C A L C U L AT I O N S D E S I G N A P P L I C AT I O N S C . 4 .

Failure Analysis by means of Microscopy

Ticona A Business of Celanese AG

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

Contents
1. 2. 2.1 2.2 3. 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12 3.13 3.14 3.15 3.16 3.17 4. 4.1 4.1.1 4.1.2 5. 5.1 5.1.1 5.1.2 5.1.3 5.1.4 5.2 5.2.1 Introduction Methodology of microscopic failure analysis Structured approach to failure analysis Microscopic analysis Application examples Trapped gas in the melt Cold slug Previous damage to the molded part Contamination by extraneous materials Cold flow Design faults Processing temperature too low Stress cracking Carbon black agglomerates Inhomogeneous color masterbatch dispersion Soft phase separation Component damage as a result of external effects Different microstructure Inhomogeneous melt Insufficient holding pressure time Cracking in the molded part Weld line too cold Principles of microscopy Principle of birefringence Refractive index Optical path difference Microscopy methods Transmitted light microscopy Bright field method in transmitted light Phase contrast method Dark field method in transmitted light Polarization method in transmitted light Reflected light microscopy Differential interference contrast method in reflected light Sample preparation Literature 3 4 6 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 29 29 29 32 32 32 33 34 34 35 36 37 38

6. 7.

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

Introduction

Methodology of microscopic failure analysis

Application examples

Principles of microscopy

Microscopy methods

Sample preparation

Literature
1

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

1. Introduction
Engineering plastics are now used to produce a diverse variety of components for virtually all sectors of industry, engineering and the home. Their economical manufacturing processes make them outstanding materials for producing high-volume components of complex design. Components such as housings, connectors and functional elements can be precision molded, are able to withstand high mechanical stresses and often play a key role in complete systems. By incorporating fillers, reinforcing materials and other additives, polymers can be varied over a wide range of properties and used in special technical applications. In this context, microscopy has become established as an important method for quality control and failure analysis. Molded part quality is largely determined by the following factors: material properties molded part design mold design processing The formation of a homogeneous structure in molded parts is particularly important. To determine the source of a defect, several analytical methods of varying complexity are available. For cost reasons, a rapid damage assessment is usually required. Costly and time-consuming analytical methods are therefore frequently ruled out. In this context, light microscopy has become established as a suitable and informative method for defect evaluation in plastics processing and is an excellent tool for rapid failure analysis. This brochure provides typical examples of possible microscopic failure analysis applications and explains how to interpret results (see Section 3). The examples relate mainly to the following engineering polymers from Ticona: Hostaform Celanex GUR POM PBT PE-UHMW

The complex interactions between processing, design and material parameters and external stress conditions influence physical properties, chemical structure and therefore the molded part itself (see Fig. 1). These interactions can create defects.

Fig. 1: Factors that influence the properties of a molded part [1]

Processing

Composition Chemical structure Morphology Physical structure

Aging

Design

External stress

Properties of the finished part

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

2. Methodology of microscopic damage analysis

Physical properties listed in material brochures are determined on standard test specimens and often differ from actual conditions obtained for a given component. In comparison with tensile, notched impact and flexural test specimens produced under optimum processing conditions, industrially molded parts often have different morphologies because of different processing and flow conditions. This results in different property profiles and sometimes in failure of the part. In failure analysis, failure criteria are determined and a failure hypothesis is proposed. Microscopic analysis can provide information on the following effects: External influences and contamination, unmelted material particles, cold slugs Dispersion of fillers, reinforcing materials and pigments Voids, flaws, weld lines Molecular and reinforcing material orientation Internal stress conditions as a result of internal and external forces (and deformation resulting from the failure process) Crazing and spherulitic structures Faults in mold construction and component design, e.g. missing radii, differences in radii at corners and edges, differences in wall thickness, notch effect due to ejector marks Methods available for microscopic studies include light microscopy, scanning electron microscopy (SEM), transmission electron microscopy (TEM) and scanning force microscopy. Decisive factors in selecting the most appropriate microscopy method include both the morphological details required and the resolution of the microscope. Fig. 2 shows the resolutions achievable with various optical systems and examples of recognizable polymer structures. In most cases, if rapid failure analysis is required, light microscopy offers a fast and conclusive method of identifying defects in materials and molded parts.
crystalline structures 10 10 100 pm multi-phase structures

