DuPont Engineering Polymers

Failure Analysis
using Microscopic Techniques

Edith Bhme

The DuPont Oval, DuPont, The miracles of science, DELRIN, ZYTEL are registered trademarks
or trademarks of DuPont or its affiliates
Caption for front page illustration:

Thin polished section (10 m) of a ball pen tip moulded

of DELRIN 500 NC010 (Acetalhomopolymer).
Microscopic photography taken in transmittant
polarized light using first order red quartz filter.

Magnification: 40
 Edith Bhme Failure Analysis
on Moulded Parts of
DELRIN and ZYTEL
using Microscopic Techniques

Delrin and Zytel are registered trademarks or trademarks of DuPont or its affiliates

1
 Note about the author

Mrs. E. Bhme is a Technical Consultant with Engineering Plastic

Materials of Du Pont de Nemours International S.A. Geneva.
For more than 20 years Mrs. Bhme has worked with processors
and end-users of engineering plastics. During this time she has
developed techniques of failure analysis using microscopic structural
analysis and demonstrated on countless occasions the value of these
studies in understanding part failure.
This work is intended to provide users of engineering materials
with a guide to the practical application of this technology.

2
Introduction

The increasing use of failure analysis on thin sections prepared from

finished parts by many laboratories demonstrates that the study of
the structure of crystalline material is a useful tool to reduce rejects,
optimize moulding conditions and improve the quality of the
moulded components. However, once a fairly good thin section has
been prepared, it is not always easy to interpret what is seen in the
structure and to pinpoint the reason of failure. Experience is needed
to be able to identify what has gone wrong in the production, tooling
or even designing.
Over the years many cases have come across the authors table
and since there are few publications on failure analysis by transmit-
tant light microscopy it was the intention to issue a kind of refer-
ence book to assist the newcomer in this field of technology.
The emphasis is on the pictures rather than on the comments
which have been kept short to provide more space for the micro-
scope photos of the thin sections. Some of the failures shown may
occur single or in combination within the same part.
The cases described herein demonstrate the usefulness of failure
analysis on crystalline materials using microscopic techniques.
However, these techniques should not be regarded as the only tool
for quality control. A thin section of 5-20 m prepared from a
moulded component does not permit a statement about the quality
of a whole series. However, if parts fail in testing or in use, an analy-
sis of the structure of a thin section from such a part in comparison
to a good one will be of great help in recognizing the possible cause
of failure and hence preventing them in the future.
The author realizes that the booklet is only the first step towards
a more complete guide book and revisions may be necessary as the
techniques advance and more knowledge is gained about the mor-
phology of crystalline materials. We hope and wish that the present
version may become useful and handy for day-to-day problem
solving.

Acknowledgment

Many moulders and end-users have generously provided us with

samples, parts and information of their products and so permitted
us to collect a variety of examples. We extend to them all our sincere
thanks for their cooperation.
Many thanks to my colleagues who assisted with their efforts and
advice towards the accomplishment of this booklet.

3
 Index

5 Methods of Obtaining Thin Sections

of Semi-Crystalline Polymers for Structural Analysis

13 Unmelt or Inhomogeneous Melt

Examples (p. 24-27)

15 Weld Lines in Mouldings

Examples (p. 48-58)

16 Mould Temperature
Examples (p. 59, 60)

19 Screw Forward Time (SFT)

Examples (p. 61)

20 Shear Orientation
Examples (p. 63-67)

20 Flash/Sharp Corner = Notch Effect

Examples (p. 72-75)

21 Voids
Examples (p. 62)

21 Colour Agglomeration
Examples (p. 68-71)

22 Glass fibre Orientation

Examples (p. 76, 77)

22 Welding
Examples (p. 78-83)

23 References, General

23 References on Preparation Techniques

4
 Methods of Obtaining Thin
Sections of Semi-Crystalline
Polymers for Structural Analysis
Structural analysis using optical microscopy is a method for
analysing the reasons for failure of unreinforced semi-crystalline
polymers, e. g. polyacetal (POM) and polyamide (PA). It is a tech-
nique often used in the laboratories of moulders as well as end-users
in the plastics industry.
From systematic studies of the influence of the processing param-
eters on the crystalline structure of moulded thermoplastic resins
in relation to their mechanical properties, data are now available
which allow many causes of failure to be determined by structural
analysis alone. Several authors in this field 3-10 give interpretations
of observed phenomena in the structure of moulded parts.
Unfortunately, this failure analysis technique is not used as fre-
quently as it should be because at first it seems to be difficult to
prepare a thin and thus light transmittant sample. Failures arising
from incorrect processing become visible only in a light transmittant
sample with the aid of polarisation filters and sometimes with the
additional use of a first order red quartz wave plate inserted in the
optical path to further delineate structural differences.
ln the following text, two preparation methods are described for
obtaining thin, light transmittant specimens. One is the method of
cutting a thin section with a microtome, as used in the medical field.
The other is a polishing method, a technique with instruments used
mainly in metallurgy. A third method, originating from the field of
medicine, but also used in metallurgy, is that of micromilling. The
use of such a device has been described in publications b, c, d, h and will
not be discussed here.

Advantages and The decision as to which method to use for the preparation of a thin
Disadvantages of section depends mainly upon the complexity of the object to be
Microtome Cutting and analysed and/or the time available for the analysis.
Polishing Techniques The microtoming technique provides a thin section for micro-
scopie inspection within 10-20 minutes. However, the microtome
holding device sets certain limits on the size and complexity of the
part to be cut.
The polishing technique provides a thin section for microscopic
inspection, but takes a few hours to achieve depending on the choice
of embedding resin and the time required to cement the sample on
to the glass support. However, the polishing technique allows frac-
tured samples to be observed as a unit and/or at angles which would
not have been possible with the sample holder device of the micro-
tome.
The polishing technique usually provides sections having better
definition than those obtained by microtoming (see photos for com-
parison, p. 6).

A. Cutting Technique The preparation of a light transmittant section by the microtoming

(Microtome Cut) technique is done in two steps:
Cutting with a microtome
Mounting of the thin section on a glass support for microscopy

5
a
Thin Polished Section

b
Microtome Cut
(done with an already abused
knife)

a b
Thin section
cut with the microtome

Cutting with a microtome Thin sections can be prepared with all microtomes available on the
market provided they have a good solid base. Special attention has
to be paid to the cutting knives and the cutting angle of the knife. The
knife should be positioned at a 40-45 angle to the part. Tungsten
carbide tips are recommended for cutting POM and PA. The part to
be sectioned must be mounted rigidly in the sample holder of the
microtome, but without applying too much pressure so as to avoid
irregular thicknesses within the cut-off section. The free portion of
the sample should not exceed 2 mm. The preferred thickness of the
section is between 10-20 m. In some cases too thin a cut does not
provide the desired information. It is advisable to prepare sections
from 25 m downwards to see in which thickness the failure can be
identified best.
6
 If it is not possible to clamp the part or a section of the part direct-
ly in the holder, it may be embedded in a suitable resin and then
cut after the resin has hardened. Often the thin section will sepa-
rate from the embedding resin when cut and can then be mounted
on the glass support. If not, the section can be mounted together
with the surrounding embedding resin. Another possibility is to
embed only part of the sample to be cut so as to provide a solid
foot to be clamped. For that purpose there are embedding resins
on the market which harden quickly within 2-10 minutes. However,
the free portion should not be too high in order to prevent bending
while cutting.

