Materials and Design 55 (2014) 381–386

Contents lists available at ScienceDirect

Materials and Design

journal homepage: www.elsevier .com/locate /matdes

Technical Report

Numerical simulation analysis of the in-cavity residual stressdistribution of lignocellulosic (wood) polymer composites used inshallow thin-walled parts formed by the injection moulding process

0261-3069/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.matdes.2013.09.041

⇑ Corresponding author at: Department of Mechanical and ManufacturingEngineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia.Tel.: +60 3 89466318; fax: +60 3 86567122.

E-mail address: sapuan@upm.edu.my (S.M. Sapuan).

M.D. Azaman a,b, S.M. Sapuan a,c,⇑, S. Sulaiman a, E.S. Zainudin a,c, A. Khalina d,e

a Department of Mechanical and Manufacturing Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysiab School of Manufacturing Engineering, Universiti Malaysia Perlis, 02600 Arau, Perlis, Malaysiac Laboratory of Biocomposite Technology, Institute of Tropical Forestry and Forest Products (INTROP), Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysiad Department of Biological and Agricultural Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysiae Aerospace Malaysia Innovation Center, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history:Received 17 June 2013Accepted 15 September 2013Available online 26 September 2013

In this paper, a numerical analysis of in-cavity residual stress formation in the thin-walled parts of injec-tion-moulded parts is presented by considering the residual stresses produced during the post-fillingstage. Injection moulding of shallow thin-walled parts with a thickness of 0.7 mm was performed usinglignocellulosic polymer composites (PP + 50 wt% wood), and the parts have been systematically investi-gated using simulation results from Autodesk MoldFlow Insight� software. In-cavity residual stressesconstitute the primary stage for analysis because of the need to control the quality of moulded partsto prevent problems with shrinkage and warpage. The analysis showed that the cooling times and pack-ing times had a less significant effect; nevertheless, the optimal levels that are required to be used in themoulding process for thin-walled parts yielded better results. The in-cavity residual stress results showthat the stress variation across the thickness exhibits a high tensile stress at the part surface, whichchanges to a low tensile stress peak value close to the surface, with the core region experiencing a par-abolic tensile stress peak. The optimum parameter ranges for obtaining the minimum in-cavity residualstresses are as follows: a mould temperature of 40–45 �C, a cooling time of 20–30 s, a packing pressure of0.85Pinject, and a packing time of 15–20 s.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Injection moulding is one of the most versatile production tech-niques used in plastic part manufacturing. Recently, manyresearchers have used lignocellulosic filler [1–6] to replace fibreglass fillers in thermoplastic composites applied to plastic partsmanufacturing.

The processing of lignocellulosic thermoplastic composites isusually limited to low temperatures below 230 �C to minimisefibre degradation [7]. Processing at low temperatures will makeit difficult for the polymer melt to flow in the mould cavities andultimately affects the inconsistent residual stress distributions inmoulded products, particularly in thin parts.

Azaman et al. [5] found that the use of shallow, thin-walledparts is preferable for moulding composite materials with lignocel-lulosic polymers because of the resulting low residual stress andwarpage. They found that shallow, thin-walled parts are structur-

ally more rigid than flat, thin-walled parts and that they can beprocessed more economically using less material.

The sources of residual stresses can be discussed in terms ofresidual flow stresses and residual thermal stresses. The residualstress is the stress left inside the moulded product under thecondition of no external loads but inconsistent temperature andpressure loads. Wang and Young [8] have briefly explained thatflow-induced stresses caused by polymer chains depend on theorientation and freeze-off packing pressure, whereas thermallyinduced residual stresses are due to the inhomogeneous coolingof moulded parts.

The characteristic residual stress distribution across the thick-ness of a part reveals tensile stresses in the surface and coreregions and compressive stresses in the intermediate region, as de-scribed by Kim et al. [9]. Zhou et al. [10] reported similar findings;however, they found that the residual stress distribution through-out the thickness of a part remains nearly constant along the flowpath and that the final residual stresses near the gate are lowerthan are those at the end of the flow path.

The residual stresses occur during the moulding and demoul-ding processes [11]. During the moulding process, the residual

Fig. 1. A shallow, thin-walled part.

