外文原文-對采用常規(guī)注塑與微孔注塑形成的織滾筒收縮率的研究.pdf

UNCORRECTED PROOF 1 2Shrinkage study of textile roller molded by conventional/microcellular 3injection-molding process 4Shyh-shin Hwang a, Peming P. Hsub, Chi-wei Chianga 5 a Department of Mechanical Engineering, Ching-Yun University, Chung-Li, 32023, Taiwan, ROC 6 b Department of Mechanical Engineering, Far-East University, Tainan, 74420, Taiwan, ROC 7 8Abstract 9Oneoftheadvantagesofmicrocellularconventionalinjectionmoldingoverconventionalinjectionmoldingisthattheshrinkageofthepartcanbe 10reduced. This project investigated the effect of the process parameters on the shrinkage of the textile roller by conventional/microcellular injection- 11molding process. Polybutyleneterephthalate (PBT) materials with 30 wt.% glass and Wollastonite fiber were used. The results showed that the 12shrinkage by microcellular injection molding is less than that of conventional injection molding. Glass fiber filled PBT has more shrinkage than 13Wollastonite fiber filled PBT due to the non-uniform cell size of the glass fiber filled PBT. 14 2008 Published by Elsevier Ltd. 15 16Keywords: Microcellular injection molding; Three-plate-mold; PBT; Glass fiber; Wollastonite; Shrinkage 17 181. Introduction 19The microcellular process was first introduced by N. P. Suh 201,2asabatchprocessin1980.Inthisprocess,apolymersample 21ishoused ina highpressurechamber. Aninert gas likeCO2or N2 22is introduced into the chamber, diffusing into the polymer until 23saturation. Then, the pressure is rapidly reduced while the poly- 24mertemperatureissimultaneouslyincreased,producingathermo- 25dynamic instability that lowers the gas solubility and creates cell 26growth. The disadvantage of the batch process is that a long 27period of time is required for the polymer to become saturated 28with the gas, due to low diffusion rates at room temperature. To 29avoid this problem, microcellular extrusion was developed. It 30reduces the time necessary for the gas to saturate the polymer by 31introducing the inert gas into the barrel while the polymer is still 32molten. The diffusion rate is high because the temperature in the 33barrel is high. Unfortunately, as parts become more complex, 34microcellular extrusion cannot be used to produce them. 35In response, microcellular injection molding was developed 363 and commercialized by Trexel Co. Ltd. as the Mucell 37process. The key insight of this process is the application of a 38supercritical fluid. The supercritical fluid is injected during the 39injection stage cycle, creating millions of micron-sized voids in 40otherwise solid thermoplastic polymer parts. 41Severalstudieshaveinvestigatedtheshrinkageandwarpageof 42injection molded parts. Bushko et al. 4,5 studied the effect of 43processing conditions on shrinkage, warpage, and residual 44stresses of a thermal viscoelastic melt. Their results showed that 45ahigherpackingpressureresultedinlessshrinkage.Liaoetal.6 46investigated optimal process conditions for shrinkage and war- 47page in thin-wall parts. They showed that the optimal process 48conditions differ for shrinkage and warpage in injected thin-wall 49cellular-phone covers. Recently, Kramschuster et al. 7 studied 50the shrinkage and warpage behavior of a grocery box in micro- 51cellular and conventional injection molding. They showed that 52the SCF level and injection speed are the most important factors 53affecting the shrinkage and warpage of microcellular injection 54molded parts. In the authors last paper 8, we have showed that 55the shrinkage rate of microcellular injection molding is less than 56that of convention injection molding. 57Thetextileroller(Fig.1)inthetextilemachinesisawornpart 58whichshouldbereplacedafteracertaintime.Thematerial ofthe 59roller is PBT. PBT has good dimensional stability, mechanical 60strength, stiffness, and fire retardant characteristics. To improve 61the mechanical strength, most of the plastics are filled with glass 62fiber. In this study, both glass and Wollastonite fiber filled PBT Available online at International Communications in Heat and Mass Transfer xx (2008) xxxxxx ICHMT-01731; No of Pages 9 Communicated by W.J. Minkowycz. Corresponding author. E-mail address: .tw (S. Hwang). 