### Transcription

International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014ISSN 2229-5518750Taguchi and ANOVA analysis of shrinkage ofInjection moulded polypropylene ComponentTejendra Singh1 , Mahendra pratap Singh2, Mohd Mukhtar Alam31M.Tech Scholar, Jagannath University, Chaksu, Jaipur, India2 Assistant Professor, Jagannath University, Chaksu, Jaipur, India3 Assistant Professor, Vivekananda Institute of Technology, Jaipur, IndiaAbstract— In the present era, competition gets tougher; there is more pressure on manufacturing sectors to improve quality andcustomer satisfaction while decreasing cost and increasing productivity. These can be achieved by using modern quality managementsystems and process improvement techniques to reduce the process variability and driven waste within manufacturing process usingeffective application of statistical tools. Taguchi technique is well known technique to solve industrial problems. In this study, effect ofinjection molding parameters on the shrinkage in polypropylene (PP) is investigated. The relationship between input and output of theprocess is studied using Taguchi method and Analysis of Variance (ANOVA) technique. The selected input parameters are meltingtemperature, injection pressure, packing pressure and packing time. Effect of these parameters on the shrinkage of above mentionedmaterials is studied using mathematical modelling. The determination of optimal process parameters were based on S/N ratios.Keywords: Injection moulding, ANOVA, Taguchi analysis, Shrinkage—————————— ——————————IJSER1. IntroductionNowadays, competitive market requires producers toproduce high quality parts, with lower price in the leastpossible time. Injection molding is known as an effectiveprocess for mass production of plastic parts withcomplicated forms and high dimensional precision. Inthis method, high pressure fluid polymer is injected to thecavity with desired form. Next, under high pressure, fluidsolidifies. During the process, plastic materials are underhigh pressure and temperature. Materials are cooled to getdesired form. Injection molding process can be dividedinto four stages: Plasticization, injection, packing andcooling. Although molding process may seem simple, themolded polymers are effected by many machineparameters and process condition.Incorrect input parameters settings will cause bad qualityof surface roughness, decreases dimensional precision,Warpage, unacceptable wastes, increases lead time andcost. Therefore, finding the optimized parameters ishighly desirable. In past scientists used trials and error tofind good Process conditions but this method is time andcost Consuming. In addition, when there are a largenumber of Input parameters, these methods can’t be used.Nowadays, The model of the process and optimalcondition are developed Using analytic methods andheuristic algorithms.The study carried by Chang and Faison [1] reported thatmore shrinkage occurs across the flow direction thanalong the flow direction. Chang and Faison studied theshrinkage behavior and optimization of PS, HDPE andABS parts by using the Taguchi and ANOVA methods.They stated that the mold and melt temperatures alongwith the holding pressure and the holding time were themost significant factors affecting the shrinkage behaviorof the three materials studied. One of the main goals ininjection moulding is the improvement of quality ofmoulded parts besides the reduction of cycle time, andlower production cost. For instant, poor cooling systemwill give rise to non uniform mould surface temperatureand irrational gate location, would lead to differentialshrinkage in moulded parts [2,3]. For the effects ofprocessing parameters on shrinkage in POM injectionmoulded parts, Postawa and Koszkul [4] reported that theclamp pressure and the injection temperature were keyparameters.As in many manufacturing industry meeting requiredspecification means keeping quality under control Qualityproblems can be material related defects i.e. black specksand splay , process related such as filling related defectsi.e. flash and shots packing and cooling related defectsi.e., sink