7 Factors Affecting Post-Mold Shrinkage and Warpage

The net result is that the parts molded in a hot mold need little or no annealing and exhibit little or no post-mold shrinkage.. Post-mold shrinkage of acetal parts molded at a variety o

Most part shrinkage takes place within a very short

time after the part is molded, typically within sixteen

to forty-eight hours after demolding The reduction in

volume during this initial time period is a result of

so-lidification and thermal contraction as the molded part

cools to room temperature This rapid size change is

influenced by the variables discussed in Chs 2–6:

ma-terial properties, part geometry, the runner and gate

systems, melt temperature, mold temperature,

injec-tion pressure, holding pressure, and so on

The same variables affect post-mold shrinkage,

oc-curring more than forty-eight hours after demolding

Especially important phenomena in post-mold

shrink-age are temperature and moisture conditions during

molding, along with in-service exposure after

manu-facturing This chapter reviews the factors of greatest

influence on post-mold shrinkage.[39]

7.1 Effects of Temperature on

Dimensions

Time and temperature conspire to allow molded-in

stress relaxation and some slight additional

crystalli-zation in semicrystalline materials after the molded part

is ejected Some semicrystalline materials such as

ac-etal, PBT, and PB can shrink as much as 0.5% after

molding The longer the time and the higher the

ambi-ent temperature, the greater the tendency for the molded

part to shrink after molding

Plastics, by their very nature, have more thermal

expansion and contraction than metals When plastics

are constrained by being attached to a metal part, they

may crack or totally fail if exposed to widely varying

temperatures This type of failure is due to the

fre-quent change in stress from tension to compression and

back again under the influence of the temperature

variations

In molding operations, the plastic material is cooled

from the outer surface Solidification occurs against

the mold surface and the solidification front proceeds

from that surface toward the center of the thickness of

the plastic part Several factors affect the rate of heat

transfer from the plastic to the mold The mold

tem-perature is the most significant factor and most

sub-ject to the control of the molder The higher the mold

temperature, the slower the plastic will cool because

the temperature gradient between the molten plastic and the mold wall is lower

Higher mold temperatures slow the cycle and in-crease the in-mold shrinkage, but reduce long-term or post-mold shrinkage The net result is that the parts molded in a hot mold need little or no annealing and exhibit little or no post-mold shrinkage For example,

in molding Delrin® at moderate temperatures, good sta-bility can be obtained with a mold temperature of 90°C (194°F) For more severe conditions, the mold tem-perature for Delrin may need to be as high as 120°C (250°F).[33]

The cooling efficiency of the mold contributes to the cooling rate of the plastic part For example, if cool-ing channels in the mold are placed very near the mold-ing surface, the heat transfer into the coolmold-ing water is quite rapid near the water channels but somewhat slower between water channels This results in a varia-tion of the temperature of the surface of the mold from

a minimum immediately over the water channel to a maximum half-way between the channels The varia-tion in mold temperature across a large, flat surface that results from cooling channels placed too near the surface may cause a visible “ripple” on the surface of the part

Placing the cooling channels at a greater distance from the molding surface results in a more uniform surface temperature At one time it was advocated that cooling channels not be placed in the inserts but in-stead be placed in the holder blocks or the plates im-mediately behind the mold inserts This resulted in very uniform temperatures on the mold surfaces initially, but the continuous, very slow heat-transfer ultimately caused a rise in the mold surface temperature This

“uniformity” theory actually can result in a reduction

of mold-temperature consistency

If there are mold details that are difficult to cool, remote cooling lines increase that difficulty and increase the mold surface-temperature variations In addition,

if there are mold cycle-time variations, as there fre-quently are with manually operated molding machines, the mold surface temperature drops more during any delays (such as when the operator sprays the mold sur-face, smokes a cigarette, drops a part, extracts a stuck part, etc.) After a delay such as this, the next few parts are molded in a cooler mold than those molded during