Mounting on Using tweezers, the section is mounted on a microscope slide

the glass support (76 26 mm) in Canada Balsam (neutral in xylene). A drop of the
balsam is spread on the slide. The sample is carefully flattened out
(a long or thick section will curl severely) and, still using tweezers, is
pressed into the balsam. The thinner the cut the less the section has
the tendency to curl. Balsam is then spread on the cover glass which
is placed over the specimen. The cover glass is pressed carefully
either by hand or with tweezers to distribute the balsam evenly.
The slide and coverglass assembly with the thin section in
between, positioned horizontally, is heated on a hot plate (50-60C)
for 5-10 minutes. A weight with a flat bottom, cut to the diameter
of the coverglass (20 mm and approximately 350 g), is placed on
the assembly.

Thin section prepared

in Canada balsam

This squeezes out the air bubbles in the Canada balsam. When
cooling, the heated weight is exchanged for a cold one of approxi-
mately the same dimensions. This quick cooling prevents a retrac-
tion of the balsam and a possible recurling of the thin section. As
soon as the mounted microtome is cooled it is ready for microscopic
inspection.

7
B. Polishing Technique c, f, h The preparation of a thin light transmittant section by the polishing
technique is done in four steps:
Embedding
First polishing
Cementing
Second polishing

Embedding The embedding is done using the procedure established in metal-

lurgy. The choice of embedding resin (Epoxy or Polyester resin)
depends on the type of polymer to be encapsulated. Some of the
embedding resins develop heat during the hardening process which
may effect some of the polymers. With regard to DELRIN and
ZYTEL there are no restrictions concerning the choice of embedding
resin. Good results have been obtained with EPOFIX (manufactured
by Struers, Denmark) and SCANDIPLAST 9101 (manufactured by
Scan DIA, Germany). Other products on the market are certainly
suitable as well.

Figure 1:
Textile machine thread guide
in PA 66

Figure 2:
Ballpoint pen tip in POM

1 2

Embedding resins having very little shrinkage are recommended

in order to prevent the embedded part falling off when polishing.
Some resins harden in a short time with or without exposure to ele-
vated temperatures. Others cure at room temperature after 16 hours
or more. To be able to make a failure analysis quickly a resin having
a short curing time is preferred.
If the part to be embedded contains hollow sections, it is sug-
gested that these are filled with the aid of a disposable hypodermic
injection needle before the part is embedded. This procedure pre-
vents the formation of air bubbles in and around the specimen as
8
 well as a possible delamination of the specimen from the glass
support during polishing. Should access to the hollow section be dif-
ficult, then a vacuum may have to be applied. Small parts can be
fixed to the bottom of the container for encapsulation by using a
double faced adhesive band to prevent the specimen floating to the
surface.

First polishing The resin with the encapsulated specimen is polished by hand on a
wet paper grinder. Starting with a paper SIC-grain of No. 320, the
polishing is continued until the level of the desired microscopic
observation is reached. Then this surface is ground with increas-
ingly finer grain paper (e. g. 500, 800,1000, 1200) to improve the
quality of the surface for microscopic inspection. If available, auto-
matic wet paper grinding machines are preferred because the overall
flatness of the specimen surface will be improved. The better the
quality of the specimens surface the better the adhesion to the glass
support in the cementing step. Consequently a better quality of the
thin polished sample can be obtained in the second polishing step.

Ballpoint pen tip

after first polishing

Cementing The polished surface of the specimen is now glued on to the glass
support (76 20 mm). Araldit D with hardener HY956 (manufactured
by CIBA-GElGY) has been used with success for POM and PA. Other
glues available on the market may also work well.
The suitability of the cement for each case should be checked
beforehand. To achieve a uniform contact between the glass support
and the specimen, and at the same time to obtain a uniform thin
layer of the cement, a weight of about 500 g is placed on top of the
embedded block. Preferably the weight should have a pin in its
centre, approximately 5 mm and a few centimetres in length, in
9
Mounting Bloc

Mounting Bloc Detail

order to apply the load centrally and uniformly onto the specimen
(see photos above).
The hardening process can be varied with time and temperature
depending on the recommendation of the manufacturer of the
cement and the material of the embedded plastic component.

Second polishing To reduce the polishing time for the thin section, the larger portion
of the embedded specimen is cut off with a small saw by cutting just
alongside the glass support. The remaining sample of about 1 mm
thickness can now be polished as in the first polishing step starting
with paper grain size No. 320 and finishing with No. 1200.
The glass support can be held with the aid of a rubber suction cap
as used in many garages to polish automotive valves. With this
accessory, thin sections of 10 m thickness and less can be obtained.
About 10 m additional thickness must be accounted for the cement.
While polishing, intermediate controls of thickness as well as of the
microscopic quality are recommended. As with the microtome

10
Figure 3:
Gear (3 mm ) after second
polishing step with paper
grain 320

preparation, sometimes too thin sections may not provide the

desired information.
Microscopic inspection of a polished thin section is improved if a
small drop of immersion oil is placed between the sample and the
cover glass. Should it be necessary to store the thin section for a
longer period of time, the oil can be removed with alcohol and
Canada balsam is then used in the same manner as for microtomed
sections.

Second polishing with the aid

of a rubber suction cap

Figure 4:
Thin polished section of
the textile machine thread guide
(Fig. 1), photographed with
polarized transmittant light

Figure 5:
Thin polished section of
the gear (Fig. 3), photographed
with polarized transmittant light

4 5

11
Summary The techniques described are, in reality, simpler to apply than it may
seem at first glance. Once they are used and some skill has been
achieved, a structural analysis can provide useful additional infor-
mation for failure analysis studies. For example, when a sudden
increase in the number of rejects is noticed in production or in the
quality control of delivered parts.

12
 Unmelt or Inhomogeneous Melt

ln some cases when moulded parts show unusually low perform-

ance or extruded material is brittle when machining, the reason may
be the presence of unmolten particles or a non-uniform structure.
Unmolten particles might be visible already with the naked eye while
the latter will become visible in a thin section inspected with trans-
mittant polarized light.
The reason for the presence of unmolten particles in most cases is
a too low or wrong temperature profile for the melt in the barrel of
the moulding machine.
The reason for non-uniform structure, i. e. non-uniform crystal-
lization can be:
insufficient mixing of the melt in the barrel (which sometimes is
more pronounced if a colour master batch is used);
inadequate screw used for moulding barrel;
a heater band on the moulding machine went out of function
causing loss of heat for the passing melt at a certain area of the
barrel;
a combination of the above.
ln all cases the mechanical properties suffer resulting in lower
elongation, lower impact and increase in notch sensitivity compared
to parts moulded at optimum processing conditions and thus having
a uniform crystallinity growth.
It should be emphazised that these failures (unmelt and inhomo-
geneous melt) are the most frequent ones in failure analyses for
parts moulded from crystalline resins (see graphs, p. 13 and 14).

POM
Type-of-Failure-Distribution PA
analysed by Structure Study PA modified
during 3 subsequent years.

Melt Void Shape Shear Cont. Assem.