Fig. 2: Resolutions achievable with various optical systems [2]

Resolution m pellets 10 3 1 mm human eye powder particles 10 4 100 m

10 5

10 m

spherulites

2
10 6 1 m light microscope

10 7

100 nm

10 8 lamellae microcrystals

10 nm

scanning electron microscope

10 9

1 nm transmission electron microscope

Electron microscopy and scanning force microscopy, on the other hand, generally require very costly and time-consuming sample preparation. Many factors influence the reliability of microscopic studies, which also often depend on parallel investigations using other analytical methods. For this reason, a structured approach to failure analysis, as described below, is of crucial importance. 2.1 Structured approach to failure analysis To carry out failure analysis, the following should be available: the pellets (virgin material) from which the molded part was produced an unused or undamaged molded part made from the same material batch as the failed part the failed part (defective molding).

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

By comparing the pellets with the undamaged part, the influence of processing can be determined; by comparing the failed part with the undamaged part, the influence of service conditions can be deduced; from the damage picture and the molded part geometry, design shortcomings can be identified. If only the failed part is available, the failure analysis is more difficult and considerably less reliable. To determine the causes of failure reliably and unambiguously, the failure analysis must be carefully planned. A flow chart for failure analysis of plastic components is shown in Fig. 3. The individual steps in this failure investigation program are described in more detail below.

Failure description In the failure description step, the failure is photographically recorded and any distinctive features that can be verified through macroscopic examination by stereomicroscopy or the human eye are documented. Pictorial documentation of the failure is very important, since the molded parts frequently have to be destroyed in the subsequent analyses. Important points to note in the failure description are listed below: Type of failure fracture, cracking, deformation, corrosion, fatigue, aging or failure as a result of temperature effects Characterization of the type of failure brittle or ductile fracture, degree of deformation, depth of corrosive attack Association with geometric shape failure features associated with corners, edges, holes, ribs and radii, weld lines Processing indications weld lines, grooves, surface treatment, thermal effects, melting structures Color and shape of the part change in color compared with the original material Surface structures mold grooves, discoloration, thermal effects, sink marks, deposits, inhomogeneities, etc. Clear evidence of fault-inducing features sharp-edged transitions, crack-initiating notches, voids, extraneous contamination, agglomerates When the failure has been described in detail, a status assessment can be prepared. Status assessment After the macroscopic examination, the background information relating to the failure should now be determined. This includes: Material exact grade, blend components, additives, processor

Fig. 3: Flow chart for failure analysis [1]

Failure event
(Primarv failure)

Failure description
(macroscopic failure description, documentation)

Situation assessment
(function of part, required condition, failure sequence)

Preparation of investigation program Material investigation

Analysis microscopy (LM, SEM, TEM, AFR) thermal analysis (DSC, TMA, DMA) IR spectroscopy chemical analysis Testing static dynamic ESC ultrasonic Simulation practical study laboratory study calculation

Evaluation of investigation results

(actual condition, comparison between actual and required condition)

Conclusion
(determination of cause of failure corrective measures)

Documentation
(failure report, database)

