CAEDSMouldandDieDesign

CoolingSystemsinInjectionMoulds
SchoolofTechnologyandManagement,PolytechnicInstituteofLeiria

The main phases in an injection moulding process involve filling, cooling and
ejection.Thecoolingphaseisthemostsignificantstepamongstthethree.Itdeter
mines the rate at which the parts are produced. In the moment of the melted
polymer injection, ideally, the moulds temperature should be like of the melted
polymerstemperatureandinthemomentofthepartsremovalthemouldmustto
be to the temperature of the environment. Of this way, the polymer would be
injected with the minimum of pressure and the difference between the surface
temperature and the nucleus temperature of the injected parts would be a mini
mum leading a slow cooling and minimising the mouldings stresses. Notice that
these technical advantages are not compatible with economical needs and the
generalizedruleistoproducepartswiththebiggestpossiblespeed.Accordingto
this rule, the most important factor is the capacity of the cooling system removes
heatofthecavitiesofthemould.Usuallythetimeofcoolingisaround50%ofthe
total cycle. The injected material loses temperature in the contact with the mould
surfaces,transferringitselfheatthroughthemould.Forspeedingtheheattransfer
process, the mould designer design specific holes in the adjacent surfaces of the
mouldedpartinthemould.Theseholes,knownbylinesofwater(bythewateris
themorefrequentfluidofcooling),constitutethecoolingsystemofamould.

Introduction

Thefundamentalrulesthatshouldbehadincountinthecoolingsystemdesignare:
i) The circuits of the water should be symmetrical and independent rela
tivelytothefillingzonesandimpression(s)ofthemould;
ii) Thermal variations in the walls of the impressions shouldnt be pro
nounced,sothelinesofwatershouldbedesignedinfunctionofitsdistance
totheimpressionwalls;
iii) The cooling fluid input and output should be placed for the mould
backwards (opposite side to the operator), or alternative for the breaks
lower;
iv)Itsimportanttoguaranteethatthecoolingflowinthechannelsbetur
bulent.TheindexofturbulenceisgivenbyReynoldsnumber:

Re =

vd

Where,
vFlowsspeed
dChanneldiameter
Fluiddensity
mDynamicviscosityofthefluid

CoolingSystemsinInjectionMoulds1

CAEDSMouldandDieDesign
Whenitproceedstothepolymerinjectionforinsidetheimpressionofamouldthe
removal energy of the polymer in the melted state is transmitted by conduction
through the mould material up to the channels of the cooling system and to the
mouldexternalsurface.Theheatexchangemechanisms(fig.1)includetheconduc
tionforthestructureoftheinjectionmouldingmachine,theforcedconvectionfor
the fluid that circulates into the cooling channels and the thermal radiation and
naturalconvectionfortheairthatsurroundthewallsofthemould[1,2].

HeatTransfer

Figure1Heatexchangeinamouldofinjection

Intheinjectionmouldingcycle,theheatcorrespondingtotheenthalpyvariationof
themouldingmaterialduringthecycle,isexchangedforthemouldingzonesurface
(orimpressionsurfaceofthemould)andofthisforhisoutside.Todefinetheen
ergy swing, is established an equilibrium between the heat powers that are
introduced in the mould, the heat power accumulated in every single moment in
their interior and the heat powers removed from the mould, being positive or
negativethosethatrespectivelyincreaseordiminishtheirinternalenergy[1,3].Ina
process analysis with accumulation of internal energy, the heat flow that is sup
pliedtothemouldandtheheatflowthatisremovedfromthemouldshouldbein
thermal equilibrium, in every single moment, with the heat accumulated in the
structureofthemould:

EnergyBalance

Q PL + Q AMB + QTM = Q ACCUM

Q PL Heatflowsuppliedbythepolymer

Q AMB Heatflowtransferredfortheenvironment

QTM Heatflowtransferredforthecoolingfluid

Q ACCUM Accumulatedenergyinthemouldmaterialpertimeunit

CoolingSystemsinInjectionMoulds2

CAEDSMouldandDieDesign
Simplifiedhypothesestoobtainresults
i)Quasistaticprocess
ii) During the cycles the temperatures and thermal flows fluctuations are
despised
iii)Duringthedifferentperiodsmediumvaluesareconsidered

