olaf zöllner -chupagem

The fundamentals of shrinkage of thermoplasticsOlaf Zllner

l

l Shrinkage

Table of contents

1 The concept of shrinkage 3

2 Factors that influenceshrinkage 9

2.1 Material 102.1.1 Amorphous and semi-crystalline

thermoplastics 102.1.2 Crystallization behavior 152.1.3 Fillers and reinforcing materials 15

2.2 Processing 172.2.1 Pressure holding time 192.2.2 Holding pressure level 192.2.3 Cavity wall temperature 202.2.4 Melt temperature 222.2.5 Injection velocity 222.2.6 Ejection temperature 23

2.3 Molded part geometry 232.3.1 Wall thicknesses 232.3.2 Reduced wall thickness at the

end of the flow path 262.3.3 Ribs 27

2.4 Mold 282.4.1 Heating/cooling 282.4.2 Gate type 292.4.3 Gate position 30

3 Phenomena of relevanceto shrinkage 31

3.1 Warpage 323.2 Internal stresses 373.3 Sink marks/voids 383.4 Corner warpage 39

4 References 40

Page 2

1 The concept of shrinkage

When thermoplastics are processed by injectionmolding, the dimensions of the molded part change as the part cools. Often, these changes arereferred to as either shrinkage or warpage.Strictly speaking, shrinkage is due to the compressibility and thermal expansion ofplastics. When thermoplastics shrink, they undergoa volume change. With warpage, by contrast,the shape changes while the overall volumeremains constant.

In order to obtain the desired dimensions in theplastic part, the mold cavity must be enlarged bythe amount of shrinkage that will be experiencedby the plastic.

For this reason, the mold builder has to predict the difference, due to shrinkage, between thedimensions of the mold cavity and those of themolded part. In many cases, this is not an easy task, since shrinkage is influenced by a large number of parameters.

Apart from process control (temperatures,pressures) and the properties of the material (e.g.its pvT behavior, filler content, and its amorphous

or semi-crystalline nature), the stiffness and wallthicknesses of the molded part also have an effecton shrinkage.

Although shrinkage is based on thermal contrac-tion, the effective reduction in dimensions after demolding is less than would be expected for pure thermal contraction.

Additional mechanisms are involved in shrinkagethat lead to a lower level of thermal contractionthan would otherwise result. The chief mechanismsinvolved here are:

inherent stresses, influenced by the temperature profile during cooling the cavity surface temperature

crystallization in the case of semi-crystalline thermoplastics, influenced by cavity surface temperature cooling profile

mechanical obstruction of shrinkage,influenced by mold constraint

Page 3



Apart from this, thermal contraction is influencedby the thermodynamic process sequence in injec-tion molding, and particularly through

the pressure profile during cooling the action of pressure and the stressing of the

melt due to the design and position of the gate the melt temperature and other processing

conditions.

The influence of the thermodynamic process profile on shrinkage behavior is presented in a very clear fashion on a pressure (p), volume (v) and temperature (T) diagram known simply as the pvT diagram which characterizes:

compressibilitychange in volume with a change in pressure

and

thermal behaviorchange in volume with a change in temperature

Page 4

TA

V1bar

Pressure p1 < p2 < p3; p1 = 1bar

Influence oftemperature

Influence ofpressure

6

5

4 3

2

1

0 P1

P2

P3

Temperature T

Sp

eci

fic

volu

me

V

VA

TD T1bar

Start ofshrinkage

V

Figure 1Schematic pressure (p), volume (v) andtemperature (T) diagram

0 1 volumetric filling1 2 compression2 3 action of holding pressure3 4 pressure reduction to ambient pressure4 5 cooling to demolding temperature (T

D)

5 6 cooling to ambient temperature (TA)


The pvT diagrams supply information on the volume shrinkage, SV, from (V = V1bar VA) andshow shrinkage potentials. Volume shrinkage isdefined below:

VC VPSV = VC

VC: volume of cold cavity VP: molded part volume

SV = 1 (1 SL)(1 SW)(1 SS)

SL: longitudinal shrinkageSW: shrinkage of widthSS: shrinkage of thickness

When it comes to the practical layout of injectionmolds, however, it is the linear shrinkage that is more important.

IW IFSI = IW

Sl = linear shrinkagelW= dimension of cold cavitylF = dimension of molded part

Shrinkage is a relative value and is given in theform of a percentage.

Page 5


A real-life part cannot shrink uniformly in all threedirections (over its length, width and thickness).Only over the thickness of the part does virtuallyunimpeded shrinkage take place. Most of the volume shrinkage, therefore, is used up in theshrinkage of the wall thickness of the molded part.Even if the mold does not impede shrinkage in anyway, the fact that the layers of the molding freezefrom the outside towards the inside means thatshrinkage is obstructed over the length and widthof the part.