382 M.D. Azaman et al. / Materials and Design 55 (2014) 381–386

stresses are generated due to the high pressure, temperaturechanges, and relaxation of polymer chains. After demoulding, theinjection-moulded parts undergo shrinkage and warpage causedby residual stresses and temperature changes. These stresses arecaused mainly by a non-uniform temperature distribution profilein the moulded part during the filling, packing, and cooling stagesof the moulding process [12].

Postawa and Kwiatkowski [13] proved that the hold pressureand injection temperature were significant parameters affectingthe residual stress distribution, and the higher injection tempera-ture and lower hold pressure seem to be more favourable. Altanand Yurci [14] investigated some moulding parameters, such asmelt temperature, mould temperature, and cooling time, and theireffects on residual stresses in surface regions of the parts. Theyfound that the most important parameters for residual stresses insurface regions of the PS and HDPE parts were the melt tempera-ture and the mould temperature, respectively. Shen and Li [15]found that mould cooling has a significant effect on part warpageand residual stresses and that mould-cooling parameters, such asmould temperature, resin temperature and cooling channelsshould be set carefully.

Residual stresses are usually developed during solidification inthe post-filling stage. The stresses can be either flow induced orthermally induced. However, in absolute values, the flow-inducedstresses are usually one order of magnitude smaller than the ther-mal stresses [16]. Because of that, Shen and Li [15] focused theirstudies on developing an optimal mould-cooling channel designto achieve uniform mould cooling to reduce the thermally inducedstress. Choi and Im [11] also studied specific parameters at thepacking and cooling stages to analyse the residual stresses. Thisis because the pressure at the filling stage is much lower than thatat the packing stage; thus, the pressure has little effect on theshrinkage and warpage of parts.

Zhou and Li [17] observed that residual stresses that accumu-lated during the post-filling and cooling stages will lead to warpagein parts after demoulding. During the packing stage, the frozen-instress caused by the packing pressure should be taken into accountwhen measuring residual stresses. Similarly, during the coolingstages, due the low thermal conductivity of polymers and the dif-ference in temperature between the molten resin and the mould,an uneven temperature field arises, particularly along the gap-wisedirection. This non-uniform temperature field distribution ulti-mately leads to differential shrinkage, thermal residual stress,and warpage in moulded parts [18].

This review indicates that limited research has been con-ducted on the concentration effect of post-filling parameters onresidual stresses in shallow thin-walled parts and the use of lig-nocellulosic fibre-reinforced thermoplastic composite materialsfor injection moulding processes. Hence, this study investigatedthe potential processability of shallow, thin-walled parts com-posed of lignocellulosic fibre-reinforced polymer compositematerials with respect to in-cavity residual stresses. Neverthe-less, the limitation of this research is lack of direct experimentalvalidation of results from numerical simulation. Focusing onnumerical simulation initially helped researchers to understandabout the behaviour residual stress distribution which causedshrinkage and warpage defects. Similarly, Young [19] also usenumerical simulation with applied a simple thermoviscoelasticmodel to calculate the residual stress developed in the post fill-ing stage for thin rectangular cover. While, Sen and Bhattacharya[20] investigated the residual stresses distribution on starch/syn-thetic polymer blends just using experimental rather than simu-lation for justified that significant parameter is injectionpressure, mould temperature and packing pressure. Therefore,the recognized material data properties from database of numer-ical software, comparison findings from literatures and also

simulation results from analysis on post-filling stage able tobring these finding into conclusion.

2. Methodology

2.1. Part design, modelling, and simulation

Autodesk Inventor Professional� was used to model the thin-walled moulded parts, as shown in Fig. 1. A 3D design of a shallow,thin-walled part was created. The part had a length, width, andthickness of 55 mm � 50 mm � 0.7 mm but exhibited differentfeatures. Autodesk MoldFlow Insight� was used to simulate andanalyse the injection moulding process. A meshing model wasdeveloped, as shown in Fig. 2. The injection moulding machineand material used in the simulation adhered to the followingspecifications: Argburg Allrounder 370c 88 ton injection mouldingmachine (30 mm) and (PP + 50 wt% wood composite) from IsoformLip CPCW50, Isokon. The process parameters, such as mould tem-perature, cooling time, packing pressure, and packing time, aresome of the most significant parameters affecting the residualstresses based on reviews from previous research findings. Theprocessing post-filling parameters for simulations are shown inTable 1. The fixed parameters used for further detailed analysisof the post-filling parameters are shown in Table 2. The simulationwas performed using set analysis (Fill + Cool + Fill + Pack + Warp)for this models. The results indicate that the in-cavity residualstress lies along the first principle direction (plotted on the centreof the surface locations).