a consistent cycle

In some cases, it is impossible to maintain

abso-lutely uniform mold surface temperature Very small

and long core-pins cannot be effectively cooled

through-out their length Usually, most of the cooling around

such core pins is from the outside surface of the part

around the cored hole, with little of the heat transferred

through the core pin A similar problem exists in the

vicinity of sharp, inside corners of a molded part This

type of uneven cooling shifts the neutral axis toward

the hot side of the part and increases the tendency

to-ward warpage

As the plastic part cools, it pulls away from the

mold surface due to volumetric shrinkage The lower

the packing pressure, the sooner the separation occurs

As the plastic pulls away from the mold wall, there is a

sharp reduction in heat transfer from the plastic to the

mold wall This happens because dead air space is an

excellent insulator A vacuum is an even more

effec-tive insulator and a vacuum is often present as the

plas-tic shrinks away from the cavity wall because there is

no source for air until the mold opens Inadequate

pack-ing pressure can cause significant variations in the

cool-ing rates thus coolcool-ing inconsistency across the surface

of a molded part as a result of this type of separation

In summary, higher mold or melt temperature

re-sults in less post-mold shrinkage However, higher mold

temperatures are often localized because of inefficient

cooling Localized hot spots cause shrinkage variation

and warpage Post-mold annealing can accelerate the

post-mold shrinkage and minimize later size change

Parts molded in cooler molds can be annealed (stress

relieved) to achieve better mechanical properties and

stability in the final part Fixturing may be required to

stabilize parts during the annealing process

Fixturing is a complex process and should only be

used when molded parts require very tight tolerances

and exposure to high temperatures for prolonged

peri-ods while in use Attempts to reach good dimensional

stability by annealing parts molded in a cold mold are

likely to lead to high post-molding shrinkage and may

introduce stresses causing uncontrolled deformation

This is especially true for semicrystalline materials such

as acetal or nylon

Post-mold shrinkage of acetal parts molded at a

variety of mold temperatures when exposed to

differ-ent temperatures for 1000 hours are shown in Fig 7.1

The annealing procedures for the parts showing the

least shrinkage in the charts in Fig 7.1 were subject to

the following guidelines:

• Parts should be exposed to air or an inert

mineral oil at 160 ±3°C for 30 minutes

plus 5 minutes per mm of wall thickness

• Overheating and hot spots should be avoided

• Parts should neither contact each other nor the walls of the container

• Parts should be left in the container to cool slowly until 80°C is reached

• Stacking or piling, which may deform the parts while they are hot, should be de-layed until the parts are cool to the touch

• Annealing can also be used to test molded parts to determine their long-term stabil-ity and size change Annealed parts closely resemble the ultimate size of the parts after long-term use

For maximum in-service stability of the molded part, mold temperatures should be near the high end of the plastic supplier’s recommendations For example, post-mold shrinkage can be estimated for Delrin® ac-etal from Fig 7.1.[33]

7.2 Effects of Moisture on

Dimensions

Post-mold size change also can come about as a result of absorption or loss of fluids such as water or plasticizers The loss of plasticizers causes a plastic part to become more brittle and to shrink How many automobile dashboards have you seen that have lost color or cracked? This type of failure is caused by the loss of plasticizers

Some materials are hygroscopic; that is, they at-tempt to absorb moisture from the environment As they absorb moisture, the material properties change Sometimes the materials become tougher, usually there

is dimensional change Figure 7.2 shows the change in size due to moisture absorption of Zytel® 101.[9] Size changes for Delrin® 100 and 500 are shown in Fig 7.3.[33] Other moisture absorption curves can be found in the material-specific data section (Ch.11 of this book) Nylons are strong materials with good chemical resistance, but they absorb large amounts of water if immersed It is not generally considered a good appli-cation for nylon if the part is to be immersed in or continually exposed to water unless full consideration

is given to the amount of post-mold growth that nylon can experience in water Applications using nylon have failed because the nylon parts that were immersed in water swelled so much that they did not allow the mov-ing parts to move freely Some nylons can absorb

mois-ture to such an extent that the totally saturated nylon part is larger than the cavity in which it was molded Figure 7.2 shows the dimensional change of nylon

as it absorbs moisture The change shown here is not necessarily equal in flow and cross flow The measure-ment direction is not specified but is probably in the flow direction.[13]