The following shrinkage situation generally resultsfrom shrinkage that is due to the mold and intern-ally obstructed shrinkage

Shrinkage over thickness SS = 0.9 0.95 * SV

Shrinkage over length or width SL/W = 0.1 0.05 * SV

SV = volume shrinkageSS = shrinkage over thicknessSL/W = shrinkage over length or width

Page 6

constrained length

fluidcenter

solidifiedframe

S

contour after shrinkagemold contour

Figure 2Internal obstruction of shrinkage in a cooling plasticpanel /4/

Shrinkage SL/W does not make allowance for anyanisotropic shrinkage that may occur on account offillers and reinforcing materials.

1 The concept of shrinkage Page 7

Since shrinkage is a time-dependent parameter,the point in time after demolding at which theshrinkage is measured must be specified in orderto achieve a precise definition.

Figure 3Change in a molded part dimension over time through shrinkage

Le

ngt

h

Time

1

2

3 4

5

6

A B C D E

7

F

1 = Mold dimension; 2 = Thermal expansion of mold;3 = Demolding shrinkage SE; 4 = Molding shrinkage SM5 = Post-shrinkage SP; 6 = Overall shrink. S tota l ; 7=Potentiallength increase through conditioning, e.g. for polyamides

A = Cold mold; B = Hot mold; C = Molded part after demolding; D = Molded part after 24h in standard climate;E = Molded part after a prolonged period or after storage in heat; F = Molded part after water absorption, e.g. forpolyamides

1 The concept of shrinkage Page 8

A distinction is drawn between demoldingshrinkage SD, which is measured immediatelyafter the injection molded part has been ejected,molding shrinkage SM and post-shrinkageSP.

According to DIN 16901, the molding shrinkage(SM ) of engineering plastics in injection molding isspecified as the difference between the dimensionsof the cold mold and those of the molded partafter 16 hours storage in a standard climate /3/.With continued storage at high temperatures, afurther change in dimensions can occur, which isthen known as post-shrinkage (SP).

Molding shrinkage SMIW IF1SM = IW

Post shrinkage SNIF1 IF2SP =

IW

lF1 = molded part dimension before SNlF2 = molded part dimension after SNlw = dimension of cold mold

The reasons for post-shrinkage are the relaxationof inherent stresses, together with re-orientationprocesses and, in the case of semi-crystalline materials, potential post-crystallization /1/.

With a number of thermoplastics, conditioning(water absorption) can also take place, in additionto post shrinkage. This effect is seen particularlywith polyamides.

The level of post-shrinkage can be quiteconsiderable if the processing is not correctly tailored to the material.

The term overall shrinkage is defined as follows:

Overall shrinkage = molding shrinkage + post-shrinkage

2 Factors that influence shrinkage Page 9

2 Factors that influence shrinkage

Figure 4Factors influencing shrinkage /2/

Shrinkage

Molded part

Plastic Molding process

Mold


The interplay between the different influencing factors is highly complex. There are factors whichmutually influence each other and which aredependent on each other. Hence, a poorly dimen-sioned gate (molded part) can have a direct negative influence on the effect of the holding pressure (processing).

The interaction between the different influencingfactors is not covered in what follows. Instead,the basic way in which shrinkage is influenced isexplained on the basis of these factors.

The decisive point in this observation is that thecorrelation between the influencing factors and the shrinkage is always viewed in terms of themechanisms listed at the outset:

thermodynamic state profile, inherent stresses and crystallization.

This makes it possible to explain the dependenceof shrinkage on the influencing factors in a clearmanner.

2.1 Material

2.1.1 Amorphous and semi-crystalline thermoplastics

The dissimilar shrinkage behavior of the two material types can be presented particularly clearlyon the pvT diagram.

Page 10

Figure 5Qualitative pressure, specific volume and temperaturediagram (pvT diagram) for an amorphous and a semi-crystalline thermoplastic .

Sp

eci

fic

volu

me

V

Temperature T

TG TC

VC (0)

VA(0)

0

Glass-like state Uncooledliquid

Liquid

amorphous

semi-crystall

inep = constant


Amorphous materials

The green curve in Fig. 5shows the dependence of the specific volume on thetemperature of an amorphousthermoplastic (e.g. ABS, PC).Undercooling of the melt is evident here as the tem-perature falls. Below the glass transition point TG, the volume undergoes a less pronounced reduction(lower gradient), and the material is in a glass-likestate undercooled liquid. The free volumeremains virtually constant during this phase /5/.

Semi-crystalline materials

The red curve in Fig. 5 shows the path of the specific volume as a function of temperature for a semi-crystalline polymer (e.g. PA,PBT).