0735-1933/$ - see front matter 2008 Published by Elsevier Ltd. doi:10.1016/j.icheatmasstransfer.2008.02.011 ARTICLE IN PRESS Please cite this article as: S. Hwang, et al., Shrinkage study of textile roller molded by conventional/microcellular injection-molding process , Int Commun Heat Mass Transf (2008), doi:10.1016/j.icheatmasstransfer.2008.02.011 UNCORRECTED PROOF 63areinvestigated. Wollastonite isacalcium metasilicate(CaSiO3) 64which can improve thermal and dimensionalstabilityat elevated 65temperatures 9. 66In this study, we have investigated the effects of process 67parameters on the variation of the rim thickness (Fig. 2Q1) by 68conventional and microcellular injection molding. The specifica- 69tion of the rim thickness is 0.02 mm. If the rim thickness 70tolerance is out of specification, there is a running noise from the 71roller,andthetextilerollershouldbereplacedafter a certaintime. 722. Experimental work 732.1. Material 74The materials used were 30 wt.% glass fiber filled PBT and 30 wt.% 75Wollastonite fiber filled PBT. The PBT material, Shinte D202G30, was supplied 76by Shinkong Synthetic Fibers Co. The material was dried at 120 C for 3 h 77before injection molding. The PVT diagram 10 is shown in Fig. 3. 782.2. Part geometry and mold design 79A three-plate-mold with 4 cavities of a textile roller was used in this study. 80Eachcavityhas six gatesaroundthe thin sectionof the roller (Fig.4). Points1, 3, 815, 7, 9 and 11 are the gate positions (Fig. 2). There is a weld-line in between 1 82and 3 and so on. The rim thickness of the textile roller is 6.4 mm in the cavity. 832.3. Injection-molding machine 84The injection-molding machineusedwas the Arburg 420CAllrounder1000- 85350, equipped with Mucell capability. Nitrogen is used as the gas source. In this 86study, the process parameters; melt temperature, injection speed, shot size, melt 87plastification pressure (MPP), SCF level, and mold temperature; were varied to 88determine the effects on the rim thickness of the textile roller. The details of the 89process parameters are shown in Table 1. 90Experimentswerecarriedout by changing one factor ata time andkeepingthe 91others constant. The rim thickness was measured by micrometer, and the mic- 92rostructure of the foamed part was examined by scanning electron microscope 93(SEM). 942.4. Microscopy 95A SEM was used to observe the morphology of the cell structure in the 96textile roller. The cell structure in the SEM image was taken on a JEOL Fig. 1. Simplified diagram of textile roller. Fig. 2. Points 1, 3, 5, 7, 9 and 11 are the gate positions. The rim thickness is measured at red area according to the 12 points direction from center. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Fig. 3. PVT diagram of the glass fiber filled PBT material 10. 2S. Hwang et al. / International Communications in Heat and Mass Transfer xx (2008) xxxxxx ARTICLE IN PRESS Please cite this article as: S. Hwang, et al., Shrinkage study of textile roller molded by conventional/microcellular injection-molding process , Int Commun Heat Mass Transf (2008), doi:10.1016/j.icheatmasstransfer.2008.02.011 UNCORRECTED PROOF 97JSM6360. Specimens were cut into smaller pieces and gold was sputtered onto 98the surface. They were then inspected using the SEM. 993. Results and discussions 100Althoughthereare4cavitiesinonemold,onlyonecavity(#1inFig.5) 101was taken for the measurement in order to get consistent data. The rim 102thicknesswastheaverageofthefivesamples.Theexperimentwascarried 103out by short shot first and showed that the runner system has unbalanced 104meltflowproblem(Fig.5)11.Someportionsarefilledwhereassomeare 105only partially filled. To improve this problem, a modified mold design is 106neededwhichwillbementionedlater.Theinjection-moldingprocesswas 107done by the conventional method first,and then the microcellular process 108was introduced as the foaming molding method. 