensions of 300 mm 350 mm 270 mm, in which four pairs of sensors for detecting tie-bar elongation (Ta–Td) and mold separation (Ma–Md) are installed on each tie bar and in each corner of the mold; here, Ta–Td are tie-bar strain sensors (GE1029, Gefran Corp., Germany) and Ma–Md are inductive sensors (IF6029, IFM Corp., US). A dog-bone-shaped specimen with a 1.5-mm thickness and 125-mm length was designed to analyze injection molding quality by referring to the test specimens in ASTM D638 (Fig. 7). In addition, PC (PC-110, Chi-Mei Corp., Taiwan) was used. To monitor the mold temperature profile during injection-molding experiments, two thermocouples (T1 and T2) were mounted on the surface of the male mold (Fig. 8). In addition, the specimen thicknesses were measured at Pa–Pd. To investigate the influence of filling balance on the mold separation, tie elongation, and part thickness, each cavity was equipped with two edge gates at each end of the specimen to allow or block the passage of the polymer melt from the runner. Therefore, the mold could be adjusted as a single cavity or as two cavities, and each cavity could be filled from a single gate or two gates. The experiments were conducted on an all-electric injection molding machine (ROBOSHOT S-2000i100B, Fanuc, Japan) with a clamping force of 100 tonf. As mentioned before, mold separation typically occurs during the period from the end of the filling stage to the end of the packing and holding stage, and its extent depends on the cavity size and materials employed because these factors influence the mold separation force. Notably, mold separation is often tolerated with an acceptable range (e.g., 75 lm) without causing flash defects. Such mold separation may cause added tie-bar extension, which in turn increases the predetermined clamping force during injection molding. Fig. 9 shows the variation in clamping force with respect to the period of injection molding in which the mold is filled with molten polymer from the injection unit. The tie-bar elongation is further developed rapidly as the molten polymer is filled, and the increase is affected by the polymer pressure. In particular, the effect of the polymer pressure becomes relatively large when the clamping force is set to a low value. On the basis of the calculation of the clamping force given in Eq. (3), additional tie-bar elongations can be treated as increments of the clamping force during injection molding.
227
M.-S. Huang, C.-Y. Lin / International Journal of Heat and Mass Transfer 109 (2017) 223–230
Fig. 7. Geometry of tested specimens.
Table 1 Experimental combinations. Two-cavity filling:
Single-cavity filling:
C1C2GaGbGcGd
C1GaGc
C1C2GaGb
C1Ga
Fig. 8. Positions for measuring thickness of injection-molded parts.
Fig. 9. Typical clamping force profile during one cycle of injection molding.
This study first investigated the effects of mold-filling patterns on the cavity pressure distribution that may generate nonuniform tie-bar elongation and mold separation at each corner. Specifically, the filling pattern of the molten polymer into a mold cavity affects the pressure distribution within the cavity, resulting in the generation of an anticlamping force that acts on the mold halves. Consequently, mold separation and tie-bar elongation occur in the filling and packing stage. Notably, the pressure distribution within a mold can be symmetric or asymmetric, depending mainly on the cavity layout and gate locations. In this study, four filling patterns for the dog-bone-shaped specimen were examined to determine their influence on the elongation of each tie bar and on the mold separation at each corner. On the basis of the symmetric and asymmetric layouts of the mold cavities, the filling patterns are categorized as (1) balanced two-cavity layouts and (2) unbalanced one-cavity layouts. Additionally, the cavities can be filled from a single gate
or from multiple gates. The types of filling patterns investigated in this study are denoted by CxCyGxGy, where Cx and Gx represent Cavity x and Gate x, respectively. For example, C1Ga indicates Cavity 1 with Gate a employed, and C1C2GaGb indicates Cavities 1 and 2 with Gates a and b employed. Table 1 lists the combinations employed in this study for experimentally evaluating the effect of flow balance, categorized as two-cavity filling and singlecavity filling, on the quality of injection-molded parts. Twocavity filling involves two types of gate locations used commonly in practice, namely a four-gate system (C1C2GaGbGcGd) and a two-gate system (C1C2GaGb). Single-cavity filling also involves two types of gate locations (C1GaGb and C1Ga). The effects of the five filling patterns mentioned previously on mold separation and tie-bar elongation were examined and compared according to the injection molding quality. Notably, C1C2GaGbGcGd is widely considered the most balanced sprue-runner-gate system, and it is expected to achieve high injection molding quality. By contrast, C1Ga is a relatively unbalanced mold-filling pattern, and it is expected to result in poor injection molding quality. Table 2 lists the material and processing parameters. 4.1. Computer simulations This study used Moldex3D R13 commercial simulation software to simulate the cavity pressure distribution immediately after mold filling and packing with respect to Cases C1C2GaGbGcGd, C1C2GaGb, C1GaGc, and C1Ga. Consequently, the corresponding mold separation patterns could be predicted and then compared with the experimental results.