Q PL + Q AMB + QTM =0
Where,

Q PL =

h m PL
or,
t arref

Q PL =

h PL V

t arref

Where,

h = hi- he;hi Polymerenthalpyattheinjectiontemperature;he Polymeren

thalpyattheejectiontemperature;mPLPolymermassinjectedinthemould;PL
Polymermediumdensitybetweentheinjectiontemperatureandtheejectiontem
perature;tarrefCoolingtimeoftheplasticpart;VVolumeoftheplasticpart

Q AMB = Q CONV + Q COND + Q RAD

Where,

Q CONV Heatflowbyconvectiononthemouldlateralwalls

Q COND Heatflowbyconductionontheinjectionmouldingwalls

Q RAD Heatflowbyconductiononthemouldlateralwalls

Q CONV =ALxhx(TambTmould)
Where,
ALMouldexposedarea;hHeattransfercoefficient,naturalconvection;Tamb
EnvironmentTemperature;TmouldMouldtemperature.

Q COND =Afixxx(TambTmould)
Where,
AfixContactareaMould/Fixingsystem;Proportionalityfactor

Tmolde

100

Q RAD =ALxxradx Tamb

100

Where,
radStefanBoltzmanconstant;Materialemissivity

When the material is inside the mould cools supplying him heat, by that QPL is
alwayspositive.Theheatchangedwiththeenvironment,canbepositiveornega
tivedependingonthetemperatureofthemould.
CoolingSystemsinInjectionMoulds3

CAEDSMouldandDieDesign
An efficient system of cooling, with optimal cooling conditions, leads to a part
uniform distribution of temperatures, minimizing the undesired effects appeared
duringdecoolingprocess,thecycletimeandtherateofrejections.Theconception
ofanefficientcoolingsystemisnotasimpletrial,becausetherearedifferentfactors
thatcancontributeforthefinalintendedresults.Someofthefactorsthatinfluence
thecoolingprocessare:thegeometryofthepart,thetemperatureofthemould,the
architectureofthecoolingchannels,thecoolingfluidtemperatureandthespeedof
theflow.
Itcanbeidentifiedtworeferencetermsforaniterativeprocessofcharacterization
ofthemouldcoolingsystem[3]:
i)Theincreaseoftheheattransferrate
ii)Uniformtemperaturedistributioninthemouldingsurface

Whereas the increase of the heat removal rate between the plastic part and the
mould is important in the economical point of view, the uniformization of the
temperaturesdistributiononthepartssurfaceswillprovidetheobtainingofparts
withestatesandqualityimproved.

CoolingTime

TheWubkenequationallowustoestimatethecoolingtime[3]

tK =

s2

Cooling
Conditions

8 T Ta

ln 2 W
TW Tb

Whereisthematerialthermaldiffusivity;sisthepartthickness;Taistheinjec
tion temperature; Tb is the ejection temperature and Tw is the medium mould
temperature.
Themediummouldtemperatureisconsideredoneofthemostsignificantvariables
inthecoolingtimedetermination[4,5].Somedeterminationsusethetemperature
ofthecoolingfluidforcalculatingthemediummouldtemperaturevariable.How
ever, such utilization ignores the temperature increases of the melted plastic
materialinthemoldingzones,duringtheinjectionphase.Duringthemoldingcycle
the mould temperature increase whilethe plastic materialis injected, diminishing
progressively up to the following injection. Also the flow regime of the cooling
fluid,thetemperatureofthecoolingfluid,thearchitectureofthechannels,thekind
ofthecoolingfluid,andthemouldmaterialproperties(namelythemouldmaterial
thermalconductivity),influencethemouldtemperature.

Table1Propertiesofatypicalresin,Aluminiumandsteel,usedinthemanufactureof
injectionmoulds.
SL Vantico
5260

Aluminium
AlZn5Mg3Cu

Steel P20

Young modulus

600 - 800 MPa

72 MPa

2500 GPa

Tensile strength

40 - 65 MPa

540 MPa

-1

-1

Thermal conductivity

0.2 W.m K

Coefficient of thermal
expansion (at 20C)

10510-6 K-1

965-300 MPa
-1

-1

120-150 W.m K

29-34 W.m-1K-1

23,610-6 K-1

12,810-6 K-1

CoolingSystemsinInjectionMoulds4

CAEDSMouldandDieDesign
Ifthecoolingchannelsarentcorrectlydesigned(fig.2),thecoreandcavitymould
walltemperaturecanbedifferent.Ifthereisastronggradientinthecavitybetween
thetwohalvesthepartmaywarpanddistortitsshape[68].