marks and voids, and post, mould related defectsi.e., warpage , dimensional changes. Vaatainen et al. [5].investigated the effect of the injection mouldingparameter on the visual quality of mouldings using theTaguchi method. They focused on the shrinkage withthree more quality characteristics: weight, weld line andsinkmarks.2 Experimental StudiesIJSER 2014http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014ISSN 2229-55182.1 MaterialsPolypropylenes were used as an amorphous and a semicrystalline polymer. The general properties of PP areshown in table 1.21222324252627Table 1 Properties of PolypropyleneDensity(g/cm3 )Melt flow index( g per 10 min )Modulus of elasticity( MPa)Charpy impact toughness( KJ/m2)0.90-0.9110.7841001.4-1.8Table 2 The process parameters and levels234FactorsLevel 1Melt200temperature, MPa)Packing6time,D(s)Level 2230Level 313121212312313231312Shrinkage is the difference between the size of moldcavity and size of finished part divided by the size of amold. The relative shrinkage of selected characteristicswere calculated with following equationS (D m - D p )/ D m x 100%Where S denotes the shrinkage, D m denotes the molddimension and D p denotes the part dimension. In thisstudy three trial of shrinkage taken and S/N ratio iscalculated by average value of the three shrinkage value.3 Results and DiscussionIJSERTable 3 The L 27 orthogonal arrayDeney No123456789101112131415161718192033333332.2 Shrinkage measurementThe experiment was conducted with four controllablethree level processing parameter: melt temperature,injection pressure, packing pressure, packing time,therefore the L 27 orthogonal array was selected for thisstudy. The process parameters and levels are shown intable 2 and the L 27 orthogonal array in table 3.S.No1751Taguchi’s philosophy is an efficient tool for design ofhigh quality manufacturing system, which has beendeveloped based on orthogonal array experiments, whichprovide much reduced variance for experiment withoptimum setting of process control parameters[8].Thesignal to noise ratio is a simple quality indicator thatresearchers and designers can use to evaluate the effectsof changing a particular design parameter on performanceof the products.[9,10] Taguchi methods [11] use a specialdesign orthogonal array to study the entire factor withonly a small number of experiments[12].It introduces anintegrated approach that is simple and efficient to find thedesigns for quality, performance and computational costIn product or process design of Taguchi method, there arethree steps:i) System design: selection of system for given objectivefunction.ii) Parameter design: selection of optimum levels ofparameteriii) Tolerance design: determination of tolerance aroundeach parameter levelTaguchi method uses the signal-to-noise (S/N) ratioinstead of average. The S/N ratio reflects both the averageand the variation of the quality characteristics [6].Asdiscussed by Oktem el at. the S/N ratio is a measure ofperformance aimed at developing products and processesinsensitive to noise factors. The standard S/N ratio usedare as follows: Nominal is best (NB), lower the better (LB)and higher the better (HB).[4]. In this study lower value ofshrinkage behavior is expected. Thus S/N ratiocharacteristics the lower – the- better is applied in theanalysis which is given in table 4 and can be calculated byusing relation.IJSER 2014http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014ISSN 2229-5518752S/N -10 Log10 ( 1/n 1/yi )260457062.5632.4272.445Where y i is the value of the quality characteristics for theith trials, n is number of repetitions.2605540102.4182.2632.32722.478Table 4.Shrinkage values and S/N ratio for ackingpressure(MPa)40Packingtime(s)6Shrinkage (%) 7.2198.2966.0947.024-2.336The response table of the S/N ratio is given in table 5, andthe best set of combination parameter can be determinedby selecting the level with highest value for each factor.As a result, the optimal process parameter combinationfor PP is A3, B2,C3,D3.The difference value given in table 5 denotes whichfactor is the most significant for shrinkage of PP molding.Packing