3. Results and discussion

In-cavity residual stresses in the first principal direction de-scribe the stress inherent along the orientation direction before apart is ejected from a mould. These residual stresses occur insidea mould and will be altered after a part is ejected. However, the re-sults obtained in this study are sufficient to relate and describepart quality in terms of the predicted shrinkage and warpage[21,22]. The positive and negative values obtained from the simu-lation indicate residual tensile and compressive stresses, respec-tively [21]. Figs. 3–6 show the simulated residual stresses in thethickness direction at the centre of the thin-walled part surface un-der different moulding parameters. The thickness in the plot wasnormalised from �1 to 1 according to the core and cavity sidesof the mould. The maximum residual tensile stresses occur nearthe wall in the core and cavity mould. The maximum residualstresses decrease when approaching a part thickness of0.175 mm and increase again near a thickness of 0.35 mm for bothsides of the parts. The distribution pattern of the residual stresses

Feed system Thin wall part

Cooling channels

Coolant inlet

Fig. 2. A meshing model for simulation.

Table 1Post-filling processing parameter settings.

Parameters Values

Mould surface temperature 40–60 �CCooling time 10–50 sPacking pressure 45–85%Packing time 10–30 s

Table 2The fixed processing parameter settings used for analysis of the effect of some post-filling parameters.

Parameters Values

Injection time 1 sMould surface temperature 45 �CMelt temperature 185 �CEjection temperature 104 �CPacking time 10 sPacking pressure 85%Cooling time 20 s

M.D. Azaman et al. / Materials and Design 55 (2014) 381–386 383

through the thickness of the part is in agreement with the patternsreported in previous studies [9,10].

Fig. 3 shows the effect of different mould temperatures from 40to 60 �C on the distribution of residual stresses through the thick-ness at the midpoint of part surface. Increasing the mould temper-ature causes increases in the tensile residual stresses near the wallfrom 21.65–30.79 MPa for both sides of the part. Similarly, residualstresses at the core region of the part also increase from 17.3–28.34 MPa when increasing the mould temperature. Wang andYoung [8] also reported that a higher mould temperature tendsto increase the tensile stress in the surface.

The width of the core regions in the parabolic profile in theresidual stresses on thin-walled parts for lignocellulosic polymercomposite increase or become wider with the mould tempera-ture. This increase is caused by the slower rate of propagationof the solid–liquid boundary at higher mould temperatures.The distribution pattern of the residual stresses is the same asthat found in previous research [8,14]. Thus, it is clear that ingeneral, the mould temperature most significantly affects thedeformation residual stresses, especially in thin-walled parts.Altan and Yurci [14] also indicate that the most importantparameter for surface regions of the HDPE parts was found tobe the mould temperature. Mould temperature was determinedto be the most effective parameter for the reduction of residualstresses because when the mould is hotter, the cooling rate isslower [23]. As a consequence, the higher mould temperature

is able to increase the crystallisation time, which consequentlyinduces slower crystallisation with lower stresses that becomefrozen with enough relaxation. In contrast to findings byKwiatkowski et al. [24], the higher mould temperature causeseasier cavity filling, and a decrease in the temperature gradientbetween the polymer and mould wall causes a decrease in theresidual stresses in the part. For lower mould temperaturevalues, an increase in the residual stresses was observed.

However, the mould temperature for moulded thin parts usinglignocellulosic polymer composite material should not be very lowand too high. The mould temperature should be maintained in theoptimum range. This is because if the temperature is too low, thiswill make it difficult for the polymer to flow in the cavity and willcause crystallisation problems, as described by previous research-ers [14]. When the mould temperature is too high, this will causedegradation of the material properties, especially for the lignocel-lulosic polymer composite. Thus, the optimal range of the mould-ing temperature required for moulding thin parts usinglignocellulosic polymer composite must be maintained. Fig. 3reveals that the optimum range of the mould temperature is40–45 �C and of the minimum residual stress distribution is21.65–23.27 MPa at part surfaces near the wall and 17.3–18.2 MPa at core regions of the thin-walled parts. These optimumranges yielded lower residual stress distributions than otherranges.