Sothetargetsthatacorrectcoolingsystemhastofollowaretheuniformityofthe
walltemperatureandagradualreductionofthepolymertemperature,inorderto
find a compromise between the necessity of reducing cycle time and allowing for
thecrystallization.

warpage

or
internal stresses

Qcore<Qcavity

Last
cool

layer

to

Ejected part

Figure2Coolingthroughthepartthickness.

Thedifferencesshowedinfigure2thatareabletomustitself,byex.,thedifferences
in the cavity and core geometry. After parts ejection, the more shrinkage on the
hottersideofthemolding(coreside)promotethewarpageorinternalstresses[6,
9].
Generally,convexareasneedhighcoolingbecauseinthesepartsthereisaconcen
tration of heat. On the contrary concave areas need less cooling because the
presenceofmorematerialhelpsthediffusionofheatinthemould.Thus,attention
mustbepaidtodesigningcornersandthecoolingsystemintheseareas[10].
The diameter and the arrangement of the cooling channels is limited by construc
tive aspects of the mould. However, a cooling system balanced and uniform
providesimprovementinthequalityofthemouldings.Alsoitshouldnotbeforget
fulthattobiggerefficiencyofthecoolingfluidisobtainedwhenhisflowregimeis
turbulent.
Accordingtotheliterature[eg.,11,12],thelocalizationofthecoolingchannelshasa
great importance for having an uniform cooling, because it decides the moulding
surfacestemperaturedistributionandevolutionduringtheperiodofcooling.

CoolingSystemsinInjectionMoulds5

CAEDSMouldandDieDesign
Inthiscontext,thedistancebetweenthecoolingchannelsandthemouldingsurface
(h) and the distance between cooling channels (e) are the main parameters to be

considered,asshownintheschemeofthefigure3.

Coolingchannels

mould

e
d

qmx

qmin
molding

s/2

Figure3Heatflowprofile[13].

Inthepracticalone,iscommontoconsider:e=2,5a3,5deh=0,8a1,5e

On the issue of dimensional criteria in designing cooling channels, three dimen

sionshavetobeconsidered:thediameterofthecrosssection(orthecrosssection
areaifnotcircular),thedistancebetweenchannelsandthedistancebetweenchan
nel and wall of the mould. The main problems that arise when choosing these
dimensions concerns the pressure losses derived from the choice of the diameter
and the design of the channel. A heating/cooling relationship reported in Zollner
[14]givesaguidelineonthechannelspositioning.Thisstatesthatthevalueresult
ingfromthesolutionoftherelationshipshouldstaybetween2.5and5%forsemi
crystallinethermoplasticsandbetween5and10%foramorphousthermoplastics.

In the injection molding process the main part of the cycle time is determined by Conformal
thecoolingprocess.Therefore,itisimportanttooptimizethecoolingcycleinorder Cooling
toreducethecoolingtime.Conformalcoolingchannels(i.e.channelsthatfollowthe
geometricshapeofthepart)havebeenusedforthispurposeallowingasignificant
coolingtimereduction.AccordingtoWohlers[15]itispossibletoreducethecool
ing cycle by 20% using conformal cooling channel. Similarly, Dimla et al. [10]
considers that cycle time can be significantly reduced with cooling taking place
uniformly in all zones if the cooling channels are made to conform to the parts
shapeasmuchaspossible.Someinvestigationshaverelatedthemouldscycletime
reduction with conformal cooling; the most relevant result associated to its use is
themouldsurfacetemperatureuniformity.Furthermore,ifthepartisejectedwith
thesametemperatureineverypointthesubsequentshrinkageoutsidethemouldis
alsouniform,whichavoidspostinjectionwarpageofparts.Thiswasalsopointed
out by Voet et al. [16], which mentioned that the goal of cooling a mould is to
obtain a uniform temperature at the mould surface and within the final injected
producttoavoidinternalstresses.