pressure was found most effective factor for PPfollowed by packing time, injection pressure and melttemperature.Table 5 The response table for S/N ratio for PPS.No.Level 1Level 2Level 3DifferenceMelttemperature, .9157.8827.369PackingPressure,C (MPa)Packingtime 214From the given data in table 5 S/N ratio responsediagram was drawn shown in fig a,b,c,d. The highest S/Nratio for each factor gave the optimal process condition,which corresponds to melt temperature 260ºC, aninjection pressure 55 MPa, packing pressure of 70 MPaand injection time of 14 s.S/N ratio V/s Melt temperature(ºC)IJSER 2014http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014ISSN 2229-5518753in SAS/STAT software for analysis of variance. TheANOVA procedure is designed to handle balanced data(that is, data with equal numbers of observations forevery combination of the classification factors), whereasthe GLM procedure can analyze both balanced andunbalanced data. Because PROC ANOVA takes intoaccount the special structure of a balanced design, it isfaster and uses less storage than PROC GLM forbalanced data.In ANOVA calculation, the degree of freedom for allfactors needs to be obtained. The calculation for degreeof freedom is as below:S/N ratio V/s Injection PressureTotal degree of freedom,ff T N– 1 27 -1 26For Factor A ,f Af A kA – 1 3 -1 2Where k A is the number of level of factor AIJSERFor Error, f E f T –( f A f B f C f D ) 26- ( 2 2 2 2) 18Sum of squares,SS T ( Z a1 2 Z a2 2 Z a3 2 . Z aN 2) - ( Z a1 Z a2 Z a3 . Z aN )2/N (3.0572 2.5202 2.0962 . 2.3362) - (3.057 2.520 . 2.336)2 /27 175.965 – 173.756 2.215S/N ratio V/s Packing pressureFor Factor A ,S A ( A 1 )2/ k A ( A 3 )2 / k A - ( Z a1 Z a2 Z a3 . Z aN )2/N ( 3.057 . 2.281 )2 /9 (2.333 . 2.917)2 /9 (2.654 . 2.336)2/9- (3.057 . 2.336)2 /27 66.178 57.188 50.693 -173.756 0.703S/N ratio V/s Packing timeFor Factor B ,f BFB k B – 1 3 -1 2Where k B is the number of level of factor BANOVA ANALYSISThe ANOVA procedure performs analysis of variance(ANOVA) for balanced data from a extensive variety ofexperimental designs. In analysis of variance, acontinuous response variable, known as a dependentvariable, is calculated under experimental conditionsrecognized by classification variables, known asindependent variables. The variation in the response isassumed to be due to effects in the classification, withrandom error accounting for the remaining variation. TheANOVA procedure is one of several procedures availableFor Error, f E f T –( f A f B f C f D ) 26- ( 2 2 2 2) 18For Factor B,S B ( B 1 )2/K B ( B 3 )2 /K B - ( Z B1 Z B2 Z B3 . Z BN )2/N (3.057 2.478)2/9 (2.520 . 2.336)2/9 (2.096 . 2.217)2/9 - (3.057 . 2.336)2 /27 66.825 51.643 55.705– 173.751 0.422IJSER 2014http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014ISSN 2229-5518754 0.422/2 0.211For Variance Error ,Ve For Factor C ,f CF C K C – 1 3 -1 2Where K C is the number of level of factor C 0.043/18 0.00238F-ratio , F for all factors are calculated as belowFor Factor B ,F B VB / V e 0.211/.00238 88.655Percentage Contribution, P B for Factor BP B (S B/ S t ) x 100 (0.422/ 2.215) x 100 19.05 %For Error, f E f T –( f A f B f C f D ) 26- (2 2 2 2) 18For Factor CS C ( C 1 )2/K C ( C 3 )2/K C - ( Z C1 Z C2 Z C3 . Z N )2/N (3.057 . 2.336)2/9 (2.520 . 2.217)29 (2.096 . 2.478)29 - (3.057 . 2.336)2 /27 64.208 63.797 46.662- 173.751 0.916The value of Variance for factor CVC S C / fC 0.916/2 0.458For Variance Error ,Ve .043/18 0.00238F-ratio , F for all factors are calculated as belowFor Factor C ,F C VC / V e 0.458/0.00238 192.43Percentage Contribution, P C for Factor CP C (S C/ S t ) x 100 (0.916/ 2.215) x 100 41.35%For Factor D ,f DF D K D – 1 3 -1 2Where K D is the number of level of factor DFor Error, f E f T –( f A f B f C f D ) 26- ( 2 2 2 2) 18IJSERFor Factor DS D ( D 1 )2 / K D ( D 3 )2 / K D - ( Z D1 Z D2 Z D3 . Z N )2/N (3.057 . 2.478)2/9 (2.520 . 2.336)2/9 (2.096 . 2.217)2 /9- (3.057 . 