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

Processing processing parameters, machine, storage, post-processing Function of the part effects deriving from this function Application Was the component in use? Environmental influences such as temperature, radiation, media, etc. Points of special note related to the failure Since this is a comprehensive list, it must always be assumed that some points will remain uncertain and others will only serve to round out the complete picture. When these questions have been answered, a failure hypothesis can be proposed and an investigation program can be worked out. Material investigation When the status assessment has been completed and a failure hypothesis prepared, a material investigation must be carried out to confirm the hypothesis. There are three main avenues of approach here: Analytical studies, material testing and simulation. Generally speaking, the focus in material investigation is on material analysis. If required to support the measured analytical values, results from testing and simulation can also be used. The selection of analytical studies is based on the failure hypothesis and on the resources and/or investment required. The simpler the better is the guiding principle. To facilitate this process, the questions to be answered by the studies should be formulated in advance. Evaluation of the investigation results Once results are available, comparison of the theoretical and actual conditions will either reveal the source of the failure or else it will point up the need for further investigations and thereby necessitate continuation of the failure analysis program. The point at which the difference between the actual and theoretical condition becomes significant depends on the type of failure and can only be assessed with requisite expertise and experience.

Conclusion Using logical connections between the failure description, status assessment and investigation results, one or more causes of failure can often be deduced. The causes of failure should then be divided into primary and secondary causes. Corrective measures then follow directly from the causes of failure. Failure report In failure analysis, it is important to ensure structured documentation of failures. Moreover, a generally accessible database in which all failure documentation is sorted according to failure characteristics can be very useful. By facilitating database searches, this offers advantages in terms of utilizing experience from previous failures thereby reducing the amount of work required to perform an investigation preparing reports efficiently cutting analysis costs Structured documentation takes the form of a failure report, which should have the following layout: 1.) Background Customer Identification of failed part Reason for failure investigation Nature and scope of failure 2.) Description and investigation Method Test parameters 3.) Results Of the status assessment Of the individual investigation 4.) Conclusion Cause of failure Recommendations Following this description of the structured approach, detailed consideration will now be given to the light microscopic methods used in failure analysis, which feature prominently in this brochure.

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

Special effects can be made clear by the microscopic methods of polarization, differential interference contrast, dark field, crystal kinetics measurement (hot stage microscope) and image analysis. 2.2 Microscopic analysis Through appropriate sample preparation, microscopy reveals internal structures in the molded part due to differences in the reflection and absorption characteristics of material structures. If necessary, when the

sample is too transmissive, the required contrast can often be obtained by optimizing the lighting system. This is achieved by light polarization or phase shifting as with the phase contrast method or the differential interference contrast method. The dark field method is based on light scattering effects from objects contained in the polymer sample. Apart from the phase contrast method, which can only be used with transmitted light, all other microscopic methods can also be used with reflected light.

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

3. Application examples
Typical failure analysis pictures analyzed by various microscopic methods are shown in this section. The pictures correspond to the following examples: 3.1 Trapped gas in the melt Problem: White streaks on the molded part surface Material: POM Cause of problem: Trapped gas in the melt Note: The white streaks are caused by voids just under the surface that cause the light to scatter differently there than in surrounding areas.

3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12 3.13 3.14 3.15 3.16 3.17

Trapped gas in the melt Cold slug Previous damage to the molded part Contamination by extraneous materials Cold flow Design faults Processing temperature too low Stress cracking Carbon black agglomerates Inhomogeneous color masterbatch dispersion Soft phase separation Component damage from external effects Different microstructure Inhomogeneous melt Insufficient holding pressure time Cracking in the molded part Weld line too cold

Molded part surface in reflected light

Section through the surface defect using differential interference contrast

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

3.2 Cold slug Problem: Failure of molded part in service Material: POM Cause of problem: Cold slug Note: The cold slug was caused by unclean separation of the cold runner from the part.

3.3 Previous damage to the molded part Problem: Failure of the molded part in service Material: POM Cause of problem: Previous damage to the molded part by the ejector during removal of the part from the mold Note: Classic dimensional accuracy problem in conjunction with sharp-angled edges.

Fracture surface in reflected light Thin section taken here

Damage point in reflected light Thin section taken here

Thin section through the cold slug in polarized light

Thin section through the damage point in polarized light

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

3.4 Contamination by extraneous materials Problem: Fracture in the molded part Material: POM Cause of problem: Contamination by extraneous materials Note: The molded part failed because of extensive contamination by an amorphous polymer.