CoolingSystemsinInjectionMoulds6

CAEDSMouldandDieDesign
Amethodthatutilisesacontourlikechannel(fig.4),constructedascloseaspossi
ble to the surface of the mould to increase the heat absorption away from the
moltenplastic,ensuresthatthepartiscooleduniformlyaswellasmoreefficiently.

Figure4Conformalcoolingchannels

Whenmoltenplasticisinjectedinthemoulditmustbesolidifiedtoformtheobject.
Themouldtemperatureisregulatedbycirculationofaliquidcooler,usuallywater
oroilthatflowsinsidechannelsinsidethemouldparts.

Table2Heatconvectioncoefficientoftheair,waterandoil.

Air
Water
Oil

Heatconvection
50
900
400

coefficientWm2k1

Whenthepartissufficientlycooleditcanbeejected.Most(95%)oftheshrinkage
happensinthemouldanditiscompensatedbytheincomingmaterial;theremain
deroftheshrinkagetakesplacesometimefollowingtheproductionofthepart[17].
Ifthechannelscarryingthewatercouldbeconformedtotheshapeofthepartand
their cross section changed to increase the heat conducting area then a more
efficient means of heat removal could be realised. This may also help to reduce

warpagewhenthepartisejected,astheplasticwouldbecooledmoreuniformly.

Anotheradvantageisthatamouldequippedwithconformalchannelsreachesthe
Modelling
operation temperature quicker than a normal one equipped with standard (or

drilled)coolingchannels[18,19].

The analysis tools utilization for the cooling systems conception that assures the
uniformityofthecoolingalongthepart,drivethesignificantimprovementsinthe
mould production and definition of the process conditions to the specifications of
theproductdemanded.

The main resistance to the transference of heat in the cooling happen of the own
materialduetothelowthermaldiffusivityoftheplasticmaterial.So,itsessential
toconsiderthedependenceofthematerialwiththetemperatureinthemodulation
oftheheatconduction.

CoolingSystemsinInjectionMoulds7

CAEDSMouldandDieDesign
Inthecoolingprocessitsessentialtoconsiderthethermalpropertiesofthemould
materialandappropriateborderconditions(e.g.theheattransferbyforcedconvec

tioninthecoolingchannels).

Inisotropicdomaintheheattransferisdescribedbytheenergyconservationequa

tion[20]:

C P
= (KT ) + Q
t

Where,CPandkrepresentthedensity,thespecificheatandthethermalconduc
tivity of the material, respectively. T represents the local temperature in each

instantmomenttandineachspatialcoordinate,whereas Q representstheenergy
generated/dissipated by unit of time and by unit of volume in the material. This
differential equation with derived partial for bidimensional heat conduction, not

stationary,inCartesianscoordinatesandinasimplifiedform,takestheform:

T
T T
+Q
= K
C P
+ K

t x x y y

Thetemperatureprofileinagivenzoneofthematerialandhisvariationwiththe
timeareabletobeobtainedresolvingthisequation.However,itisnecessaryspec

ifythetemperatureprofileintheinitialinstantandtheborderconditions.

To optimise the design and construction of the mould, with attention on refining

the tool design through application of finite element and thermal flow analyses,

specific commercial software for injection moulding have been used. In the next

sectionitwillbemadeabriefdescriptionaboutthe heattransferprocessanalysis

usingsomecommercialsoftware.

ThelatestcommercialsoftwareofCAEallowsthreedimensionalsimulationofthe Coolingsystem
injection molding process. This software has modules for conception efficient simulation
coolingsystems.Thecoolinganalysisisbasedinthemethodoftheborderelements
approach.
InthecoolingmoduleofthecommercialCAEsoftware,thetransferenceofheatin
the polymer is treatedas onedimensional conduction locatedin transient regime.
Theheatexchangebetweenthesurfaceofthecoolingchannelsandthecoolingfluid
areconsideredinstationaryregime,consideringthecorrelationfortheheattrans
ferencecoefficientbyconvection.Tosolvesimultaneouslytheprominentequations
oftransferenceofheatinthisprocess,theprogramutilizesahybridschemewhere
the transference of heat is calculated by the approach modified analyzes of the
element of three dimensional border for the region of the mould, and one
dimensional heat transference analysis, along the part thickness for the region of
meltedplastic.Thesetwoanalysesareconjugatedofformitequalthetemperature
andtheheatflowintheinterfacepolymer/mould.