Figure 7.3 implies that the molded part was prob-ably a tensile test (dog-bone) specimen and that the measurements were along the long or flow-direction axis There is no indication that the cross-flow changes are the same

The presence of moisture during molding inhibits

a glossy surface Moisture usually causes surface splay-ing (which normally manifests itself as silvery streaks parallel to the flow direction of the plastic, sometimes

as irregularly shaped silver spots) or other imperfec-tions because it inhibits close contact with the cavity wall and can cause foaming or voids within the molded part

Moisture in the plastic pellets as they enter the heating section of the molding machine often cause plas-tic-property degradation because of chemical reactions between the plastic and superheated steam

Table 7.1 shows the equilibrium water absorption percentages for several polyamides.[9] Nylons must be molded dry to avoid material degradation, but in the dry condition, they tend to be brittle When they have absorbed moisture, they become tougher

Figure 7.1 Post-molding shrinkage of Delrin® acetal

resins [33] (Courtesy of DuPont.)

Figure 7.2 Size change of Zytel® 101 vs moisture absorption [9] (Courtesy of DuPont.)

The 24-hour absorption levels of water by nylon compared to the equilibrium levels of water in nylon in

an environment where the relative humidity is less than about 25% are as follows:

Type of nylon 24 hours Equilibrium

in water % of water content

Figure 7.4 shows longer-term water absorption for Nylon 11 and two other grades.[13] Note that Nylon 6 absorbs significantly more water than the other grades

In most cases, it is a good idea to condition nylon parts

in hot water before placing them in service to stabilize the moisture absorption and increase the toughness of the nylon Dry nylon as molded is relatively brittle Suppose a flat part is exposed to water on one side and a dry environment on the other The bow-shaped warpage as shown in Fig 7.5 could take place The same sort of warpage can take place if one side of a part is coated with an impermeable layer and the other side is left uncoated

Plastics will absorb all kinds of fluids to a mea-surable level Inspection of the chemical compatibility

of the plastic in question will give a good indication of likely absorption of a particular fluid If a supplier states that a plastic is compatible with a particular fluid or is resistant to that fluid, it can be assumed that after two weeks of immersion, the plastic will absorb an amount

of fluid that is less than 1% of the weight of the part.[13]

Figure 7.3 The effect of temperature and moisture content

on the dimensions of Delrin ® 100 and Delrin ® 500 [33]

(Courtesy of DuPont.)

Absorption Polyamides In Water at 20°C

(%)

In Air at 50%

RH, 23°C (%)

Table 7.1 Water Absorption of Nylons in Air and

Water

Figure 7.4 The percentage of water absorbed by some

grades of nylon over long periods of time.

Many plastics contain mobile fluids such as

plas-ticizers, antistatic agents, lubricating oils, dyes, etc

Most users are aware of the problem of plasticizer

mi-gration and that plasticizer loss will cause significant

changes in dimensions (shrinkage) The migration of

mobile fluids is accelerated by contact with a wide range

of organic fluids which, having greater affinity for the

plasticizer than the molded plastic, may cause rapid

shrinkage.[13] Some materials contain plasticizers

with-out this being explicitly stated Flexible grades of

cel-lulosics and nylons (particularly Nylon 11 and Nylon

12) are quite common, and these will be prone to

mi-gration-induced shrinkage, just as will any plastic

con-taining mobile fluids

Figure 7.6 shows the moisture absorption as a

per-centage of the weight of the part of certain glass-fiber

plastics immersed in water.[40] This figure does not

dif-ferentiate between hygroscopic and non-hygroscopic

materials, but rather suggests at least some moisture

migration along the glass fibers into the plastic part

From Fig 7.7 it is obvious that nylon is

hygro-scopic and its level of water is strongly affected by the

environment.[35] The more water that is available, the

more nylon absorbs to reach equilibrium

The time that is required for a plastic part to reach

an equilibrium condition, for any given moisture

con-centration, is affected by the environmental

tempera-ture and thickness of the plastic part The thicker the

part, the longer it takes for the moisture to migrate

through the plastic and uniformly permeate the part

Figure 7.8 shows how thicker walls of Zytel® 101 take

longer to reach equilibrium.[35]