Page 11

In the case of semi-crystalline thermoplastics,crystallization takes place during cooling. If theplastic falls below the crystallization temperature,TC, then linear, or only slightly branched, chainswith easily alignable repeat units will reconfigurethemselves into an ordered (crystalline) state with a lower free enthalpy.

The polymer chains lie parallel to each other,tightly packed in some cases, forming fine crystal-lites in which almost complete order prevails,with the exception of a few defect points. Contraryto the case with purely crystalline materials, thereis no single point of crystallization here but a transition range instead /5/.

The superposition of advancing crystallization andthermal contraction leads to the parallel-shapedcurve profile in the semi-crystalline range. Highershrinkage values are seen than with purely amorph-ous materials.

The influence of the pressure that prevails duringcooling on the specific volume is shown in Fig. 6.


Figure 6pvT diagram of an amorphous (left) and semi-crystalline (right) plastic during slow cooling under different pressures p0, p1, p2

Page 12

Sp

eci

fic

volu

me

V

Temperature T

TR

Glass

Sp

eci

fic

volu

me

V

Temperature T

Freezing line

Pressure p

G0

G1

G2

S0

S1

S2

2

1

3

Melt

Freevolume

p0

p1

p2

TG0 TG1 TM

(0 > 1 > 2)S,G Pressure p

Solidification line

S0 > S1 > S2

Melt

Semi-crystallinerange

p0

p1

p2

TR TCO TC1 TM

2

1

3

S0

S1

S2

Olaf Zoellner


Figure 6 shows that the higher packing densitythat results with rising pressure causes an upwardshift in the glass transition point TG for amorphousplastics and in the crystallization temperature TCfor semi-crystalline plastics /5/. The specific volumeachieved at room temperature under a high pres-sure is lower than the specific volume at atmo-spheric pressure. When the pressure is removed,the plastic attempts to achieve the specific volumeat a normal pressure of 1 bar.

Figure 7 shows linear shrinkage ranges for a number of amorphous and semi-crystalline thermoplastics. These shrinkage ranges result fromthe influence of processing, the process, the moldand the molded part (e.g. the wall thickness) andthe fillers or reinforcing materials. The shrinkage of the non-reinforced amorphous polymers (Novodur, Bayblend, Makrolon, Apec) is lessthan 1 %, while that of the non-reinforced semi-crystalline polymers (Durethan, Pocan) is morethan 1 %. With the reinforced materials (GF), anincreased anisotropy (directional dependence) isobserved in the shrinkage.

Page 13

Amorphousthermoplastics

S < 1 %

Relatively isotropicshrinkage values,hence only a slighttendency to warp

Semi-crystallineand reinforcedthermoplastics

S > 1 %

Shrinkage as afunction of fiberand molecule orientation, hencea pronounced tendency to warp


Figure 7 Shrinkage ranges for amorphous and semi-crystalline thermoplastics /2/

Page 14

Shrinkage in %

0

longitudinaltransverseisotropic

Novodur/Lustran ABSBayblend (PC+ABS)

(PC+ABS), GFMakrolon PC

PC, GF30Apec HT PC-HT

PC-HT, GFDurethan PA6

PA6, GFPocan PBT

PBT, GF30Triax (ABS+PA)

(ABS+PA), GFDesmopan TPU, hard

TPU, mediumTPU, softTPU, GF

0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Speciality products up to 0.8 %

2.2 2.4

Established on shrinkage measurement plaques (150 x 90 x 3 mm) under the recommended proces-sing conditions. Holding pressure approx. 500 bar.

Shrinkage values relate to the product class and not to individual grades and cannot be used for theprecise layout of a mold.


2.1.2 Crystallization behavior

In the case of semi-crystalline thermoplastics,crystallization occurs during cooling. This crystallization process is time and temperature-dependent. The cooling rate has a major influenceon nucleation and nucleus growth and hence onthe structure that develops.

The more slowly cooling takes place (through highcavity surface temperatures), the higher the degreeof crystallization and the greater the level ofshrinkage.

If the temperature falls too rapidly, then it is possible for nucleation and nucleus growth to besuppressed. This will give rise to a structure with a low degree of crystallization and hence to a low molding shrinkage. This can then lead to more pronounced post-crystallization and hence to undesirable post-shrinkage.

2.1.3 Fillers and reinforcing materials

Fillers of a spherical shape lead to reduced shrinkage on account of their lower CLTE whencompared with plastics /1/.

Glass fibers have an even more pronouncedeffect on shrinkage than spherically-shaped fillers/6/. The glass fibers constitute an additional internalrestraint, which impedes thermal contraction in the direction of the glass fibers and thus leads to lower shrinkage values. Perpendicular to thedirection of orientation, the fibers act in the sameway as the fillers referred to above and similarlyreduce shrinkage, albeit to a lesser extent /4/.