CoolingSystemsinInjectionMoulds8

CAEDSMouldandDieDesign
Theequationsfortheflowofthefluidinacircuitofcoolingareresolvedthrough CAESoftware
the iterative approach of NewtonRaphson, to obtain the torrent and the fall of Simplifications
pressureineachchannelofthecoolingsystem.Then,theheattransferencecoeffi
cientsbetweenthesurfacesofthechannelsandthecoolingfluidarecalculated.

Thechangeofheatbynaturalconvectionbetweentheenvironmentandthewallsof

themouldarealsocalculated.Forthiscalculation,commercialsoftwareconsiders
theexteriorsurfaceofthemouldasaspherewithanareaequivalenttothesurface
ofabox,inthatthechannelsofcoolingwillbeincluded,thefeedingsystemandthe
moldingzones.

Theprocesssimulationstartsinthephaseofthemouldfilling.Whenthecooling
module of cooling is used, the polymer injection temperature is assumed as con
stant. This assumption has some associated errors; therefore the injection
temperature can be a superior due to the heating by viscous dissipation of the
materialinthesprue.Thattemperaturewouldbeabletogoupuntil30Cdepend
ingonthespeedofinjectionandofthematerialproperties[21].

The thermal resistance in the interface polymer/mould defines the heat transmis
sioncoefficient(hint)intheinterfacebetweenthepolymerandthemoldingsurfaces.

Thiscoefficientisusedforsimulatetheresistancetotheexistingheatinthecontact
betweenthetwomaterialsbythefollowingequations:

hint Tint TM

x=b

T
= k

n x =b

hint Tint TM+

x=+b

T
= k

n x =+ b

where,Tintisthemelttemperatureintheinterfaceofthetwomaterials; TM and TM+

arethemoldingzonestemperatures,onthecavityside(negativeside)andonthe
coreside(positiveside),respectively.Theindicesband+bindicatethepositive
andnegativesideofthedistancerelativelytothecenterofthepart(equivalentthe
halfofitsthickness).
If the thermal conductivity assumes the zero value, (thermal isolated border), the
changes between the two materials do not exist. If it assumes an elevated value,
existaperfectthermalcontactbetweenthematerialsandtheinterfacetemperature
is considered equivalent at the mould wall temperature. Many times, and by
defect,thisvalueisof25000w/m2C,incommercialsoftware.
The case study presented shows some important aspects when different cooling
systemsareconsidered.

CoolingSystemsinInjectionMoulds9

CAEDSMouldandDieDesign

CaseStudy

Figure5Coolingsystemcasestudy.

Coolingsysteminthecavityside

a)Conventionalcoolingsystem

Figure6Temperaturedistributiononthe
partssurfaces

Figure7Partsdeflection

Figure8Partscoolingtime

Figure9Percentagefrozenlayer

CoolingSystemsinInjectionMoulds10

CAEDSMouldandDieDesign
b)Bafflecoolingsystem

Figure10Temperaturedistributiononthe
partssurfaces

Figure11Partsdeflection

Figure12Partscoolingtime

Figure13Percentagefrozenlayer

c)Conformalcoolingsystem

Figure14Temperaturedistributiononthe
partssurfaces

Figure15Partsdeflection

Figure16Partscoolingtime

Figure17Percentagefrozenlayer

CoolingSystemsinInjectionMoulds11

CAEDSMouldandDieDesign

Coolingsysteminthecavityandcoresides

a)Conventionalcoolingsystemsinthecavityandcoresides

Figure18Temperaturedistributiononthe
partssurfaces

Figure19Partsdeflection

Figure20Partscoolingtime

Figure21Percentagefrozenlayer

b)Bafflecoolingsystemsinthecavityandcoresides

Figure22Temperaturedistributiononthe
partssurfaces

Figure23Partsdeflection

Figure24Partscoolingtime

Figure25Percentagefrozenlayer
CoolingSystemsinInjectionMoulds12

CAEDSMouldandDieDesign
c)Conformalandbafflecoolingsystemsinthecavityandcoresides,respectively