3.5 Cold displacement Problem: Fracture in an integral hinge Material: POM Cause of problem: Cold flow Note: The molded part failed because of cold flow caused by melt flow stagnation ahead of the hinge during mold filling.

Fracture surface in reflected light

Thin section in polarized light

Inhomogeneities

Critical areas

Thin section in polarized light

Thin section in polarized light Section enlargement

Stress cracks

Foreign particles
9

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

3.6 Design faults Problem: In-service fracture of a molded pin Material: POM Cause of problem: Sharp-angled corner in conjunction with stress concentration Note: The molded part failed because of the sharp-angled corner (notch effect) in conjunction with stress concentration in this area. Thin section of the defective sample part in polarized light Fracture surface

3.7 Processing temperature too low Problem: In-service failure of molded part Material: POM Cause of problem: Premature freezing of the sprue due to low processing temperature Note: The sprue froze prematurely because the processing temperature was too low. This resulted in insufficient holding pressure, which created a sheath-core structure. Thin section of the sprue in polarized light

Thin section of the control sample part in polarized light, mirror-inverted sharp-angled transitions

Thin section from the fracture zone in polarized light sheath-core structure

stress concentration

10

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

3.8 Stress cracking Problem: Stress cracking Material: POM Cause of problem: Stress cracking as a result of contact with biodiesel

3.9 Agglomerates Problem: Failure of the molded part during assembly Material: PBT Cause of problem: Failure due to carbon black agglomerate in the snapfit hook Note: Given that carbon black is hygroscopic, an additional hydrolysis effect in the PBT due to moisture might have weakened the surrounding material.

Surface of the contaminated sample sheet in reflected light bright field

Fracture surface in reflected light Fracture initiation by the carbon black particle

Thin section through the contaminated sample sheet in polarized light

Thin section through the agglomerate in a bright field

Carbon black agglomerate

11

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

3.10 Inhomogeneous color masterbatch dispersion Problem: In-service failure of molded part Material: POM Cause of problem: Failure due to inhomogeneous pigment dispersion, particularly as a result of color masterbatch accumulations in the flow front zones. Note: The areas of large pigment accumulation at the flow fronts are potential crack initiation points, as evidenced by these very striking thin section images.

Thin section in direct proximity to the fracture in polarized light

Thin section in direct proximity to the fracture in a dark field

12

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

Fracture surface in reflected light

3.11 Soft phase separation Problem: Failure of the molded part due to stress Material: POM Cause of problem: Soft phase separation and elongation of the soft phase due to unsuitable injection molding parameters. Note: The reflected light photo shows the fracture surface of the component. It can be clearly seen here that this is a smooth brittle fracture. In impact-resistant POM under normal temperature conditions, such brittle fractures can only occur at flow fronts or at interfaces between elongated soft phase regions and the matrix. On the phase contrast photo, the light areas represent the matrix and the gray areas the soft phase. The photo shows soft phase separations, elongated soft phase flow fronts and areas with high soft phase stretching where cracks have already formed.

Thin section from the fracture zone in phase contrast

Cracking

13

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

3.12 Component damage as a result of external effects Problem: Failure of the molded part in a high-temperature test Material: POM Cause of problem: Thermal damage to the external wall Note: The component failed as a result of thermal stress and the associated surface damage after installation of the part. During high-temperature testing, a crack formed on the outside surface of the component precisely at the point where the outer layer had melted the most. It then spread progressively towards the inside of the component. The crack was initiated by the rough surface (notch effect) and also by internal stresses due to recrystallization under high-temperature stress.

Damage point in reflected light

Damaged component surface

Thin section at the damage point, in polarized light Melting zone

14

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

Non-nucleated material with long cooling time in polarized light

3.13 Different structure Problem: Prolonged cooling time Material: POM Reason for problem: Use of non-nucleated material in place of nucleated material Note: Nucleation reduced the crystallization half-life in the specific case from 300 s to 60 s.