Table 1 shows that, in polyamide 6, it is possible to influence shrinkage to a large extent throughthe filler content and filler type.

Page 15


Through the use of glass fibers, it is possible toreduce shrinkage by 50 to 80 % in the longitudinalfiber direction. Adding more than 20 to 25 % glassfiber has no further effect on the shrinkage behavior of semi-crystalline thermoplastics /2/.

This is illustrated in Figure 8, which shows how the shrinkage changes with varying glass fiber content and progresses towards a limit in semi-crystalline Durethan AKV PA66 resin.

Page 16

Table 1 Influence of filler on the shrinkage behavior of polyamid PA6 (Holding pressure: 500 bar)

Material

Durethan B 30S

Durethan BM 240 H 2.0

Durethan BKV 30 H 2.0

Durethan BG 30 X

Filler

non-reinforced

mineral

30 % glass fibers

15 % spheres / 25 % glass fibers

Shrinkage longitudinal / transverse

1.0 / 1.2

1.2 / 1.2

0.2 / 0.8

0.3 / 0,9


Figure 8Influence of glass fiber content on shrinkage for semi-crystalline Durethan AKV (PA66, GF) /2/

Mo

ldin

g sh

rin

kag

e i

n %

AKV 15PA66GF15

0

1.4

0.2

0.4

0.6

0.8

1.0

1.2

longitudinaltransverse

Shrinkage measurement panel 150 x 90 x 3 mmMold temperature 80 C

AKV 25PA66GF25

AKV 30PA66GF30

AKV 35PA66GF35

2.2 Processing

The qualitative correlations between the individualprocess parameters and the molding shrinkage aredisplayed in Fig. 9.

The correlations between molding shrinkage andprocess parameters are explained in more detailbelow.


Figure 9 Influence of processing parameters on shrinkage behavior /2/

Sh

rin

kag

e

Pressure holding time

Sh

rin

kag

e

Holding pressure level

Sh

rin

kag

e

Cavity wall temperature

Sh

rin

kag

eDemolding temperature

Sh

rin

kag

e

Injection velocity

Sh

rin

kag

e

Melt temperature


2.2.1 Pressure holding time

The purpose of theholding pressure is tooffset material shrinkageby conveying more melt into the cavity.Under the action of the holding pressure,the material in the cavity is compressedand volumetric contraction due to the cooling process is offset. The holding pressure time thushas a major influence on the amount of extra meltinjected and on the shrinkage compensation. Thelonger the hold time, the lower the shrinkage.

With amorphous materials, the holding pressuretime has slightly less influence than with semi-crystalline materials due to the reduced additionalvolume conveyed into the cavity /1/.

Using large gates with generous cross-section dramatically increases the time over which holding pressure can act. Always position gates in the area of thickest wall section.

2.2.2 Holding pressure level

With both amorphousand semi-crystallinethermoplastics, the levelof holding pressure hasa decisive influence onthe degree of shrinkage.The higher the holdingpressure, the lower themold shrinkage.

The extent to which shrinkage can be influenced by holding pressure is regressive, however. In otherwords, as the holding pressure increases, thereduction in shrinkage becomes less pronounced.

With an optimally designed gate system and molded part, it is possible to reduce shrinkage byup to 0.5 % in semi-crystalline thermoplastics byincreasing the holding pressure. With amorphousmaterials, values of no more than 0.2 % areachieved on account of the lower shrinkage potential /1/.

Page 19

Sh

rin

kag

ePressure holding time

Sh

rin

kag

e

Holding pressure level


Figure 10 shows how dissimilar shrinkage occursdue to the dissimilar action of the holding pressureclose to and remote from the gate.

Figure 10 Dissimilar action of holding pressure close to andremote from the gate with a centrally-gated circulardisk in a non-reinforced thermoplastic /2/

2.2.3 Cavity wall temperature

The molding shrinkageincreases with the cavity wall temperature.Due to the influence ofcavity wall temperature,various different factorsare superimposed whichaffect the flow process(holding pressurephase), the crystallization and the inherent stress profile. This holds true for semi-crystallinematerials in particular.

Page 20

Sh

rin

kag

e

Cavity wall temperature

200 bar

400 bar

600 bar

0.5 %

0.3 %

Intended shape

Gate

Dome warpage


Apart from enhancing or impairing the action ofthe holding pressure, the mold wall temperaturealso has a strong influence on the cooling rate (Fig. 11).

A low mold temperature TW (40 C) leads to high cooling rates and hence to a low level of crystallization in the polyamide, as shown in Fig. 11. This leads to low molding shrinkage followed by high post-shrinkage.