Figure26Temperaturedistributiononthe
partssurfaces

Figure27Partsdeflection

Figure28Partscoolingtime

Figure29Percentagefrozenlayer

d)Conformalcoolingsysteminthecavityandcoresides

Figure30Temperaturedistributiononthe
partssurfaces

Figure31Partsdeflection

Figure32Partscoolingtime

Figure33Percentagefrozenlayer

CoolingSystemsinInjectionMoulds13

CAEDSMouldandDieDesign

References
[1] MENGES, G.; MOHREN, P. How to Make Injection Moulds. 2nd ed, Hanser
Pulishers,1993.ISBN3446163050
[2] LIMA, S. P. Evaluation of the rapid prototyping incorporation in injection
moulds,MasterThesis,October2002.
[3] POUZADA, A.S. Heat transfer in injection moulds Support texts to the
MouldDesignandManufacturingMasterDegree
[4]BARROS,I.;TEIXEIRA,S.F.C.;TEIXEIRA,J.C.;CUNHA,A.M.Evaluationof
thethermalBehaviourofInjectionMoulds.Intern.PolymerProcessing,Vol.15,No.
1(2000),pp.95102.
[5] BOELL, K.M. Predicting the cooling time of na injection moulded part, Pro
ceedings of the 53th Annual Technical Conference & Exhibition, ANTEC 1995,
Boston,711May1995,pp.42424246.
[6] MALLOY, ROBERT A. Plastic part design for injection molding. New York:
HanserPublishers,1994.460p.ISBN1569901295.
[7]WANG,T.J.;YOON,C.K.Shrinkageandwarpageanalysisofinjectionmolded
parts.Orlando:SPEANTEC2000,p.687692.
[8] JOHANNABER, F. Injection molding machines. Third Edition. New York:
HanserPublishers,1994.315p.ISBN1569901694.
[9] MARTINHO, P.G. Warpage study in injection moulding parts. University of
Minho,Guimares,2002.98p.MasterThesis
[10]DIMLA,D.E.;CAMILOTTO,M.;MIANI,F.Designandoptimisationofcon
formalcoolingchannelsininjectionmouldingtools.JournalofMaterialsProcessing
Technology,164165,pp.12941300,2005.
[11]SINGH,K.J.MoldCooling.InBERNHARDT,E.C.CAE:ComputerAided
EngineeringforInjectionMolding.Munich:CarlHanserVerlag,1983.ISBN3446
139508.p.326347.
[12]YANG,S.Y.;CHANG,H.C.Studyontheperformanceofcoolingsystemsin
precisioninjectionmolds.Intern.Polym.Proc.Vol.10,n2(1995),p.255261.
[13] POTSCH, G.; MICHAELI, W. Injection molding: an introduction. Munich:
CarlHanserVerlag,1995.195p.ISBN1569901937.
[14]ZOLLNER,O.Optimisedmouldtemperaturecontrol,Appl.Technol.Inform.
(1997)1104.
[15]WOHLERS,T.,WohlersReport2006Rapidprototypingandmanufacturing
StateoftheIndustry.AnnualWorldwideProgressReport,WohlersAssociates,Inc.,
2006.
[16]VOET,A.;PEE,B.V.;MINGNEAU,J.;CARDON,L.;HOUTEKIER,R.Optimi
zationofconformalcoolingbyuseofdesignofexperiments:industrialcasestudy
of an injection molded product, RPD 2006 Building the future by innovation,
MarinhaGrande,1317November2006.

CoolingSystemsinInjectionMoulds14

CAEDSMouldandDieDesign
[17]BRYCE,D.M.PlasticInjectionMoulding,SocietyofManufacturingEngineers,
Dearborn,MI,1996.

[18] SACHS, E.; WYLONIS, E.; ALLEN, S.; CIMA, M.; GUO, H. Production of

injection molding with conformal cooling channels using the three dimensional

printingprocess,Polym.Eng.Sci.,2000,40(5),12321247.
[19] DELGARNO, K., W., Layer manufactured production tooling incorporating
conformalheatingchannelsfortransfermouldingofelastomercompounds,Plastic
RubberCompos.30(8)(2001)384388.

[20]HOLMAN,J.P.HeatTransfer,NewYork:MacGrawHill,Inc,1989.ISBN0

071004874.
[21]CMOLDusersmanual,ACTechnology,IthacaNewYork,1997.

CoolingSystemsinInjectionMoulds15