228
M.-S. Huang, C.-Y. Lin / International Journal of Heat and Mass Transfer 109 (2017) 223–230
Table 2 Material and processing parameters (PC). Material
Specification
Manufacturer
Polycarbonate
PC-110
Chi-Mei Corp.
Temperature parameters
Unit
Values
Melt temperature Mold temperature Ejection temperature Solidification temperature Room temperature
°C °C °C °C °C
295 80 145 165 25
Filling parameters
Unit
Values
Injection speed Injection pressure
mm/s MPa
100 240
Holding parameters
Unit
Values
Holding time Holding pressure
s MPa
5 150
Cooling parameters
Unit
Values
Cooling time
s
20
Clamping parameters
Unit
Values
Initial clamping force
tonf
80
Fig. 10(a) depicts the simulation result of Case C1C2GaGbGcGd regarding the cavity pressure distribution after mold filling and packing. Because of the symmetry of the cavity layouts and gate locations, the central gravity of the total force acting on the mold is located at the central point of the mold. Fig. 10(b) depicts the simulation result of Case C1C2GaGb regarding the cavity pressure distribution after mold filling and packing. The right and left cavity pressures are symmetric because of the symmetry of the cavity layouts. However, the cavity pressures near and far from the gates are quite different. Compared with the former case, in this case, the central gravity of the total force acting on the mold is shifted upward to a point 24.9% higher than the central point of the mold. Fig. 10(c) depicts the simulation result of Case C1GaGc regarding the cavity pressure distribution after mold filling and packing. Because the single-cavity layout in the mold is asymmetric, in this case, the central gravity of the total force acting on the mold is shifted to a point 6.3% leftward from the central point of the mold. Fig. 10(d) depicts the simulation result of Case C1Ga regarding the cavity pressure distribution after mold filling and packing. Because the single-cavity layout in the mold is asymmetric and
(a) C1C2GaGbGcGd
(b) C1C2GaGb
(c) C1GaGc
(d) C1Ga
Fig. 10. Cavity pressure distribution after mold filling and packing.
229
M.-S. Huang, C.-Y. Lin / International Journal of Heat and Mass Transfer 109 (2017) 223–230 Table 5 Material and processing parameters (ABS).
Fig. 11. Maximum clamping force increment and part thickness with respect to various clamping force set values in Case C1C2GaGbGcGd.
Table 3 Predicted clamping forces (PC materials; Fanuc ROBOSHOT S-2000i100B). Cavity layout and gate location
Reference set value (tonf)
Actual acting value (tonf)
Predicted set value (tonf)
C1C2GaGbGcGd C1C2GaGb C1GaGc C1Ga
60 60 60 70
60.5 60.3 59.4 67.7
60 60 59 64
Table 4 Predicted clamping forces (PC materials; FCS AF-110). Cavity layout and gate location
Reference set value (tonf)
Actual acting value (tonf)
Predicted set value (tonf)
C1C2GaGbGcGd C1C2GaGb C1GaGc C1Ga
60 60 60 70
64.5 64.5 64.5 73.3
72 72 54 72
Material
Specification
Manufacturer
Polycarbonate
PA756
Chi-Mei Corp.
Temperature parameters
Unit
Values
Melt temperature Mold temperature Ejection temperature Solidification temperature Room temperature
°C °C °C °C °C
205 60 100 120 25
Filling parameters
Unit
Values
Injection speed Injection pressure
mm/s MPa
100 230
Holding parameters
Unit
Values
Holding time Holding pressure
s MPa
5 150
Cooling parameters
Unit
Values
Cooling time
s
20
Clamping parameters
Unit
Values
Initial clamping force
tonf
100
Table 6 Predicted clamping forces (ABS materials; Fanuc ROBOSHOT S-2000i100B). Cavity layout and gate location
Reference set value (tonf)
Actual acting value (tonf)
Predicted set value (tonf)
C1C2GaGbGcGd C1C2GaGb C1GaGc C1Ga
60 50 50 50
61.3 52.4 52.0 51.9
50 50 58 43
has one gate, in this case, the central gravity of the total force acting on the mold is shifted upward to a point 23.6% higher and 7.1% leftward from the central point of the mold.
prediction of the proper clamping force is close to the observed clamping force settings that result in high-quality injection molding. Moreover, the more symmetric the cavity layout and gate location, the more accurately the proper clamping force can be predicted. This study also tested ABS polymer materials (Table 5), which have low viscosity and are beneficial to thin-walled molding, to evaluate the feasibility of the proposed method. The experimental results shown in Table 6 indicate that the prediction of the proper clamping force is also feasible for these materials.