With a high cavity temperature TW (120 C), agreater proportion of molding shrinkage takesplace right away (high degree of crystallization).The post-shrinkage potential is drastically reduced.Overall shrinkage is more or less identical at bothprocessing temperatures (Fig. 13).

TW VCR SMTW VCR SMTW = mold temperatureVCR = cooling rateSM = molding shrinkage

Figure 11 Shrinkage as a function of the cavity surface temper-ature for Durethan B 30 S (PA6, non-reinforced)

Page 21

longit.

0

Molding shrinkage

Post shrinkage

Shrinkage measurement panel150 x 90 x 3 mm

0.5

1.0

1.5

transv. longit. transv.

40 C 120 C

Sh

rin

kag

e i

n %


2.2.4 Melt temperature

The melt temperaturealso has an influence onshrinkage behavior. Twocounteracting effectsare at play here. Firstly an elevated melt tem-perature increases thepotential for thermalcontraction in the resin(increased shrinkage, A) and, secondly, it leads to a reduction in the meltviscosity and hence to better packing and,ultimately, to a reduction in shrinkage (B).As a general rule, curve profile B is measured andobserved. The improved packing predominates over the contraction potential. If, however, an unfavorable wall thickness situation and poor packing conditions exist, then raising the melt temperature can actually increase shrinkage.(sink marks, curve profile A) /1/.

When optimizing other process parameters,it is often helpful to maintain a constant melt temperature

2.2.5 Injection velocity

The injection velocityhas almost no influenceon overall shrinkage.Apparently, countereffects such as orienta-tion vs. re-orientationand shear heating vs.pressure distributioncancel each other out.

Page 22

Sh

rin

kag

eMelt temperature

A

B

Sh

rin

kag

e

Injection velocity


2.2.6 Ejection temperature (longer cooling time, increased mold restraint)

The longer the partremains in the mold, thelonger the cooling timeand lower the ejectiontemperature will be.At the same time, thepart will see increasedmold restraint. Particu-larly with semi-crys-talline thermoplastics, lower ejection temperaturestypically cause less mold shrinkage.

With higher ejection temperatures (shorter coolingtimes), a more pronounced temperature increaseoccurs in the outer layers of the molded part.A kind of virtual heat storage takes place. Thisrelieves surface stresses and increases shrinkage/8/.

2.3 Molded part geometry

2.3.1 Wall thicknesses

When the thickness of the molded part is varied,this leads to a quantitative change in all the othervariables that affect shrinkage (such as the actionof holding pressure or the cooling rate).

Given identical melt and mold temperatures, a thinner molded part will cool more rapidly than athick one. All the different thermally-conditionedprocesses (such as crystallization and internalstresses due to cooling) thus have less time inwhich to act. Because the boundary conditionschange with wall thickness it is very difficult tomake a general statement on the influence of thickness /4/.

Despite this, it is still possible to point to basiccorrelations between the shrinkage and the thick-ness of the molded part.

Page 23

Sh

rin

kag

e

Demolding temperature


Figures 12 and 13 show the correlation betweenshrinkage and wall thickness for semi-crystallineand amorphous thermoplastics. Especially in thecase of non-reinforced semi-crystalline thermo-plastics (Fig. 12 left), the wall thickness is seen tohave a major influence on shrinkage. The thickerwalls lead to slower cooling, and thus give rise to more favorable crystallization conditions.

The degree of crystallization is higher, increasingthe level of shrinkage. These figures show thatthere can be large shrinkage differentials if thereare different wall thicknesses in a plastic moldedpart. If dissimilar shrinkage occurs, then warpage can result. This is more pronounced with semi-crystalline thermoplastics than with amorphousthermoplastics (Fig.13).

Page 24

Figure 12 Correlation between shrinkage and wall thickness for semi-crystalline thermoplastics /2/

0

freshlymolded

00

1 2 3 4 5 6 7 8 9 10

Sh

rin

kag

e i

n %

Wall thickness in mm

Sh

rin

kag

e i

n %


0 1 2 3 4 5 6 7 8 9 10

1.01.0

2.0

Durethan B 30 S (PA6, non-reinforced) Durethan BKV 30 (PA6, GF30)

freshlymolded

after 2.5 %waterabsorption

after 1.5 %waterabsorption

2.0


Figure 13 Correlation between shrinkage and wall thickness for amorphous thermoplastics /2/

0

A

0

1 2 3 4 5 6 7 8 9 10


A: Makrolon 2805 (PC, non-reinforced)B: Makrolon 8035/9425 (PC, GF)

Sh

rin

kag

e i

n %

0.2

0.4

0.6

0.8

1.0

0

0

1 2 3 4 5 6 7 8 9 10


Novodur P2H-AT und P2M(ABS, non-reinforced)

Sh

rin

kag

e i

n %

0.2

0.4

0.6

0.8

1.0

B


2.3.2 Reduced wall thickness at the endof the flow path

As Fig. 14 shows, the holding pressure has less ofan effect farther from the gate than close to thegate. This leads to higher shrinkage at points farfrom the gate, which means that it is useful toreduce the wall thickness at these points (at theend of the flow path). The wall thickness reductionmeans that there is a lower shrinkage potentialin these areas, so that the shrinkage differentialcompared with the region close to the gate is nolonger so pronounced.