13
14
Weld Lines in Mouldings

Weld lines are almost inevitable when a moulded part has an insert,
a boss, a hole, etc. Any insert or core is a sort of obstacle for the
melt filling the cavity. The material flow is separated into two or
more separate flows which will meet again to complete the cavity
filling.
Assuming the melt temperature was already on the low side and
hitting in most cases a cold insert, the material will not merge
well when the separated flows meet. Thus a weld line becomes
visible. Usually it can be seen on the surface of the part, almost
always exactly opposite the gate location.
A visible weld line may not always be harmful as long as it does
not extend throughout the total wall thickness. A microtomed section
prepared perpendicular to the weld line will show how far the weld
line goes through the cross section. ln some cases it is a clear line
which may not always be a straight one, but can form different
shapes, e.g. an S. The part in question is likely to have a reduced
performance. Pressure, impact or other mechanical load in service
may cause premature breakage which will usually occur along the
weld line.
However, in some cases mechanical properties will not suffer from
the presence of one or more weld lines as performance tests may
reveal. Often only empirical evaluation will help to find the accept-
able limits.

15
 Mould Temperature

ln general mouId temperature affects mould shrinkage and dimen-

sional stability of a part. It also influences mechanical properties,
i. e. with increasing mould temperature stiffness increases and elon-
gation at break decreases. Also, the hotter the mould the greater is
the initial mould shrinkage and the better the dimensional stability
after moulding. This is easily proven, e. g. on POM, by annealing a
part after moulding.
The structure of the cross section of ASTM test bars moulded in
DELRIN (POM) under the same conditions but at different temper-
atures is shown in photo 1. The second crystallization zone is much
thicker in the part moulded at the lowest temperature because this
sample has a higher temperature gradient and faster cooling rate
than those in the hot mould. However, the thickness of the second
and center zone in relation to each other is influenced by the wall
thickness of the part (see photo 3, page 18).
As other studies have shown, the centre zone which has the larger
spherulites without preferred orientation, is the most stable one. It
is desirable to mouId polyacetals hot to obtain the stiffest and
dimensionally most stable parts.
The structure of ZYTEL (PA) is influenced as well (see photo 2) by
mould temperature. However, its effect on mechanical properties is
far less and is compensated by the moisture absorption at ambient
atmosphere.

16
Influence of Mould Temperature on the Structure
Microtome sections of moulded tensile bars (3,2 mm thickness)

Mould Wall Temperature Photo 1 Photo 2

30 C

30 C

60 C

50 C

80 C

70 C

100 C

125 C 95 C

DELRIN (POM) ZYTEL (PA 66)

17
 60

35

Photo 3

Influence of mould temperature

and wall thickness on the
structure of DELRIN 500
(Acetal homopolymer)

18
 Screw Forward Time (SFT)

The screw forward time (injection time plus hold time) is certainly
one of the most important factors in moulding parts for satisfacto-
ry performance. When the melt enters the mould, the material will
freeze immediately along the mould wall and shrink. More material
A = 25 sec. SFT can then be pushed into the mould by maintaining the pressure until
such time as the gate has frozen off. Thus it is obvious that gate
design and area are also important factors. ln case the SFT is not
maintained until the gate freezes off, the sudden drop of pressure
when the screw is retracted will influence the material due to the
gate still being open. Pressure variations during screw forward time
affect mould shrinkage, part weight and mechanical properties as
B = 15 sec. SFT
well as crystal growth and configuration.
For example, with a tensile test bar of 3 mm thickness moulded of
standard viscosity acetal homopolymer (DELRIN 500) the following
phenomena occur with increasing SFT: decrease in mould shrinkage,
increase in part weight and increase in elongation at break.
These effects are discussed (under ref. 5). Under optimal mouIding
C = 5 sec. SFT conditions, i. e. if the SFT is maintained long enough until the whole
material is frozen, proper mould shrinkage and better mechanical
properties will result. The SFT is a function of the thickness of the
part and the mould temperature. For a part with a thickness of 3 mm,
such as an ASTM test bar, an SFT of 25 sec. should be sufficient at
a mouId temperature of 90C.
The photo shows the cross sections of test bars of standard vis-
cosity acetal homopolymer (DELRIN 500) moulded under the same
conditions but one with 5 sec. SFT (C), one with 15 sec. SFT (B) and
with 25 sec. SFT (A). ln picture C a second line can be seen in the
structure below the surface which appears all around the cross
section. This line does not appear in the picture of the properly
moulded part, and its presence can be explained as follows:
At the end of the 5 sec. SFT there is a sudden drop of injection
pressure to zero and a correspondingly abrupt decrease in the freez-
ing point. At this moment there is a sudden change of temperature
gradient and hence of crystallization rate. This results in the forma-
tion of a new zone of orientation.
As the SFT is increased (B) the line of new orientation is much
closer to the centre of the bar. This line moves closer to the centre as
the SFT is increased and finally, it disappears completely. With a
25 sec. SFT for a 3 mm thick test bar we reach the point at which part
weight is at maximum together with optimum values for all other
properties.
These structural changes show why it is necessary to maintain
injection pressure constant until the melt is completely frozen.
Variations in pressure result in structural variations which can be
noticed in the mechanical properties.

19
 Shear Orientation

Shear orientation occurs when the molecules receive a preferred

orientation while crystallizing. That is to a certain degree the case
during the first phase of filling the cavity when the material freezes
along the mould surface while still more melt is being injected.
Thus the skin of a moulded part has a preferred orientation perpen-
dicular to the surface. The high orientation prevents the growth of
spherulites and in a thin section prepared from a moulded part the
skin looks transparent (similar to amorphous material) when
observed with polarized transmittant light.
The same kind of effect can be seen in the weld of a part welded
by rotation or vibration because the material molten by high friction
becomes oriented in the direction of rotation or vibration and
freezes/crystallizes under that condition.
Shear orientation will also and inevitably occur at the injection
point. How far this oriented zone reaches into the part can be influ-
enced to some extend by the injection speed. This kind of shear
orientation, however, does not result in a preferred order of crystal-
lization and may cause internal stress. ln cases where the part suffers
impact or tension on such areas, mechanical strength will be
reduced and premature breakage may occur.
Sharp corners due to too small a radius or bad mouId manufacture
can create a shear orientation as well causing premature breakage
of the part.

Flash/Sharp Corner = Notch Effect

Flash on moulded parts can occur when either the clamping force of
the moulding machine is insufficient or the viscosity of the molten
resin is very low or the mould is abused and does not close correct-
ly any longer.
ln case a force is applied at the area of a part where flashing
occured the flash may break and cause a notch effect into the part
which would not break otherwise.
Sharp corners may also cause a notch effect but for different
reasons. Comparing thin sections of a moulded part having a sharp
corner but not yet broken will reveal high stress concentration in that
area while a rounded off corner does not show this effect. The high
stress concentration will cause premature breakage under load.
The reasons can be:
design not respecting the recommended minimum radius for
moulded plastic parts;
given radii had not been respected in tooling the mould;
inserts or mould has been abused.

20
Voids

Large voids or areas of microporosity can be the result of insufficient

filling of the cavity for various reasons such as:
gate size too small and thus the material freezes off too early
preventing injection of more material into cavity;
filling thick wall sections through a thin wall by unfavourable loca-
tion of the gate. Material of the thin section is already crystallized
preventing sufficient filling of material into the thicker section;
cycle too short and by subsequent shrinkage of the material voids
will develop;
material degraded during the moulding process. Consequently,
development of gases prevents complete filling of the cavity.

Colour Agglomeration

ln case a resin is not fully colour compounded, a masterbatch or

liquid colouring is used to obtain the desired shades of colour. Only
very rarely are powder pigments applied.
ln all cases, uniform distribution of the pigment within the pigment
carrier is necessary, otherwise not only the colour aspect is unsat-
isfactory. Additional to the visual effect a decrease in mechanical
properties can be noticed. The pigment agglomeration reacts like a
contaminant resulting in stress concentration and premature breakage.