The effect of different cooling times from 10 to 50 s on theresidual stresses is shown in Fig. 4. No obvious change was foundin this case. Altan and Yurci [14] reported that the cooling timehad very little effect on residual stresses. The cooling time becomesan important parameter if correlated with the mould temperaturein terms of the residual stress distribution due to the effect of thecooling rate during the crystallisation process. Increasing the cool-ing time causes decreases in the tensile residual stresses near thewall from 23.8 to 22.73 MPa, whereas in the core region of the part,a decrease from 18.47 to 17.98 MPa was observed.

Azdast et al. [25] indicate that increasing the cooling time willchange the profile of the residual stresses from compressive to ten-sile residual stresses. The results obtained in this study are differ-ent. Increasing the cooling times will yield an improveddistribution of residual stresses from a high tensile profile to a low-er tensile profile. No compressive residual stresses exist on themoulded thin-walled part composed of lignocellulosic polymercomposite. However, both results agreed that increasing the cool-ing time will reduce the gap in the residual stress distribution inthe thickness direction and replace it with a consistent distribu-tion. This means the distribution of residual stresses existing inmoulded parts should not be too high or low between tensileand compressive residual stresses.

The gap width of the parabolic portion of the core region in theresidual stress profile decreases with cooling time. This decreasesis caused by the slower rate of propagation of the liquid–solidboundary at longer cooling times. The longer cooling times weresupposed to increase the crystallisation time due to induced slowercrystallisation with lower stresses that become frozen with enoughrelaxation [25]. From this study, the results reveal that the opti-mum cooling time is in the range of 20–30 s. Other values thatare less or more than this range are likely to affect the inconsistentdistribution of residual stresses and increase the manufacturingcost for thin-walled parts.

There are three things that should be discussed regarding theeffect of packing pressure on the residual stresses on thin walledparts: (1) A low packing pressure caused cases of polymer meltearly solidification in the packing stage. (2) The optimum packingpressure must be sufficient to reduce the residual stress distribu-tion. (3) When the packing pressure is higher than the optimumpacking pressure, higher inner stresses on the parts will occur [13].

21.65

17.3

9.94

23.27

18.20

11.32

26.86

21.10

13.42

29.25

14.64

24.49

30.79

14.61

28.34

Remarks: Normalized thickness

1 = 0.35 mm (upper),

0 = 0 mm (centre) and

-1 = 0.35 mm (lower).

Fig. 3. The tensile residual stress distribution occurs at the centre of the surface of the thin-walled parts at mould temperatures of 40 �C, 45 �C, 50 �C, 55 �C, and 60 �C.

Remarks: Normalized thickness

1 = 0.35 mm (upper),

0 = 0 mm (centre) and

-1 = 0.35 mm (lower).

22.92 22.73

18.01 17.98

11.08 11.00

23.80

11.70

18.47

23.27

11.31

18.20

23.01

18.15

11.12

Fig. 4. The tensile residual stress distribution occurs at the centre of the surface of the thin-walled parts at cooling times of 10, 20, 30, 40, and 50 s.

384 M.D. Azaman et al. / Materials and Design 55 (2014) 381–386

Fig. 5 shows the effects of different packing pressures from0.45Pinject to 0.85Pinject through the thickness on the cavity residualstresses at the middle of the part surface. An increase in the pack-ing pressure will cause decreases in the tensile residual stress dis-tribution from 46.03 to 23.25 MPa and from 46.32 to 18.15 MPa forthe part surface (both sides) and the core region, respectively. Theoptimum value of the packing pressure is 0.85Pinject for mouldedthin-walled parts using lignocellulosic polymer composite. Thepacking pressure of 0.85Pinject is the best level to ensure replenish-ment of polymer melt into the moulded part to account for the lossdue to volume shrinkage of the cooled part during the solidificationphase. Therefore, at this level, the distribution of residual stressesis lower than the results from other packing pressure levels. Thisstatement also agrees with Postawa and Kwiatwoski [13], who

note that the hold pressure should be maintained at a certain levelto prevent problems with volume shrinkage of moulded parts.Holding pressures that are too high are inadvisable due to theappearance of high internal stresses in the moulded parts [13,18].

A packing pressure less than 0.85Pinject will cause early solidifi-cation phenomena at the packing stage. At this point, the polymerhad completed the glass transition. As a result, these packing pres-sures will only increase the inner stress on the part. The resultsshown in Fig. 5 reveal that a packing pressure less than 0.85Pinject

causes the distribution of residual stresses in the core region to al-most be higher than at the part surface (both sides). During earlysolidification, the packing pressure will be applied in the directionof the flow path, resulting in a higher or inconsistent residual stressdistribution on the part. Nevertheless, the packing pressure

Remarks: Normalized thickness

1 = 0.35 mm (upper),

0 = 0 mm (centre) and

-1 = 0.35 mm (lower).