1093.1. The effect of process conditions on rim thickness of the textile 110roller of PBT with glass fiber by conventional injection molding 111Accordingtothespruerunnersystemofthemold(Fig.4).Thegates3 112and5(Fig.2)arefarfromtherunnerascomparedtotheothergatesonthe 113up right cavity. So there is an unbalanced melt flow problem around this 114area.Fig.6showstherimthicknessoftherollerbyconventionalmolding. 115The maximumthicknessis6.075 mmongate11.Thethicknessvariation 116is more than 0.10 mm, and points 4 and 5 have the least thickness. 1173.2. The effect of process conditions on rim thickness of the textile 118roller of PBT with glass fiber by microcellular injection molding 119It needs certain time to make the microcellular injection molding 120stable when the supercritical fluid is introduced into the barrel. Parts are 121sampled after half an hour of operation to make sure the cell is uniform. 122Fig. 7 shows the rim thickness variation of the glass fiber filled PBT by 123microcellular injection molding. The thickness variation is around 1240.04 mm, which is on the margin of the specification, and the thinnest 125occurs at point 5. The thickness trend is similar to that of solid molding, 126but the thickness variation is smaller for microcellular molding. How- 127ever the maximum thickness occurs at point 12 (weld-line position) and 128the value is 6.27 mm. This is larger than that by conventional injection 129molding and it is caused by the expansion of the cell whereas it is 130compression on conventional injection molding. So the shrinkage rate 131for microcellular injection molding is less than that of conventional 132injection molding. 1333.3. The effect of process conditions on rim thickness of the textile 134roller of PBT with mineral fiber by microcellular injection molding 135Figs. 813 show the rim thickness variation of Wollastonite filled 136PBT by microcellular molding. From the curves observed, the thickness 137variation of Wollastonite filled PBT is smaller than that of glass fiber 138filled PBT, and the curves of the Wollastonite filled PBT are smoother 139thanthose ofglassfiberfilled PBT. Forthe process conditions used, only 140the MPP has the trend whereas the MPP is increased, the rim thickness 141variation is decreased. MPP is the driving force to make cells smaller. 142Figs. 14 and 15 show the microstructure of the cell near the gate of 143glass fiber and Wollastonite fiber filled PBT. The cell structure of 144Wollastonite fiber filled PBT is more uniform than that of glass fiber 145filled PBT, and the cell size is around 10 m. 146The large rim thickness variation of glass fiber filled PBT may be 147attributed to the non-uniform cell structure. For the shrinkage rate 12, 148the average thickness of solid glass filled, foamed glass fiber filled, and 149foamed Wollastonite filled PBT is 6.004, 6.243, and 6.256 mm 150respectively. The thickness on the mold is 6.40 mm. In turn, the 151shrinkage rate is 6%, 2.4%, and 2.2% for solid glass filled, foamed glass 152fiber filled, and foamed Wollastonite filled PBT respectively. 153The method to improve the rim thickness variation is changing the 154runner design as shown in Fig. 16. By this design, the melt has more 155balanced flow characteristics. Because the mold is too complex to 156modify, computer simulation, Modex3D 10, is used to simulate the 157flow characteristics of the old and modified design. Fig. 17 shows the 158rim shrinkage of the original and modified designs. It shows that the 159modified design has less shrinkage compared to the original design. 1604. Conclusions 161The effect of the process parameters on the rim thickness of 162glass fiber and Wollastonite filled PBT by conventional and Fig. 4. The configuration of sprue, runner and gate. Table 1t1:1 Process parameters for microcellular injection-molding process of the textile rollert1:2 t1:3123 t1:4Melt temp. (C)255265275 t1:5Injection speed (cm3/s)130140150 t1:6Shot size (cm3)394143 t1:7MPP (bar)130140150 t1:8SCF level (%)0.180.280.38 t1:9Mold temp. (C)405060 Fig. 5. Short shot of the four cavities. 