Figure 14 Lower tendency to warp through a reduced wall thickness at the end of the flow path on a centrally-injected circular disk in a non-reinforced thermo-plastic

200 bar

400 bar

600 bar

0.35

%0.3

%

Molding less proneto warpage

Gate

Wall thickness reduction at the end of theflow path

S1 S2

S1 > S2


2.3.3 Ribs

Ribs can have a pronounced influence on moldedpart shrinkage and, in particular, on the uniformityof shrinkage. Ribs should be made thinner than the wall to which they are attached, observing aspecific rib wall thickness to nominal wall thicknessratio. The correlations between shrinkage and wallthickness referred to above mean that the ribs generally shrink less (i.e. they remain longer)than the other molded part dimensions. The resultcan be a warped part.

Figure 15Warpage due to differences in the rib wall thicknessand nominal wall thickness /2/

Page 27

SG

SR

SR < SG

Olaf Zoellner


2.4 Mold

2.4.1 Heating/cooling

Different levels of heating and cooling in differentareas of the molded part produce dissimilar shrinkage and hence dissimilar molded part properties (Fig. 16).

When a different mold temperature exists on the inside and outside of the molded part, the plastic on the hotter side will undergo more pronounced shrinkage. During this uneven cooling,there will be a shift in the temperature profile in the solidifying melt, and hence dissimilar shrinkage potentials and cooling stresses willresult. This type of differential shrinkage is similarto the effect that dissimilar metals experiencewhen coupled together as in a bi-metallic structureand then are exposed to a certain temperaturerange (as in a thermostat gage). Differing degreesof thermal expansion and contraction occur within each metal.

Figure 16 Warpage as a result of heating/cooling differentials inthe mold

Page 28

Temperature

Te

mp

era

ture

60

C

Te

mp

era

ture

60

C

S

Stress

+

Fluid plastic center

Te

mp

era

ture

60

C

Te

mp

era

ture

90

C

S

+

Temperature

Stress

60 C

60 C

60 C

90 C


Warpage can occur in the molded part as a result of the asymmetrical stress distribution thatdevelops (see lower part of Fig. 16).

Therefore, take care to ensure a uniform mold walltemperature in the injection mold. The influence of heating/cooling in the mold is considerablygreater with non-reinforced thermoplastics thanwith glass fiber reinforced thermoplastics, forexample.

In the case of glass fiber reinforced thermoplastics,it is the influence of the fibers on the shrinkagethat predominates.

2.4.2 Gate type

Different molded part geometries frequently callfor different types of gating.

Design gates to allow good packing regardless ofgate type (e.g. film, pin-point or tunnel gate, etc.).Then, the holding or packing pressure is muchmore effective at minimizing shrinkage.

The modified tunnel gate depicted in Fig. 17 willonly provide optimum conditions for an effectiveholding pressure phase (in terms of the influenceon shrinkage) if it is of the optimum design interms of the plastic and the molded part.

The version with an unfavorable design not onlyreduces the effect of the holding pressure but alsoimposes greater thermal and mechanical stressingon the melt during injection.

Page 29


Figure 17 Recommended design for a modified tunnel gate /2/

2.4.3 Gate position

Position the gate in the thickest area of the part,when possible. This allows shrinkage differentials tobe minimized through wall thickness adjustment,if neccessary.

Page 30

not recommended

preferreddead end gate

gate-cross-section

3060

3 Phenomena of relevance to shrinkage


Due to the large number of influencing parameters a large number of shrinkage effects occur, as shownin Fig. 18.

Page 31

Figure 18 Shrinkage and warpage phenomena /1/

A number of these phenomena are explained ingreater detail in the following pages.