25.26

23.89

15.29

46.03

32.53

46.32 45.5640.80

45.27 40.54

32.00

23.25

11.30

26.58

18.18

Fig. 5. The tensile residual stresses distribution occurs at the centre of the surface of the thin-walled parts at packing pressures of 0.45Pinject, 0.55Pinject, 0.65Pinject, 0.75Pinject,and 0.85Pinject MPa.

Remarks: Normalized thickness

1 = 0.35 mm (upper),

0 = 0 mm (centre) and

-1 = 0.35 mm (lower).

18.03 18.00

18.17 17.94 18.15

23.0123.0223.25

22.96

11.0811.12

11.30 11.14 11.16

22.92

Fig. 6. The tensile residual stress distribution occurs at the centre of the surface of the thin-walled parts at packing times of 10, 15, 20, 25, and 30 s.

M.D. Azaman et al. / Materials and Design 55 (2014) 381–386 385

increases until 0.75Pinject, and then at 0.85Pinject, the distribution ofthe residual stresses decreased.

Wong and Young [8] reported that the effect of packing pres-sure on residual stresses is usually found at low packing pressures.In this range, the packing pressure drops to zero prior to thecomplete glass transition. At high packing pressures, the pressuredistributions for different packing pressures prior to completeglass transition only differ by a constant, resulting in almost thesame distribution of residual stresses. This means that the residualstresses are affected by packing pressure changes before completeglass transition [26].

Fig. 6 shows the residual stress distribution obtained for differ-ent packing times from 10 to 30 s. The data from the simulationresults found almost no change in the distribution of residual stres-ses in part due to an increase in the packing times. Nevertheless,the packing times had very little effect on the tensile residualstresses at the part surface regions (both sides) compared with

the core regions of the thin parts. At the core regions, almost nochange in the residual stresses was observed; the value remainedat approximately 18 MPa.

Although the packing time did not affect the distribution ofresidual stresses, it is closely related to the packing pressureduring the packing stage. Adjusting the packing time to the opti-mum level will ensure the packing pressure can spread consis-tently to the whole part. This will make it possible to overcomethe problem of volume shrinkage of the polymer melt duringthe filling phase and also contribute to the formation of innerstresses on thin-walled parts. The optimum level for the packingtime is not too short or too long. From the results in Fig. 6, theoptimum packing times for moulded thin-walled parts usinglignocellulosic polymer composite are in the range of 15–20 s.The value of the residual stress distribution is 23.01–23.02 MPaat the nearest part surfaces (both sides) and 17.94–18.1 MPa inthe core regions.

386 M.D. Azaman et al. / Materials and Design 55 (2014) 381–386

4. Conclusions

In conclusion, this study shows that lignocellulosic polymercomposites are suitable for moulding a shallow, thin-walled partto obtain a minimum residual stress distribution. Therefore, theoptimum parameter range for the mould temperature is40–45 �C, for the cooling time is 20–30 s, for the packing pressureis 0.85Pinject, and for the packing time is 15–20 s. These valuesshould be used when processing thin-walled parts using lignocel-lulosic polymer composite for the injection moulding process.

Finally, the numerical simulation analyses have been done onprocessing thin-walled parts using lignocellulosic polymercomposite but further research is required to make experimentalverification. The incremental hole-drilling methods should beemployed to evaluate the residual stresses in moulded parts. Theexperimental results will be compared with numerical results toindentify the similarity on trend of residual stresses distributionwith respect to the thickness. For more accurate measurementand prediction of residual stresses distribution, the comparing re-sults between experimental and numerical simulation will ableto minimize such errors.

Acknowledgement

The authors would like to thank Universiti Putra Malaysia forthe financial support through the Research University GrantScheme (Project No: RUGS/05-02-12-1917RU).

References

[1] Nourbakhsh A, Hosseinzadeh A, Basiji F. Effects of filler content andcompatibilizing agents on mechanical behavior of the particle-reinforcedcomposites. J Polym Environ 2011;19:908–11.

[2] Kuo PY, Wang SY, Chen JH, Hsueh HC, Tsai MJ. Effects of material compositionson the mechanical properties of wood–plastic composites manufactured byinjection moulding. Mater Des 2009;30:3489–96.