21
 Glass Fibre Orientation

Evaluation of the glass fibre orientation may sometimes help to

explain differences in shrinkage and/or warpage observed in a part
injection moulded from a glass reinforced crystalline material.
Microtome sections are not easy to obtain because the glass fibres
may be torn out of the matrix resin while cutting. The polishing tech-
nique is recommended and microscopic inspection is often best
using no filter at all.
With reflecting light and interference contrast, the orientation of
glass fibres can also be observed on samples which have been
printed and highly polished for the microscopic observation.
Once the main orientations and/or agglomerations of the glass
fibres are recognized in the thin polished section prepared from the
critical area of the part, moving the gate and/or changing the mould-
ing parameters (e. g. injection speed) may alter the undesired behav-
iour of the moulded component.

Welding

Today, the welding of parts moulded in engineering thermoplastic

resins is widely used as an assembly technique. Pressure tests can
be used to check the strength of the weld, but the reason for a failure
can only be detected by looking at the cross section of the weld area,
or even better by examining a microtome section. ln particular,
microtome examination can be used to maintain control over weld
time, weld pressure, dimensional precision and material condition.
Sometimes the microscopic examination reveals at the same time
that the moulding conditions have not been at the optimum, but the
mechanical strength of the weld may be within the limits of require-
ments. The evidence of crystallinity not being at its optimum might
indicate which actions to take to improve the quality of the product
further.

22
References

General

1
D.H. Reneker / P.H. Geil: Journal Appl. Phys. 31 (1960),
p. 1916ff.
2
P.H. Geil: C & EN, August 16 (1965), p. 73-84.
3
E.S. Clark: SPE-Journal 23, No.7 (1967), p. 46-49.
4
J.J. Dietl: Kunststoffe 59, H.8 (1969), p. 575-619.
5
E. Bhme: Kunststoffe 60 (1970), p. 273-276.
6
Th. Speil: Plastverarbeiter 22, H.11 (1971), p. 785-792.
7
H.U. lschebeck: Kunststoffe 62, H.2 (1972), p. 110-114.
8
E. Bhme: Kunststoffe-Rundschau, H. 12 (1972).
9
E.S. Clark: Appl. Polymer Symposium No.20 (1973), p. 325ff.
10
Th. Speil: Kunststoffe 63, H.7 (1973), p. 463-468.
11
E.S. Clark: Plastic Engineering, March (1974), p. 73-76.
12
E. Bhme: SPE Techn. Paper 20 (1974), p. 159ff.
13
E. Bhme: Industrie-Anzeiger 97, Jg. 23 (1975).
14
E. Bhme: Kunststoffe Plastics, No. 12 (1982), p. 8-11.

References on Preparation Techniques

a
G. Elssner / G. Kiessler / L. Gessner:
Praktische Metallographie 14, H.9 (1977), p. 445-461.
b
M. Jestdt / R. Henkhaus / J. Hessel / H.E. Bhler:
Praktische Metallographie 16, H.6 (1979), p. 268-283.
c
N. Berndtsen / F.J. Jansen:
Prparation von Polymeren fr die Licht- und
Elektronenmikroskopie. IKV-Aachen (1979).
d
R. Henkhaus / J. Hessel / H.E. Bhler:
Praktische Metallographie 11 (1980), p. 307-320.
e
U. Linke / W.U. Kopp:
Prakt. Metallographie 17, H.10 (1980), p. 479-488.
f
E. Bhme:
Plast-Verarbeiter, H.12 (1982), p. 1464-1466.
g
P. Rohner / J. Brunner:
Textilveredlung 18, H.2 (1983), p. 77-99.
h
F. Jrg:
Textil Praxis International, H.6 (1977), p. 716-719.
i
H.U.lschebeck:
Kunststoffe 74 (1983), p. 153-157.
j
G.W. Ehrenstein / V. Altstdt:
Practical metallography 21 (1984)

23
 Inhomogeneous Melt

Part:
Counter wheels

Resin:
POM

Problem:
Occasional breakage of teeth in long
term test series.

Reason for Failure:

Inhomogeneous melt increased
brittleness and lowered resistance
to impact load on start-up of gear.

A Parts

Microtome Cut

Thin Polished
Section

Thickness 15 m

Microscopic Light

Photo A B C D

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 15 M: 22

B Inhomogeneous melt C Enlargement of B

24
 Inhomogeneous Melt

Part:
Conveyor chain

Resin:
POM

Problem:
Breakage during testing below
specified load limits on some chain
segments.

Reason for Failure:

Insufficient filling of cavity (voids)
and inhomogeneous melt was
reason for premature failure.

Remark:
The thick section should be
redesigned to remove
the tendency to form voids.

A Parts

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A B

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 25

B Void and inhomogeneous melt

25
 Inhomogeneous Melt

Part:
Gear, 45 mm

Resin:
POM, natural

Problem:
Some teeth broke off in test
at extremely low load.

Reason for Failure:

Unmolten particles due to too low
melt temperature reducing strength
of gear teeth.

A Part

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A B

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 20

B Break area

26
 Inhomogeneous Melt

Part:
Socket, overmoulded

Resin:
POM, black

Problem:
Some parts in one delivery had
whitish spots. As the POM was
a blend of 1:1 natural with black
resin, insufficient mixing in moulding
process was assumed.

Reason for Failure:

Microtome cross section prepared
from area of white spots revealed
unmolten particles. These would
cause premature failure.

Remark:
Additional determination of melting
point from inspected area confirmed
that white spots were the same POM
as the homogeneous black resin.

A Part

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A B

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 17

B Section from white spots

27
 Inhomogeneous Melt

Part:
Bobbin for video tape

Resin:
POM

Problem:
ln test series, whitish spots
on surface and no dimensional
stability from shot to shot.

Reason for Failure:

Both problems due to
inhomogeneous melt.

A Part

Microtome Cut x

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 22 M: 22

B Inhomogeneous melt C Homogeneous melt

28
 Inhomogeneous Melt

Part:
Safety clamping part

Resin:
POM, black

Problem:
Parts moulded of black and natural
resin (blend 1:1) gave extremely
low impact values in test. Break area
reveals white spots.

Reason for Failure:

White spots are particles of the
natural resin which due to too Iow
a melt temperature were not
completely molten and thus acted
like a contaminant reducing impact
strength.

Remark:
ln addition, parts are insufficiently
filled resulting in voids.

A Part, break area

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A B

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 33

B Unmolten particles and voids

29
 Inhomogeneous Melt

Part:
Screw

Resin:
POM, black

Problem:
High reject rate in assembly line.
Heads of screws break off.
Cutting screw in haIf reveals void
(see arrow in photo A)
but voids are not at break area.

Reason for Failure:

Inhomogeneous melt and
sharp angle at first thread caused
brittleness.

Remark:
If the microtome cut or polished
section becomes too thin,
recognition of failure in structure
is not possible (see photo D).

A Part

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A B C D

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 10 M: 17

B Head, horizontal C Head, vertical

30
Inhomogeneous Melt

M: 25

D As C, but too thinly polished

31
 Inhomogeneous Melt

Part:
Gear with 3 springs

Resin:
POM, natural and grey

Problem:
After changing from natural to grey
coloured material springs became
brittle.

Reason for Failure:

Inhomogeneous melt in grey parts.