4.2. Experimental verification
5. Conclusions
This study initially used Case C1C2GaGbGcGd as an example to illustrate the experimental testing and results. We collected the maximal increment in tie-bar elongation corresponding to various clamping force settings of 20%–100% of the machine specification in increments of 10%. These observations were then transformed into the maximal increment of the clamping force. Furthermore, the part thicknesses at Pa–Pd were measured using a micrometer. Fig. 11 shows the relationship of the maximal increment in the clamping force measured at Ta and the part thickness corresponding to various clamping force settings. When the setting is lower than 30 tonf, flash defects occur in the plastic part. By contrast, the increment and part thickness tend to be stably low when the setting is higher than 60 tonf. On the basis of the experimental observations, we conclude that the proper clamping force is 60 tonf. This is used as a reference to evaluate the predicted value obtained using the proposed method (see Table 3). Tables 4 and 5 show the clamping force settings predicted using PC and injection molding with different machines (Fanuc ROBOSHOT S-2000i100B and FCS AF-110). The results indicate that the
On the basis of a previous series of experiments that was conducted to systematically examine the effect of the clamping force setting on tie-bar elongation, mold separation, and injection molding quality, this study proposes a novel searching method for identifying the proper clamping force. The method is based on information about tie-bar elongation sensed during mold filling and packing, and it is reliable. The conclusions of this study are summarized as follows: (1) The thin-walled dog-shaped specimen was molded under various clamping force settings, and the experimental observation shows that the part thickness tends to be stably uniform under high clamping force. (2) For asymmetric cavity layouts or gate locations, the elongations of the four tie bars are not uniform and cause asymmetric mold separation that may dramatically increase damage to the mold and reduce its lifetime. Additionally, the minimal clamping force required to ensure part quality is higher than that expected.
230
M.-S. Huang, C.-Y. Lin / International Journal of Heat and Mass Transfer 109 (2017) 223–230
(3) The proposed method was experimentally verified using a special mold that allows the combination of various cavity layouts and gate locations, different polymer materials, and injection molding machines. The results indicate that the predicted values of the proper clamping force setting are consistent with the ideal reference values.
References [1] M.S. Huang, Cavity pressure based grey prediction of the filling-to-packing switchover point for injection molding, J. Mater. Process. Technol. 183 (2007) 419–424. [2] M.S. Huang, T.Y. Lin, An innovative regression-model based searching method for setting the robust injection molding parameters, J. Mater. Process. Technol. 198 (2008) 436–444. [3] M.S. Huang, T.Y. Lin, Simulation of a regression-model & PCA based searching method developed for setting the robust injection molding parameters of multi-quality characteristics, Int. J. Heat Mass Transf. 51 (2008) 5828–5837. [4] S.C. Nian, M.H. Lee, M.S. Huang, Warpage control of headlight lampshades fabricated using external gas-assisted injection molding, Int. J. Heat Mass Transf. 86 (2015) 358–368. [5] S.C. Chen, E.L. Liaw, P.L. Su, M.H. Chung, Investigation of mold plate separation in thin-wall injection molding, Adv. Polym. Technol. 22 (2003) 306–319. [6] M.S. Huang, J.C. Yu, Y.Z. Lin, Effect of rapid mold surface induction heating on the replication ability of microinjection molding light-guided plates with v-grooved microfeatures, J. Appl. Polym. Sci. 118 (2010) 3058–3065.
[7] M.S. Huang, C.F. Chung, Injection molding and injection compression molding of thin-walled light-guided plates with v-grooved microfeatures, J. Appl. Polym. Sci. 121 (2011) 1151–1159. [8] S.C. Nian, C.Y. Wu, M.S. Huang, Warpage control of thin-walled injection molding using local mold temperatures, Int. Commun. Heat Mass 61 (2015) 102–110. [9] Z. Chen, L.S. Turng, K.K. Wang, Adaptive online quality control for injectionmolding by monitoring and controlling mold separation, Polym. Eng. Sci. 46 (2006) 569–580. [10] H.Y. Wong, K.T. Fung, F. Gao, Development of a transducer for in-line and through cycle monitoring of key process and quality variables in injection molding, Polym. Eng. Sci. 141 (2008) 712–722. [11] Y. Fei, M. Huajie, H. Lin, A hybrid of back propagation neural network and genetic algorithm for optimization of injection molding process parameters, Mater. Des. 32 (2011) 3457–3464. [12] S. Kitayama, S. Natsume, Multi-objective optimization of volume shrinkage and clamping force for plastic injection molding via sequential approximate optimization, Simul. Model Pract. Theory 48 (2014) 35–44. [13] M.H. Chiang, Y.P. Yeh, F.L. Yang, Y.N. Chen, Integrated control of clamping force and energy-saving in hydraulic injection moulding machines using decoupling fuzzy sliding-mode control, Int. J. Adv. Manuf. Technol. 27 (2005) 53–62. [14] Y. Kudo, T. Hakoda, N.K. Nakamura, S. Morozumi, Method of Setting Mold Clamping Force of Injection Molding Machine, US Patent US 2010/0109182 A1, 2010. [15] M.S. Huang, K.Y. Chen, R.F. Fung, Comparison between mathematical modeling and experimental identification of a spatial slider-crank mechanism, Appl. Math. Model 34 (2010) 2059–2073.