Sink marks Linear shrinkage Inherent stress Warpage

SV

SV

Voids

S11S12

SA

SH

SH < SA

T(y) v(y)

TU (y)

Stress cracks

hR

hP

hP > hR

hR > hP

hP = hRTR

TPTR > TP

TW1 < TW2

TW2

TH

Sr = St

St > Sr Sr > St

SV

y x

z

V1

V2


3.1 Warpage

Warpage always results from differing degrees ofshrinkage. This can take the form of

a shrinkage differential in the direction of flow and at right angles to this direction for glass fiber reinforced materials

shrinkage differentials due to dissimilar molded part wall thicknesses

shrinkage differentials due to locally dissimilar mold temperatures

shrinkage differentials due to locally dissimilar holding pressure action

The high shrinkage potential of semi-crystallineplastics means that these materials generally suffergreater warpage than amorphous plastics.Figure 19 shows a molded part made of three different plastics:

Pocan 3235 PBT 30 GF (top of Fig. 19) Durethan BKV 30 H PA6 30 GF (center of

Fig. 19) Makrolon 8035 PC 30 GF (bottom of Fig. 19)

It is clear from Fig. 19 that greater warpage occursin semi-crystalline materials than in the amorphousPC with 30 % GF. The reason for this is that theshrinkage potential and the difference in direction-ally-dependent shrinkage behavior is greater in thesemi-crystalline PA and PBT and thus gives rise tomore pronounced warpage.

Figure 19 Warpage of a support plate in different thermo-plastics /2/

Page 32


Figure 20 shows the basic warpage of a circularplate in glass fiber reinforced thermoplastic. Thedifferent warpage variants that can occur are evident here.

Page 33

Transverseshrinkage

Longitudinalshrinkage

Circumferentialshrinkage correspondsto longitudinalshrinkage

Glass fiber

Possibilities ofwarpage

Gate

Figure 20 Warpage of a circular plate in glass fiber thermo-plastic /2/

3 Phenomena of relevance to shrinkage Page 34

Figure 21 shows that it is possible for differentwarpage behavior to result in reinforced and non-reinforced materials when identical geometries andgate types are employed.

In glass fiber reinforced plastics, shrinkage is caused less through the correlation with wall thickness and more by the orientation.

non-reinforced reinforced

Shrinkage differentialsdue to dissimilar wallthicknesses lead towarpage

Shrinkage differentialsdue to glass fiberorientation lead towarpage

Figure 21 Dissimilar shrinkage behavior in non-reinforced andreinforced thermoplastics /2/


Even when the glass fibers display a uniform orienta-tion, warpage can still occur in flat panels (Fig. 22).This is due to the glass fibers turning around at the end of the flow path, which leads to shrinkage differentials between the center section of thepanel and the area at the end of the panel. If the

inherent rigidity of the panel is too low (wall thick-ness), then warpage will result. This warpage can bereduced by making changes to the geometry (Fig. 23).

Figure 23Eliminating warpage on flat panels /2/

Page 35

Mixed shrinkage

Warpage with thinwall thicknesses

Transverse orientation of thefibers at the end of flow leads tolower transverse shrinkage

High transverse shrinkage

~23 s

s

Warpage

Reducedwarpage

Reducedwarpage

better

Orientation of glassfibers

23 s

s

Figure 22 Glass fiber orientation at the end of the flow path /2/

Olaf Zoellner


Figures 24 and 25 show how the relocation of the gate changed the glass fiber orientation andreduced the amount of warpage of a waffle-iron inPocan 3235 (PBT 35 GF).

Figure 24 Warpage due to unfavorable orientation of the glassfibers

Figure 25 Reduced warpage through longitudinal orientation ofglass fibers /2/

Page 36

Current situation

Optimization


3.2 Internal stresses

Positive interlocking with the mold can preventcertain dimensional changes in the molded partwhile it is still in the mold. When forces act on themolded part from the outside they are known asexternal contraction restraints. Also, internalcontraction restraint can also exist.

The solidification model according to STITZ /4/ canbe used to explain the phenomena that prevail withan internal restraint.

Figure 26 shows the molded part wall thicknessdivided up into layers. As a result of the temper-ature profile over the wall thickness (A), the individual layers have a different shrinkage poten-tial. If the layers are viewed in the uncoupled state,then they can slide over each other and thus contract to differing extents (B).

Figure 26 Thermal contraction in the layer model /4/

Page 37

A Temperature profile

thermal contraction

B Without mechanical coupling of layers

C True profile (mechanically coupled)

s

molded partthickness s

s > s > s

TA T w T = f(y)

s

TA = Ambience Temperature; Tw = Mold Temperature


Since the layers are coupled to each other, how-ever, they are prevented from sliding over eachother. The inside layers are extended and the outerlayers compressed (C).

This coupling makes it difficult for the inside sections to undergo thermal contraction in thelongitudinal and transverse directions. Since thereis no resistance to deformation over the thicknessof the part, the volume reduction is achieved primarily through a reduction in the thickness.

All forms of obstruction to thermal contractionproduce internal stresses in the molded part,which, in turn, affect the level of shrinkage.