Remark:
Appearance of brittleness is
independant of the change in colour
and purely coincidental.

A Parts

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A B C D E

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 7 M: 7

B Gear, natural good C Gear, grey brittle

32
Inhomogeneous Melt

M:40

D Spring, natural good

M:40

E Spring, grey brittle

33
 Inhomogeneous Melt

Part:
Bushing, overmoulded

Resin:
POM with PTFE-Fibres (20%)

Problem:
Overmoulding cracked under load.

Reason for Failure:

Inhomogeneous melt caused
brittleness.

M: 40

A Inhomogeneous melt

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A B

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 40

B Homogeneous melt

34
 Inhomogeneous Melt

Part:
Refill bottle for lighter

Resin:
PA 66

Problem:
Quality control rejected parts
showing surface defects easily
detectable by naked eye.

Reason for Failure:

Cold thread from nozzle of moulding
machine caused notch effect on
surface (see arrow) emphasized
by cold material underneath.

M: 20

A Thread of cold material near surface

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

35
 Inhomogeneous Melt

Part:
Distance rings

Resin:
PA 66, coloured blue with water
soluble dye

Problem:
No uniform colouring with one
moulding series.

Reason for Failure:

Inhomogeneous melt prevented
uniform absorption of dye.

Remark:
Speed of water absorption is
influenced by size of spherulites.
(Smaller spherulites = slower water
absorption.) Thus non-uniform size
of spherulites caused non-uniform
absorption of water containing
dissolved dye.

A Parts

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A B C D E

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 17 M: 17

B Inhomogeneous melt B Homogeneous melt

36
Inhomogeneous Melt

M: 25

D Non-uniform absorption of dye

M: 25

E Uniform absorption of dye

37
 Inhomogeneous Melt

Part:
Container for break fluid

Resin:
PA 66

Problem:
Brittleness in pressure test.

Reason for Failure:

Notch effect in areas where
homogeneous and
nonhomogeneous melt meet.

Remark:
Visible already on surface
with naked eye.

M: 25

A Surface

Microtome Cut

Thin Polished
Section

Thickness

Microscopic Light

Photo A+B

Transmittant Reflecting
Light

Pol. Filter Darkfield

Pol. Filter +
Red Quartz

M: 25

B Surface

38
 Inhomogeneous Melt

Part:
Extruded rod, 15 mm

Resin:
PA 66, high viscosity resin

Problem:
Brittle breakage occured when
machining the threads.

Reason for Failure:

Unmolten particles in extruded
material caused brittleness.

Remark:
The material giving no problem in
machining showed a homogeneous
structure.

A Rod with brittle break

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 21 M: 21

B Cross section of A C Cross section of good rod

39
 Inhomogeneous Melt

Part:
Ball bearing cage

Resin:
PA 66

Problem:
Series of delivery having varying
performances:
series likely to be distorted
series having strong distortion
so that balls are falling out
good series from earlier delivery

Reason for Failure:

Series of decreasing quality show
decrease in homogeneity of
structure, e. g. melt temperature
was getting lower from one series
to the next.

Remark:
Series with strong distortion show M: 14
in addition weld line due to too low
a melt temperature.

A Parts B Likely to be distorted

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A B C D

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 14 M: 14

C Strong distortion D No distortion

40
 Inhomogeneous Melt

Part:
Lever

Resin:
PA 66, heat stabilized

Problem:
Breakage of parts in use.

Reason for Failure:

Inhomogeneous melt reduced
impact strength and elongation.

A Part

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A B

Transmittant

Pol. Filter

Pol. Filter
+ Red Quartz

M: 25

B Inhomogeneous melt

41
 Inhomogeneous Melt

Part:
Sealing ring

Resin:
PA 66

Problem:
Snap-fit of sealing ring was too soft
and did not hold tight.

Reason for Failure:

In comparison to the rings which
performed satisfactorily the soft
rings had a lower crystallinity
(more transparent appearance).
Thus stiff-ness was reduced.

M: 50

A High crystallinity

Microtome Cut

Thin Polished
Section

Thickness 15 m

Microscopic Light

Photo A B

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 50

B Low crystallinity

42
 Inhomogeneous Melt

Part:
Standard impact bar
(Charpy 50 6 4 mm)

Resin:
PA 66, modified

Problem:
High variation in impact strength on
standard bars moulded at customer.

Reason for Failure:

Poor melt quality and low melt
temperature caused irregular impact
resistance.

Remark:
As proof, impact bars were moulded
with correct and too low melt
temperature. Failure could be
reproduced. Moulding parameters
other than cold melt are difficult
to recognize in the structure of M: 20
modified PA.

A Correct melt

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A B

Transmittant

Pol. Filter

Pol. + Red Quartz

M: 20

B Inhomogeneous melt

43
 Inhomogeneous Melt

Part:
Cable strap, different types

Resin:
PA 66, natural and black

Problem:
Brittle break in bending test or
in assembly.

Reason for Failure:

Inhomogeneous melt. In some cases
sharp corner in tooth base. One has
high shear orientation around the
base of the teeth. This shear might
be caused by too low melt
temperature (D).

Remark:
Sharp corners in the base of the teeth
cause stress concentration and notch
effect resulting in premature break
under load.

A Part

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A B to J

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 21 M: 28

B Inhomogeneous melt C Homogeneous melt

44
Inhomogeneous Melt

M: 17 M: 55

D Shear along teeth E Inhomogeneous melt

M: 9

F Inhomogeneous melt

45
 Inhomogeneous Melt

M: 7 M: 7

G Inhomogeneous melt H Homogeneous melt

M: 22 M: 22

I Enlarged section of G J Enlarged section of H

46
 Inhomogeneous Melt

Part:
Standard specimens (tensile bars)

Resin:
PA 66 with mineral (40%)

Problem:
Evaluation of possibility to recognize
low or inhomogeneous melt in
66-nylon containing mineral.

Reason for Failure:

Standard tensile bars (3,2 mm
thickness) moulded with correct
and 20 C lower melt temperature
reveal evidence of cold melt in
thin section.

M: 25

A Correct melt temperature

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A B

Transmittant

Pol. Filter 30

Pol. Filter +
Red Quartz

M: 20

B Inhomogeneous melt

47
 Weld Line

Part:
Gear, 4 mm , 6 mm height

Resin:
POM

Problem:
Some teeth break off in test run
(gear has 2 injection points).

Reason for Failure:

Weld line due to too cold melt
temperature when melt flows meet.
In some gears the weld line is located
at a tooth base angle (D).

A Part B Thinly polished gears

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A B C D

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 9 M: 11

C Weld line between two teeth D Weld line at tooth base angle

48
 Weld Line

Part:
Trim clip

Resin:
POM

Problem:
Parts break during assembly.
Break area almost opposite to gate.

Reason for Failure:

Weld line formed on surface only due
to too fast cooling at mould surface
when melt flows around the core.
Thus notch effect.

Remark:
Weld line does not appear
throughout the whole cross section.
Therefore notch effect from
surface only when stressed during
assembling (see area between
arrows on photo D).

A Trim clip B Weld line marked

with soft pencil

Microtome Cut

Thin Polished for D

Section

Thickness 10 m

Microscopic Light

Photo A B C D

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 7 M: 7

C Break at marked area D Cross section at marked area

49
 Weld Line

Part:
Functional part

Resin:
POM

Problem:
Part of the production gave breakage
below threshold limit in quality
control testing.