[3] Rahman WAWA, Sin LT, Rahmat AR. Injection molding simulation analysis ofnatural fiber composite window frame. J Mater Process Technol2008;197:22–30.

[4] Rahman WAWA, Sin LT, Rahmat AR, Isa NM, Salleh MSN, Mokhtar M.Comparison of rice husk-filled polyethylene composite and natural woodunder weathering effects. J Compos Mater 2011;45(13):1403–10.

[5] Azaman MD, Sapuan SM, Sulaiman S, Zainudin ES, Abdan K. An investigation ofthe processability of natural fibre reinforced polymer composites on shallow

and flat thin-walled parts by injection moulding process. Mater Des2013;50:451–6.

[6] Rosa SML, Santos EF, Ferreiraa CA, Nachtigall SMB. Studies on the properties ofrice-husk-filled-PP composites – effect of maleated PP. Mater Res2009;12(3):333–8.

[7] Sanadi A, Caulfield DF, Rowell RM. Lignocellulosic/plastic composites, Spring,Newsletter; 1998. p. 8–12.

[8] Wang TH, Young WB. Study on residual stresses of thin-walled injectionmolding. Eur Polym J 2005;41:2511–7.

[9] Kim SK, Lee SW, Youn JR. Measurement of residual stresses in injection moldedshort fiber composites considering anisotropy and modulus variation. Korea–Aust Rheol J 2002;14(3):107–14.

[10] Zhou H, Xi G, Liu F. Residual stress simulation of injection molding. J Mater EngPerform 2008;17(3):422–7.

[11] Choi DS, Im YT. Prediction of shrinkage and warpage in consideration ofresidual stress in integrated simulation of injection molding. Compos Struct1999;47:655–65.

[12] Kim CH, Youn JR. Determination of residual stresses in injection molded flatplate: simulation and experiments. Polym Test 2007;26:862–8.

[13] Postawa P, Kwiatkowski D. Residual stress distribution in injection moldedparts. J Achievements Mater Manuf Eng 2006;18(1-2):171–4.

[14] Altan M, Yurci ME. Optimization of residual stresses in the surface regions ofinjection moldings. Polym-Plast Technol Eng 2010;49:32–7.

[15] Shen CY, Li HM. Numerical simulation for effects of injection mold cooling onwarpage and residual stresses. Polym-Plast Technol Eng 2003;42(5):971–82.

[16] Zoetelief WF, Douven LFA, Ingen Housz A. Residual thermal stresses ininjection molded products. J Polym Eng Sci 1996;36:19.

[17] Zhou H, Li D. Residual stress analysis of the post-filling stage in injectionmoulding. Int J Adv Manuf Technol 2005;25:700–4.

[18] Zhou H, Li D. Numerical simulation and experimental study of warpage ofinjection-molded parts. Polym-Plast Technol Eng 2005;44(4):603–17.

[19] Young WB. Residual stress induced by solidification of thermoviscoelasticmelts in the postfilling stage. J Mater Process Technol 2004;145:317–24.

[20] Sen A, Bhattacharya M. Residual stresses and density gradient in injectionmolded starch/synthetic polymer blends. Polymer 2000;41:9177–90.

[21] In-cavity residual stresses in first principal direction, Autodesk MoldflowInsight help, Autodesk Moldflow, Insight, 2011.

[22] Nita A, Oanta E. Improving the quality of the molded polymeric parts byreducing the residual stress. In: Proceedings of the 2nd internationalconference on manufacturing engineering, quality and production systems,p. 77–82 (ISBN: 978-960-474-220-2).

[23] Michaeli W, Po’tsh G. Injection molding: an introduction. New York: HanserPublishers; 1995.

[24] Kwiatkowski D, Gnatowski A, Nabiałek J. Numerical analysis of residualstresses and deformation of injection moulded parts manufactured frompolymeric composite with different processing conditions. Kompozyty2011;11(4):294–8.

[25] Azdast T, Behravesh AH, Mazaheri K, Darvishi MM. Numerical simulation andexperimental validation of residual stress induced constrained shrinkage ofinjection molded parts. Polimery 2008;53:304–10.

[26] Bushko WC, Stokes VK. Solidification of thermoviscoelastic melts. Part 3:Effects of mold surface temperature differences on warpage and residualstresses. Polym Eng Sci 1996;36(3).