Nucleated material with short cooling time in polarized light

15

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

3.14 Inhomogeneous melt Problem: Failure of molded part in low-temperature test Material: PBT Cause of problem: Inhomogeneous melt Note: The sample with the fracture shows far more pronounced flow fronts than the sample without the fracture. This prominence of the flow fronts is due to pigment accumulation in the flow front zones and is generally caused by processing at a too low temperature or with insufficient back pressure.

The sample with a fracture in a dark field

The sample without a fracture in a dark field

16

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

Fracture surface in reflected light

3.15 Insufficient holding pressure time Problem: In-service failure of molded part

Sheath-core structure

Material: POM Cause of problem: Insufficient holding pressure time Note: On the reflected light photo, a sheath-core structure can be seen, which can be examined more closely in the thin section taken from this zone. Along the flow stream, two red lines are evident, which delineate the core structure from the sheath structure. These are holding pressure lines.

Thin section from the fracture zone in polarized light

At the point in time when these lines were formed, there was a sudden drop in injection pressure. This led to a change in the temperature gradient and the crystallization rate. The result was the formation of a new orientation zone. From this, it can be deduced that the holding pressure time was insufficient, leading to a loss of component toughness.

Holding pressure lines

17

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

3.16 Cracking in the molded part Problem: Molded part not leaktight Material: PPS

Cause of problem: Microcracking in the molded part Note: The cracking was probably due to internal stresses in flow limit areas

Polished section in reflected light bright field

18

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

Overview of a thin section from the weld line zone in polarized light

3.17 Weld line too cold Problem: Failure of the molded part in a test Material: POM Cause of problem: The structureless weld line indicates that the flow fronts were too cold in the reunification zone.

Enlargement of thin section from the weld line zone in polarized light

19

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

4. Theoretical principles of microscopy

4.1 Principle of birefringence To explain the principle of birefringence, the terms refractive index and optical path difference are discussed in greater detail below. 4.1.1 Refractive index
n =1.5

Fig. 4: Illustration of optical path difference using the example of three waves [3]
n =1

A1

In a vacuum, all electromagnetic waves travel at the speed of light: 299,550 km/s. In transparent objects, the speed of light is slower. The ratio of the speed of light in a vacuum to that in the specified medium is known as the refractive index. Every object has three refractive indices na, ny, nz. If these refractive indices are identical, then the object is called isotropic and exhibits no birefringence. 4.1.2 Optical path difference Fig. 4 shows three initially equiphase waves, a, b, and c, and how they are slowed down in passing through media of various refractive indices. The wave trains

B b B1

n=3

C c C1

reach points A, B and C in the three media at the same moment. After three cycles, the waves reach points A, B and C. Wave b is half a wavelength ( /2 ) behind

Fig. 5: Illustration of birefringence [3]

A
D

E E

= F

B B A

C C

20

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

wave a and wave c is two wavelength units (2 ) behind wave a. Such delays are known in optics as optical path differences or phase differences. Birefringent objects appear alternately bright and dark when rotated between crossed polarizers. After each 90 turn, the object becomes dark; this is known as the normal position. The maximum brightness of the object occurs at 45 and 135 between two dark positions; this is known as the diagonal position. The normal position occurs when the transmitting direction of the object is parallel to the polarizer. Since in this case the vibration direction (main birefringence) of the object is parallel to the polarizer, the second beam produced by the birefringence is impeded and the object appears dark (see Fig. 5). In the diagonal position, the beam coming from the polarizer passes through the object and is split into two beams with the vibration directions a and y. Beams a and y have different velocities due to their

different refractive indices na and ny caused by their different orientations within the object. This gives rise to a delay, which is known as optical path difference. The birefringence depends both on the thickness d of the object and on the velocity difference (path difference) n of the two beams and is calculated according to the following formula: = nd