Reason for Failure:

In comparison to parts within limits
of test specification, the failing parts
show a distinct weld line in cross
section.

M: 9

A Weld line

Microtome Cut

Thin Polished x
Section

Thickness 20 m

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 44 M: 44

B Enlarged area of weld line C Same cross section as B from

good part

50
 Weld Line

Part:
Bushing

Resin:
POM

Problem:
Bushings crack opposite to gate
when forced onto axle.

Reason for Failure:

Melt has been cooled off too much
while flowing in cavity around
the core preventing a homogeneous
weld. Reason can be too low melt
temperature or too slow injection
speed.

M: 25

A Weld line

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

51
 Weld Line

Part:
Plate with threaded boss

Resin:
POM/POM, modified (blend)

Problem:
Part cracked at weld line when stud
was screwed in. Notch effect on
surface (see arrow).

Reason for Failure:

Weld line in structure goes through
the whole cross section. Weld line
appears when melt is already too
cold and material flow meets behind
the core.

A Part showing crack

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 17 M: 44

B Weld line in cross section C Enlarged area of B

52
 Weld Line

Part:
Window handle

Resin:
POM, coloured with masterbatch

Problem:
Parts break in test.

Reason for Failure:

Position of gate in relation to load
direction not optimum. In addition,
low melt temperature caused weld
line.

Remark:
The low melt temperature also
prevented homogeneous dispersion
of masterbatch.

A Parts

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 7 M: 17

B Weld line and inhomogeneous

melt C Enlarged, weld line

53
 Weld Line

Part:
Bushing

Resin:
PA 66, high viscosity resin

Problem:
Breakage at loads under the specified
limits. Breaks occured opposite gate.

Reason for Failure:

Weld line clearly visible in thin
section due to too low melt
temperature.

M: 10

A Low melt temperature = weld line

Microtome Cut

Thin Polished
Section

Thickness 15 m

Microscopic Light

Photo A B

Transmittant

Pol. Filter

Pol. + Red Quartz

M: 10

B Homogeneous melt

54
 Weld Line

Part:
Bushing

Resin:
PA 66, unreinforced

Problem:
High reject rate in pressure test.
Position of break seemed to be
at weld line.

Reason for Failure:

Two reasons, both due to too low
melt temperature.
a) weld line (in S-shape)
b) partially finer structure similar to
nucleated PA. The latter has lower
elongation in dry-as moulded
condition than normal structure

M:40

A Weld line

Microtome Cut
perpen-
dicular
to flow

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

55
 Weld Line

Part:
Radiator tanks

Resin:
PA 66, 30% glass reinforced

Problem:
Weld line opposite to gate was
visible to naked eye. Quality of
weld checked by mechanical tests
(Dynstat impact strength) and thin
polished section.

Reason for Failure:

No failure observed. Although weld
line is visible in thin polished section,
no loss in impact strength compared
to other areas of same part.

Remark:
S-shape of weld line in glass
reinforced PA seems to give
good mechanical strength.
Best observation is without any filter.

A Part

Microtome Cut

Thin Polished
Section

Thickness 15 m

Microscopic Light

Photo A B C D

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M:12 M:12

B Weld line C Weld line

56
Weld Line

M:17

D Weldline

57
 Weld Line

Part:
Spoke wheel

Resin:
PA 66, modified, 15% glassfiber
content

Problem:
Control of weld line between
two gates. Break in test at weld line.

Reason for Failure:

Presence of weld line can not be
eliminate. In this case impact tests
confirmed sufficient mechanical
strength.

A Wheels

Microtome Cut

Thin Polished
Section

Thickness 30 m

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M:28

B Weld line C Weld line Schema

58
 Mould Temperature

Part:
Spring

Resin:
POM

Problem:
In long term testing a few springs
showed less fatigue resistance than
other ones.

Reason for Failure:

Too low mould temperature of failing
springs, as seen by comparison of
cross sections, resulted in a more
flexible part explaining differences.

Remark:
Systematic studies on effect of
mould temperature on mechanical
properties of POM-homopolymer
have explained different behaviour
(see literature). M: 10

A Part

Microtome Cut

Thin Polished
Section

Thickness 15 m

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 35 M: 35

B Low mould temperature C Correct mould temperature

59
 Mould Temperature

Part:
Spring in light switch

Resin:
POM

Problem:
Prototype machined from extruded
rod performed 180000 cycles without
fatigue. Same part moulded resisted
in test 80000 cycles only.

Reason for Failure:

Machined part consisted of highly
crystalline material with big
spherulites without preferred
orientation. In comparison the
moulded part shows 3 different
zones of orientation of the
spherulites resulting in lower
stiffness/higher elasticity thus less
fatigue resistance.
M: 10
Remark:
See literature ref. 5.

A Part

Microtome Cut

Thin Polished
Section

Thickness 15 m

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. + Red Quartz

M: 11 M: 11

B Machined prototype C Moulded prototype

60
 Screw Forward Time

Part:
Spring

Resin:
POM

Problem:
In test series some of the parts had
less fatigue resistance than other
ones.

Reason for Failure:

Structure shows line of too short SFT,
reducing mechanical properties.

M: 10

A Part

Microtome Cut

Thin Polished
Section

Thickness 15 m

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 35 M: 35

B Cross section of good part C Cross section showing SFTline

61
 Voids

Part:
Zipper

Resin:
POM

Problem:
Teeth break off.

Reason for Failure:

Insufficient filling of mould cavity
resulting in voids. Thus, reduced
cross section gives lower mechanical
strength.

M: 17

A Zipper B Cross section

Microtome Cut

Polished Section Reflectant

light
Darkfield

Thickness

Microscopic Light

Photo A B C D

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 17 M: 28

C Section lengthwise D Section lengthwise

62
 Shear Orientation

Part:
Snap ring for eyepiece of binocular

Resin:
POM, black

Problem:
Ring breaks at snap fit area (100%
reject).

Reason for Failure:

Strong shear orientation at gate
(photo A) reduced elongation of ring.
Slower injection speed reduced shear
orientation considerably to produce
acceptable function of the parts.

Remark:
Failing parts were also not well filled
showing big voids in center. Thus the
functional cross section was reduced
too.
M: 13

A Strong shear at gate

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A B

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 13

B Reduced shear at gate

63
 Shear Orientation

Part:
Seal

Resin:
PA 66

Problem:
No problem, but control of desired
break area designed on purpose to
act as safety valve.

Reason for Failure:

Shear orientation is clearly seen at
corners of thin wall cross section.
The reduced strength perpendicular
to the orientation results in breakage
in the desired area.

M: 80

A Shear orientation

Microtome Cut

Thin Polished
Section

Thickness 10 m

Microscopic Light

Photo A

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

64
 Shear Orientation

Part:
Bobbin

Resin:
PA 66

Problem:
Bobbin heads break in assembly.

Reason for Failure:

High shear at gate located at head
of bobbin. In addition sharp corner
at shoulder (see arrow).

Remark:
Sharp corners increase brittleness
even with parts of conditioned nylon.

M: 25

A Shear at gate

Microtome Cut

Thin Polished
Section

Thickness 15 m

Microscopic Light

Photo A

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

65
 Shear Orientation

Part:
Snap-hinge

Resin:
PA 66, modified, black

Problem:
Break in hinge after few flexions.

Reason for Failure:

Too high shear orientation due to too
low melt temperature.

Remark:
Additional notch effect in corner of
flexioned area caused by damaged
surface of mould (see arrow in
photo B).