3S. Hwang et al. / International Communications in Heat and Mass Transfer xx (2008) xxxxxx ARTICLE IN PRESS Please cite this article as: S. Hwang, et al., Shrinkage study of textile roller molded by conventional/microcellular injection-molding process , Int Commun Heat Mass Transf (2008), doi:10.1016/j.icheatmasstransfer.2008.02.011 UNCORRECTED PROOF 163microcellular injection-molding process has been conducted. It 164has been found that: 165(1) For the rim thickness, microcellular injection molding has 166smaller thickness variation than conventional molding. 167(2) The textile roller has more uniform thickness with 168Wollastonite filled PBT than glass filled PBT. 169(3) Parts have less shrinkage by microcellular injection 170molding compared to that by conventional molding. 171(4) Wollastonite filled PBT has more uniform cell struc- 172ture but lower shrinkage rate than that of glass filled 173PBT. 174Acknowledgment 175This work was supported by National Science Council of 176Taiwan under contract # 94-2622-E231-004-CC3. Fig. 6. Rim thickness variation vs. injection speed for solid glass fiber filled PBT. Fig. 7. Rim thickness variation vs. injection speed for foamed glass fiber filled PBT. 4S. Hwang et al. / International Communications in Heat and Mass Transfer xx (2008) xxxxxx ARTICLE IN PRESS Please cite this article as: S. Hwang, et al., Shrinkage study of textile roller molded by conventional/microcellular injection-molding process , Int Commun Heat Mass Transf (2008), doi:10.1016/j.icheatmasstransfer.2008.02.011 UNCORRECTED PROOF Fig. 8. Rim thickness variation vs. melt temperature for foamed Wollastonite fiber filled PBT. Fig. 9. Rim thickness variation vs. injection speed for foamed Wollastonite fiber filled PBT. 5S. Hwang et al. / International Communications in Heat and Mass Transfer xx (2008) xxxxxx ARTICLE IN PRESS Please cite this article as: S. Hwang, et al., Shrinkage study of textile roller molded by conventional/microcellular injection-molding process , Int Commun Heat Mass Transf (2008), doi:10.1016/j.icheatmasstransfer.2008.02.011 UNCORRECTED PROOF Fig. 10. Rim thickness variation vs. MPP for foamed Wollastonite fiber filled PBT. Fig. 11. Rim thickness variation vs. SCF level for foamed Wollastonite fiber filled PBT. 6S. Hwang et al. / International Communications in Heat and Mass Transfer xx (2008) xxxxxx ARTICLE IN PRESS Please cite this article as: S. Hwang, et al., Shrinkage study of textile roller molded by conventional/microcellular injection-molding process , Int Commun Heat Mass Transf (2008), doi:10.1016/j.icheatmasstransfer.2008.02.011 UNCORRECTED PROOF Fig. 12. Rim thickness variation vs. shot size for foamed Wollastonite fiber filled PBT. Fig. 13. Rim thickness variation vs. mold temperature for foamed Wollastonite fiber filled PBT. 7S. Hwang et al. / International Communications in Heat and Mass Transfer xx (2008) xxxxxx ARTICLE IN PRESS Please cite this article as: S. Hwang, et al., Shrinkage study of textile roller molded by conventional/microcellular injection-molding process , Int Commun Heat Mass Transf (2008), doi:10.1016/j.icheatmasstransfer.2008.02.011 UNCORRECTED PROOF Fig. 15. The cell structure of the Wollastonite fiber filled PBT near the gate. Fig. 14. The cell structure of the glass fiber filled PBT near the gate. Fig. 16. More balanced design of the runner system. 8S. Hwang et al. / International Communications in Heat and Mass Transfer xx (2008) xxxxxx ARTICLE IN PRESS Please cite this article as: S. Hwang, et al., Shrinkage study of textile roller molded by conventional/microcellular injection-molding process , Int Commun Heat Mass Transf (2008), doi:10.1016/j.icheatmasstransfer.2008.02.011 UNCORRECTED PROOF 177References 1781 V. Martini, J.E., Nam. P. Suh, F.A. Waldman, U.S. Patent No. 4,473,665, 1791984. 1802 Jonathan S. Colton, Nam P. Suh, Nucleation of microcellular foam: theory 181and practice, Polymer Engineering and Science 27 (7) (1987) 500503. 1823 J.E. Martini, F.A. Waldman, N.P. Suh, Microcellular closed cell foams and 183theirmethodofmanufacture,SPEANTECTech.Papers,1982,pp.674678. 1844 W.C. Bushko, V.K. Stokes, Solidification of thermoviscoelastic melts. Part 185II: effects of processing conditions on shrinkage and residual stresses, 186Polymer Engineering and Science 35 (4) (1995) 365373. 1875 W.C. Bush