The equilibrium condition for this material is the same, about 2% to more than 5% moisture, no matter how thick the walls are This graph indicates that a 1.5-mm thick wall reaches equilibrium in about 6 months, but the thicker walls may not reach equilib-rium in a year

Figure 7.9 shows another nylon resin that has not reached equilibrium in thicker sections in a year.[35] When immersed in water, these same two resins approach equilibrium more rapidly than at 50% RH in air See Fig 7.10.[35]

Figure 7.11 shows the time required to condition Zytel® 101 to 3% moisture and to saturation for vari-ous wall thicknesses.[35]

Figure 7.12 shows that nylon can increase in size

as a result of moisture absorption as much or more than it can shrink out of the mold (as much as 0.025 inches per inch).[35]

We have dealt here primarily with size change of nylon due to absorption of water The wrong chemical can affect any plastic While water is probably the most common environmental fluid that is likely to be ab-sorbed by a plastic, and some plastics react more strongly to its presence than others, many plastics re-act adversely to hydrocarbons that are quite common

in the petroleum and automotive industry Check the plastic’s reaction to known or suspected chemicals that are likely to be present in the expected environment

Figure 7.5 Potential warpage (exaggerated) due to

non-uniform exposure to moisture.

Figure 7.6 The percentage of moisture absorption (but not

the size change) of a variety of plastics as a result of immersion in water [40](Courtesy of Hoechst Celanese.)

Figure 7.7 The equilibrium conditions of moisture content

vs relative humidity for a variety of Zytel ® nylon resins [35]

(Courtesy of DuPont.)

Figure 7.8 Moisture content vs time for Zytel® 101F exposed

to 50% RH air at 23ºC [35] (Courtesy of DuPont.)

Figure 7.9 Moisture content of Zytel® 151 as time passes

when the Zytel is exposed to air at 50% RH at 23°C Three

different thicknesses are shown [35] (Courtesy of DuPont.)

Figure 7.10 Moisture content vs time for Zytel® 101 and Zytel ® 151 when immersed in water at 23°C [35] (Courtesy

of DuPont.)

7.3 Creep

While it is not strictly a shrink or warp

phenom-enon, if a plastic part is loaded to a significant fraction

of its tensile strength, it can be subject to creep failure

For most practical purposes, plastic can be thought of

as molasses in January in Alaska Fiber fillers increase

the stiffness of plastics but they do not eliminate the

tendency to creep As a general rule, it is unwise to use

thermoplastics as load-bearing structures without huge

safety factors or extensive, long-term,

elevated-tem-perature testing For this type of application, the creep

data for the plastic is much more significant than the

tensile or compressive strength

Creep is a phenomenon that is foreign to most

de-signers Most thermoplastics are subject to at least some

creep Amorphous thermoplastics are similar to glass;

the slow rate of creep has no limit Semicrystalline

materials are somewhat more rigid and the creep rates

tend to diminish over time The physical property data

for a given plastic is for short-term loading Long-term

deflection versus stress is rarely published Before

marketing a product that is exposed to long-term stress

Figure 7.11 Boiling times to condition Zytel® 101 [35]

(Courtesy of DuPont.)

Figure 7.12 The size change of Zytel® 101 in the stress-free (annealed) condition as it absorbs moisture [35]

(Courtesy of DuPont.)

of any significant part of the tensile strength of the material, long-term measurement of deflection (six months minimum exposure) should be conducted The test should be conducted at the highest expected stress and at the highest expected environmental temperature Any significant deflection over time would indicate the need for additional structural support

It does happen that product suppliers do introduce new resins that have had only short-term testing A few years ago, a company introduced a new large prod-uct line in which the thermoplastic was expected to carry significant structural loads The initial short-term testing of the product yielded outstanding results How-ever, after six months to a year in the field, the product sagged to the point that it became unacceptable for the intended purpose This ultimately led to bankruptcy of the company Had the long-term creep characteristics

of the thermoplastics been recognized, other structural elements could have been included in the design that would have produced an excellent product Unfortu-nately, the failure to recognize the creep characteris-tics of the plastic led to the company failure and added another black mark to consumers’ concepts of plastic