3.3 Sink marks/voids

In the case of sink marks, volume contraction leadsto local depressions on the molded part surface,since the solidified edge layer is not yet stableenough to absorb the internal contraction forces.

Sink marks occur primarily at points where a hotmelt core and a thin edge are present at the sametime. The forces of contraction will be all thegreater the bigger the melt core is by comparisonto the overall cross-section. This is particularlytrue with ribs, since heat elimination is less efficient on the side facing the rib (Fig. 27, top).

Page 38


Voids are air cavities inside the molded part cross-section. They occur when the solidified edge layerwithstands the contraction forces and the insidelayers become detached from the outer layers withfurther cooling (Fig. 27, bottom).

Figure 27 Voids and sink marks as shrinkage phenomena

3.4 Corner warpage

The phenomenon of corner warpage is similarlyattributable to shrinkage. The uneven coolingbehavior in the corners causes the inside of thecorner to shrink to a greater extent (Fig. 28). Thisthen leads to stresses and forces which producecorner warpage.

Figure 28 Corner warpage due to uneven thermal behavior

Page 39

accumulated melt leadsto sink marks, henceensure an appropriaterib thickness to basicwall thickness ratio

vaccum/void

s

sR

sink mark

residualmelt

4 References

4 References

/1/ Ptsch, H. G.Prozesssimulation zur Abschtzung von Schwindung und Verzug Thermoplastischer Spritzgussteile Dissertation RWTH Aachen, 1991

/2/ Bayer

/3/ AnonDIN 50014, Normklimate Klimate und ihre technische Anwendung Juli 1985

/4/ Stitz, S.Analyse der Formteilbildung beim Spritz-gieen von Plastomeren als Grundlage fr die ProzesssteuerungDissertation RWTH Aachen, 1973

Page 40

/5/ Zllner, O.Prozessgren beim Spritzgieen von Thermoplasten als Produktionskostenfaktor paper presented at the SKZ Wrzburg,Bayer AG, 1993

/6/ Hoven-Nievelstein, W. B.Die Verarbeitungsschwindung thermo-plastischer FormmassenDissertation RWTH Aachen, 1984

Die vorstehenden Informationen und unsere anwendungs-technische Beratung in Wort, Schrift und durch Versucheerfolgen nach bestem Wissen, gelten jedoch nur als unver-bindliche Hinweise, auch in Bezug auf etwaige SchutzrechteDritter. Die Beratung befreit Sie nicht von einer eigenenPrfung unserer aktuellen Beratungshinweise insbesondereunserer Sicherheitsdatenbltter und technischen Informa-tionen und unserer Produkte im Hinblick auf ihre Eignungfr die beabsichtigten Verfahren und Zwecke. Anwendung,Verwendung und Verarbeitung unserer Produkte und deraufgrund unserer anwendungstechnischen Beratung vonIhnen hergestellten Produkte erfolgen auerhalb unsererKontrollmglichkeiten und liegen daher ausschlielich inIhrem Verantwortungsbereich. Der Verkauf unserer Produkteerfolgt nach Magabe unserer jeweils aktuellen AllgemeinenVerkaufs- und Lieferbedingungen.

Die angegebenen Werte wurden, wenn nicht ausdrcklichanders angegeben, an genormten Prfkrpern bei Raum-temperatur ermittelt. Die Angaben sind als Richtwerte anzusehen, nicht aber als verbindliche Mindestwerte. Bittebeachten Sie, dass die Eigenschaften durch die Werkzeug-gestaltung, die Verarbeitungsbedingungen und durch die Einfrbung unter Umstnden erheblich beeinflusst werdenknnen.

ATI 1120 eEdition: 2001-03-01

KU 21120-0103 e

Bayer AGGeschftsbereich KunststoffeD-51368 LeverkusenInternet: http://plastics.bayer.com

Table of contents1 The concept of shrinkage2 Factors that influence shrinkage2.1 Material2.1.1 Amorphous and semi-crystalline thermoplastics2.1.2 Crystallization behavior2.1.3 Fillers and reinforcing materials

2.2 Processing2.2.1 Pressure holding time2.2.2 Holding pressure level2.2.3 Cavity wall temperature2.2.4 Melt temperature2.2.5 Injection velocity2.2.6 Ejection temperature

2.3 Molded part geometry2.3.1 Wall thicknesses2.3.2 Reduced wall thickness at the end of the flow path2.3.3 Ribs

2.4 Mold2.4.1 Heating/cooling2.4.2 Gate type2.4.3 Gate position

3 Phenomena of relevance to shrinkage3.1 Warpage3.2 Internal stresses3.3 Sink marks/voids3.4 Corner warpage

4 References