M: 13

A Shear and inhomogeneous melt

Microtome Cut

Thin Polished
Section

Thickness 15 m

Microscopic Light

Photo A B

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 32

B Enlargement of A

66
 Shear Orientation

Part:
Connector hinge

Resin:
PA 66, modified

Problem:
After one flexion when closing hinge,
breakage occured. High reject rate
on assembly line.

Reason for Failure:

The shear orientation inevitable
in hinges was increased due to too
low or inhomogeneous melt
temperature.

Remark:
Even without presence of shear
orientation cold or inhomogeneous
melt causes brittleness of parts.

M: 14

A Hinge

Microtome Cut

Thin Polished
Section

Thickness 10 m

Microscopic Light

Photo A

Transmittant

Pol. Filter

Pol. Filter +
Red Quarz

67
 Pigment
Agglomeration

Part:
Window handles

Resin:
POM, coloured with masterbatch

Problem:
Part had been moulded from natural
resin coloured by adding master-
batch. Parts showed coloured streaks
on surface.

Reason for Failure:

Inhomogeneous melt or too low
melt temperature prevented
uniform distribution of the colour
masterbatch.

A Part

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A B

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 25

B Cross section at area with streaks

68
 Pigment
Agglomeration

Part:
Cover

Resin:
POM, coloured with red masterbatch

Problem:
Spots of darker red visible on
surface. Rejected by quality control.

Reason for Failure:

Microtome cut through area with
dark spot revealed agglomeration of
masterbatch. The high concentration
of red batch caused darker spots
of red on surface.

Remark:
Insufficiently dispersed masterbatch
may indicate inhomogeneous melt.

M: 50

A Particle of masterbatch

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

69
 Pigment
Agglomeration

Part:
Numbering wheel

Resin:
POM

Problem:
Material had been coloured black
with dry pigment. Gears were brittle
in use.

Reason for Failure:

Colouring with dry pigment caused
pigment agglomeration which acts
like foreign particles causing
premature breakage.

M: 80

A Breakage area with pigment concentration

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A

Transmittant

Pol. Filter

Pol. Filter +
Red. Quartz

70
 Pigment
Agglomeration

Part:
Extruded rod

Resin:
PA 66, modified, coloured with
masterbatch

Problem:
No uniform colour was obtained
with red masterbatch. Red streaks
on surface visible.

Reason for Failure:

Inhomogeneous melt prevented
uniform dispersion of masterbatch.

M: 25

A Section with streaks

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

71
 Notch Effect due to
flashing/
sharp corners

Part:
Cam shaft gear

Resin:
POM

Problem:
Breakage of teeth.

Reason for Failure:

Flash in tooth base due to damaged
mould caused notch effect resulting
in premature breakage.

M: 65

A Notch due to flash

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

72
 Notch Effect due to
flashing/
sharp corners

Part:
Body with ribs

Resin:
POM

Problem:
Surface cracks at ribs in some parts
of the delivered series.

Reason for Failure:

Cracks caused by notch effect from
sharp corners in tool. High stress
orientation at corners in comparison
to non-failing parts.

M: 50

A Stress orientation at sharp corners

Microtome Cut

Thin Polished
Section

Thickness 15 m

Microscopic Light

Photo A B

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 50

B Corners well rounded off

73
 Notch Effect
due to flashing

Part:
Ball bearing cage

Resin:
POM

Problem:
Trial to evaluate overmoulding
causing flashing on parts.

Reason for Failure:

Flash resulted in notch effect during
assembly.

A Part

Microtome Cut

Thin Polished
Section

Thickness

Microscopic Light

Photo A B

Transmittant

Pol. Filter

Pol. +
Red Quartz

M: 35

B Flash

74
 Notch Effect
due to flashing

Part:
Distance rings

Resin:
PA 66, dyed after moulding

Problem:
Rings crack when mounted
on an axle.

Reason for Failure:

The thin flash cracked under load
propagating a notch effect causing
the breakage of the ring.

Remark:
Insufficient clamping force of
moulding machine could be the
reason for the pronounced flash.

A Part

Microtome Cut

Thin Polished
Section

Thickness

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 28 M: 28

B Flash C Broken flash causing notch effect

75
Glass fibre Orientation

Part:
Gear

Resin:
PA 66, 30% glass fibre content

Problem:
No reject. Study of glass fibre
orientation in teeth of gear.

Reason for Failure:

Glass fibres follow the shape of
the teeth well. No turbulences
or glass fibre agglomeration can
be seen.

M: 12

A Teeth of gear

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 22 M: 17

B Tooth C Between two teeth

76
Glass fibre Orientation

Part:
Connector

Resin:
PETP, 30 % glass fibre content

Problem:
Dimensional differences of holes
within one connector.

Reason for Failure:

By recognizing existing glass fibre
orientation, slight changes in
moulding conditions reduced
differences in dimension of holes
within part.

A Parts

Microtome Cut

Thin Polished
Section

Thickness 10 m

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 7 M: 22

B Fibre orientation C Enlargement of B

77
 Welding

Part:
Ball, spin welded
(part for demonstration)

Resin:
POM

Problem:
Two sphere halves are spin welded.

Remark:
Good weld giving resistance to high
burst pressure.

M: 32

A Good spin weld

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

78
 Welding

Part:
Several spin welded parts

Resin:
POM

Problem:
Part not gas-tight in welded area.

Reason for Failure:

Area of spin weld not optimally
designed. Too sharp angle which
cannot be filled and thus did not
weld.

A Parts, spin welded

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 17 M: 17

B Weld area not gas-tight C Good spin weld

79
 Welding

Part:
Different vibration welded parts

Resin:
POM, PA 66

Problem:
Quality control of welded area.

Remark:
All welded areas are good.

A Parts, vibration welded

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 22.5 M: 22.5

B Good weld POM C Good weld PA 66/PA 66 GR

80
 Welding

Part:
Fuel filter, spin welded

Resin:
PA 66

Problem:
Weld has sufficient mechanical
strength but some are not gas-tight.

Reason for Failure:

Poor quality of weld preventing good
bond between the two halves.

Remark:
Parts which are considered to be
of sufficient mechanical strength
do not have a perfect weld either.
Furthermore, melt temperature
of the moulded parts has been too
low and inhomogeneous.

A Part, spin welded

Microtome Cut

Thin Polished
Section

Thickness 25 m

Microscopic Light

Photo A B C

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 35 M: 35

B Not gas-tight but mechanically

acceptable C Not gas-tight and insufficient
mechanical strength

81
 Welding

Part:
Box, ultrasonically welded
(part for demonstration)

Resin:
POM, PA 66

Problem:
POM = insufficient weld strength
PA 66 = no welding effect

Reason for Failure:

POM
photo B: too long weld time
photo C: bad dimensional fitting
photo D: good weld
photo E: good weld, different part
PA 66
photo F: material contains too much
moisture
photo G: good weld, dry material

A Part

Microtome Cut

Thin Polished
Section

Thickness 20 m

Microscopic Light

Photo A B to G

Transmittant

Pol. Filter

Pol. Filter +
Red Quartz

M: 22 M: 22

B Weld time too long POM C Bad fitting POM

82
Welding

M: 22 M: 17

D Good weld POM E Good weld POM, different part

M: 22 M: 22

F Material moist PA 66 G Good weld PA 66

83
 We believe this information is the best currently available on the subject. It is
offered as a possible helpful suggestion in experimentation you may care to under-
take along these lines. It is subject to revision as additional knowledge and expe-
rience are gained. Du Pont makes no guarantee of results and assumes no
obligation or liability whatsoever in connection with this information. This publi-
cation is not a licence to operate under, or intended to suggest infringement of any
existing patents.