2.336)2 /27 59.449 59.840 54.493 - 173.751 0.131For Error, S eS e S T - (S A S B S C S D ) 2.215 - (0.703 0.422 0.916 0.131) 0.043The values of variances for all factors are then calculated,For Factor A,VA 0.703/2 0.3515For Variance Error ,Ve 0.043/18 0.00238F-ratio , F for all factors are calculated as belowFor Factor A ,F A 0.3515/0.00238 147.68Percentage Contribution, P A for Factor AP A X 100 (.703/2.215) X 100 31.74%The value of Variance for factor BVB S B / fBThe value of Variance for factor DVD S D / fD .131/2 .0655For Variance Error ,Ve .043/18 0.002388F-ratio , F for all factors are calculated as belowFor Factor D ,F D VD / V e .0655/.00238 27.52Percentage Contribution, P D for Factor DP D (S D/ S t ) x 100 (0.131/ 2.215) x 100 mperature, A (ºC)InjectionPressure ,B(MPa)PackingPressure,C(MPa)Packingtime D .0431.95262.2151.129100Table 6ANOVA TableHence From the table it is clearly mentioned that theParameters we found in taguchi analysis is also acceptedhere.IJSER 2014http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014ISSN 2229-5518755The ANOVA table is shown in Table 9,giving thepercentage contribution and most significant factorcontributing to the shrinking of PP mold.4 References[1]K.Mao,W.Li,C.J.Hooke and D.Walton, “Friction .267,pp 639-645,2009.[2]H.Xiao Yan, L.De-Qun abd X. Qiang, “Gate locationoptimization in injection molding based on empiricalsearch method”, Material Science Forum, Vol 575578,pp55-62,2008,[3] H. Zhou, Y.Zhang J.Wen,D.Li’An acceleartionmethod for minimization of shrinkage” gy,37,pp1006-1022.[4] P.Postawa and J.k.Koszkul,” Change in injectionmolding parts shrinkage and weightas a function ofprocessing conditions”, Journal of Materials ProcessingTechnology, vol.162-163, pp.109-115,2005.[5]Vaatainen O, Pentti J. “Effect of processingparameters on quality of injection molded parts usingTaguchi parameter design method”, Plastic RubberCompos 1994;21:2117[6] Datta, Saurav; & Bandyopadhyay, Asish,& Pal PradipKumar-(2008), Grey based taguchi method foroptimization of bead geometry in submerged arc bead-onplate welding, International Journal of AdavnceManufacturing Technology,33,pp1136-1143.[7] Taguchi G.An introduction to quality engineering,Asian Productivity Organisation,1990.[8] Fowlkes WY, Creveling CM. Engineering methodsfor robust product design,using Taguchi methods intechnology and product development. Addison-WesleyPublication,1995[9] Forouraghi,B(May 2002), Worst- Case ToleranceDesign and Quality Assurance via Genetic Algoritm,Journal of optimization theory and application,Vol113,No 2,pp251-268[10] Deng, Chyn-Shu,Chin,Jih-Hua, Hole roundness indeep hole drilling as analyzed by Taguchi method,International Journal of Adavnce ManufacturingTechnology,25, pp420-426.[11] TaguchiG,.Introduction to quality engineering. NewYork;Mc Graw Hill;1990.[12]TaguchiG, Konishi S. Taguchi methods, orthogonalarrays and linear graphs , tool for qyality Americansupplier institute. American supplier Institute;1987,pp835.IJSERAuthor t,Jecrc Jaipur ,Has passed his B.E fromRajasthan University and pursuing M.Tech fromJagannath University,Jaipur. He is Working on PlasticInjection Technology for shrinkage reduction techniques.Mahendra Pratap Singh ,Asso. Prof.,JagannathUniversity,Jaipur, has passed his B.E. form Rajasthanuniversity and his Master form MNIT,Jaipur.Presently heis doing his P.Hd. He has Wide knowledge in field ofProduction and Design. He has also authored the bookson Mechanics of Solid and Workshop Technology.Mohd Mukhtar Alam has done his B.E. in MechanicalEngineering from University of Rajasthan and hisM.Tech from Rajasthan Technical University. He hasattended 4 international conferences and have writtenmany research paper in field of production Engineering.He had also written 4 books on Engineering MechanicsFluid Mechanics, Machining and Machining Tool andRenewable Energy Technology. He is now working inVivekananda Institute of Technology, Jaipur.IJSER 2014http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014ISSN 2229-5518IJSERIJSER 2014http://www.ijser.org756

injection molding parameters on the shrinkage in polypropylene (PP) is investigated. The relationship between input and output of the process is studied using Taguchi method and Analysis of Variance (ANOVA) technique. The selected input parameters are melting temperature, inject