Stmpfli-Design: Peter Sennhauser

Printed by: Stmpfli + Co Ltd., Berne, Switzerland
Belgique / Belgi Isral Schweiz / Suisse / Svizzera Brasil
Du Pont de Nemours (Belgium) Gadot Chemical Terminals (1985) Ltd. Biesterfeld Plastic Suisse GmbH Du Pont do Brasil S.A.
Antoon Spinoystraat 6 16 Habonim Street Dufourstrasse 21 Al. Itapecuru, 506 Alphaville
B-2800 Mechelen Netanya South Ind. Zone Postfach 14695 06454-080 Barueri-So Paulo
Tel. (15) 44 14 11 IL-42504 Netanya CH-4010 Basel Tel. (5511) 7266 8229
Telex 22 554 Tel. (3) 526 42 41 Tel. +41 (0) 61 201 31 50
Telefax (15) 44 14 09 Telefax (3) 528 27 17 Telefax +41 (0) 61 201 31 69 Asia Pacific
Du Pont Kabushiki Kaisha
Bulgaria Italia Slovenija Arco Tower
Serviced by Biesterfeld Interowa GmbH Du Pont de Nemours Italiana S.r.L. Serviced by Biesterfeld Interowa GmbH 8-1, Shimomeguro 1-chome
& Co. KG. Via Volta, 16 & Co. KG. Meguro-ku, Tokyo 153-0064
See under sterreich. I-20093 Cologno Monzese See under sterreich. Tel. (03) 5434-6935
Tel. (02) 25 30 21 Telefax (03) 5434-6965
Cesk Republika a Slovensk Telefax (02) 25 30 23 06 Suomi / Finland
Republika Du Pont Suomi Oy South Africa
Du Pont CZ, s.r.o. Magyarorszg Box 62 Plastamid (Pty) Ltd.
Pekarska 14/268 Serviced by Biesterfeld Interowa GmbH FIN-02131 Espoo 43 Coleman Street
CZ-155 00 Praha 5 Jinonice & Co. KG. Tel. (9) 72 56 61 00 P.O. Box 59
Tel. (2) 57 41 41 11 See under sterreich. Telefax (9) 72 56 61 66 Elsies River 7480
Telefax (2) 57 41 41 50-51 Cape Town
Maroc Sverige Tel. (21) 592 12 00
Danmark Deborel Maroc S.A. Serviced by Du Pont Danmark ApS. Telefax (21) 592 14 09
Du Pont Danmark ApS 40, boulevard dAnfa 10 See under Danmark.
Skjtevej 26 MA-Casablanca USA
P.O. Box 3000 Tel. (2) 27 48 75 Trkiye DuPont Engineering Polymers
DK-2770 Kastrup Telefax (2) 26 54 34 Du Pont Products S.A. Barley Mill Plaza, Building #22
Tel. 32 47 98 00 Buyukdere Caddesi No. 122 P.O. Box 80022
Telefax 32 47 98 05 Norway / Norge Ozsezen Ismerkezi, A block, Kat: 3 Wilmington, Delaware 19880
Distrupol Nordic Esentepe, 34394 Istanbul, Turkey Tel. (302) 999 45 92
Deutschland Ostenssjoveien 36 Tel. +90 212 340 0400 Telefax (302) 892 07 37
Du Pont de Nemours N-0677 Olso Telefax +90 212 340 0430
(Deutschland) GmbH Tel. 23 16 80 62
DuPont Strae 1 Telefax 23 16 80 62 Ukraine
D-61343 Bad Homburg Du Pont de Nemours International S.A.
Tel. (06172) 87 0 sterreich Representative Office
Telex 410 676 DPD D Biesterfeld Interowa GmbH & Co. KG 3, Glazunova Street
Telefax (06172) 87 27 01 Bruhausgasse 3-5 Kyiv 252042
P.O. Box 19 Tel. (044) 294 96 33 / 269 13 02
Egypt AT-1051 Wien Telefax (044) 269 11 81
Du Pont Products S.A. Tel. (01) 512 35 71-0
Bldg no. 6, Land #7, Block 1 Fax (01) 512 35 71-31 United Kingdom
New Maadi e-mail: [email protected] Du Pont (U.K.) Limited
ET-Cairo internet: www.interowa.at Maylands Avenue
Tel. (00202) 754 65 80 GB-Hemel Hempstead Requests for further information
Telefax (00202) 516 87 81 Polska Herts. HP2 7DP from countries not listed above
Du Pont Poland Sp. z o.o. Tel. (01442) 34 65 00 should be sent to:
Espaa ul. Powazkowska 44C Telefax (01442) 24 94 63
Du Pont Ibrica S.A. PL-01-797 Warsaw Du Pont de Nemours
Edificio LIlla Tel. +48 22 320 0900 Argentina International S.A.
Avda. Diagonal 561 Telefax +48 22 320 0910 Du Pont Argentina S.A. 2, chemin du Pavillon
E-08029 Barcelona Avda. Mitre y Calle 5 CH-1218 Le Grand-Saconnex/Geneva
Tel. (3) 227 60 00 Portugal (1884) Berazategui-Bs.As. Tel. (022) 717 51 11
Telefax (3) 227 62 00 Biesterfeld Iberica S.L. Tel. +54-11-4229-3468 Telex 415 777 DUP CH
Rua das Matas Telefax +54-11-4229-3117 Telefax (022) 717 52 00
France P-4445-135 Alfena
Du Pont de Nemours (France) S.A. Tel. 229 698 760
137, rue de lUniversit Telefax 229 698 769
F-75334 Paris Cedex 07
The information provided in this documentation corresponds to our knowledge on the subject at
Tel. 01 45 50 65 50 Romania the date of its publication. This information may be subject to revision as new knowledge and
Telex 206 772 dupon Serviced by Biesterfeld Interowa GmbH experience becomes available. The data provided fall within the normal range of product prop-
Telefax 01 47 53 09 67 & Co. KG. erties and relate only to the specific material designated; these data may not be valid for such
See under sterreich. material used in combination with any other materials or additives or in any process, unless
expressly indicated otherwise. The data provided should not be used to establish specification
Hellas limits nor used alone as the basis of design; they are not intended to substitute for any testing
Biesterfeld Hellas Intralink S.A. Russia you may need to conduct to determine for yourself the suitability of a specific material for your
Trading Establishment OOO DUPONT RUSSIA particular purposes. Since DuPont cannot anticipate all variations in actual end-use conditions
149, AG. Triados Menidi Acharnes B. Palashevsky pereulok 13, str. 2 DuPont makes no warranties and assumes no liability in connection with any use of this infor-
mation. Nothing in this publication is to be considered as a license to operate under or a
GR-13671 Athens 123104 Moscow, Russia recommendation to infringe any patent rights.
Tel. (210) 24 02 900 Tel. +7 095 7972200 Caution: Do not use this product in medical applications involving permanent implantation in
Telefax (210) 24 02 141 Telefax +7 095 7972201 the human body. For other medical applications see DuPont Medical Caution Statement .

The DuPont Oval, DuPont, The miracles of science, DELRIN, ZYTEL are registered trademarks
or trademarks of DuPont or its affiliates