The mixed white light used in light microscopy consists of different colors or wavelengths of the visible light spectrum. The different wavelengths are extinguished in a quartz wedge at different thicknesses, while other wavelengths with full or reduced intensity appear at the same points. These different mixtures in the quartz wedge give rise to the sequence of interference colors shown in Fig. 6. Interference colors occur in repeating groups known as orders. Each order spans 551 nm. This unit width was established because 551 nm represents the subjectively perceived brightest point in the solar spectrum. With increasing order, interference colors become progressively paler. Since, in light microscopic analysis, the thickness of the object is virtually the same throughout, the interference colors reveal the state of orientation of the object. In practice, the first order occurs most often, which can only be seen in black and white photographs. To show stress differences better the human eye can distinguish color differences better than grayscale values the order is increased by 1 with a lambda plate so that the second order is obtained. Since in practice it is mainly the relative orientation differences in an object that are important, it is possible to dispense with absolute measurement in most cases. For practical purposes, the following can be concluded: The more frequently the sequence of interference colors is repeated, the more pronounced the molecular orientation in the component.

Fig. 6: Sequence of interference colors [3]

black lavender gray blue-gray pale yellow bright yellow reddish orange deep red indigo sky blue greenish blue pale green pure yellow orangish red dark violet red indigo greenish blue sea green greenish yellow flesh colors chimney red matt purple blue gray bluish green pale greenish gray whitish gray flesh red

1st order

200

400

1
600

2nd order

800

Path difference 3

1000

3rd order

1200

1400

1600

4st order

1800

2000

21

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

5. Microscopy methods
5.1 Transmitted light microscopy To visualize the internal structures of unreinforced polymers, the most common approaches involve various transmitted light microscopy methods, e.g. bright field, dark field, polarization and phase contrast methods. Their specific applications depend on the particular requirements and on the problems being analyzed. With transmitted light, it is possible to discern 5.1.2 Phase contrast method the occurrence and development of cracks phase separation and poor dispersion in multiphase materials weld lines, which permit conclusions to be drawn about mold filling and the quality of the joining processes the morphology of weld lines, which gives an indication of processing parameters the size and shape of spherulites in semi-crystalline thermoplastics the stress or orientation condition as a function of processing parameters 5.1.1 Bright field method in transmitted light In bright field microscopy (Fig. 7), light is aimed through specimen without any further interference. The transmitted light method uses the light beams transmitted through the object and the reflected light The primary beam is weakened in intensity by 25% and shifted by 1/4 through the phase ring fixed into In the Zernike phase contrast method, phase differences in the light wave, which depend on differences between the refractive indices of the object and surrounding medium, are converted into visible intensity differences. This is therefore a type of optical staining without changing the object. For this purpose, the pinhole diaphragm is replaced by a diaphragm ring. This is sharply imaged by the condenser in a plane above the objective lens. In this plane there is also the phase ring, which is dimensioned so that it corresponds precisely to the diaphragm ring (see Fig. 8). method uses the beams reflected from the object. Microscopic objects in the sample can be more easily identified if they have a different absorption coefficient from that of the surrounding medium, e.g. carbon black, talc, pigments, etc. in a polymer. This is referred to as an amplitude contrast. The transmission difference should be more than 10% to obtain a sufficient contrast. In this way, it is possible to visualize object structures in grayscale values.

Fig. 8: Principle of the phase contrast method according to Zernike

Fig. 7: Principle of bright field microscopy

Eyepiece Objective lens Object Condensor diaphragm Luminous-field diaphragm Objective lens Preparation Condenser Phase plates

/4

Diaphragm ring

22

/2

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

the objective lens. As a result of this, the phase with the higher refractive index appears dark, which also corresponds to our intuitive expectation. As a result of refraction in the object, secondary beams are produced. When the object is imaged, the secondary and primary beams are combined and interfere with each other so that small refractive index differences appear in the photograph as brightness differences. 5.1.3 Dark field method in transmitted light Unlike the previously described microscopy methods, dark field microscopy uses only the diffuse light scattered by the object to form an image. The direct beams are guided past the objective lens. Interfaces and cracks on which light is refracted are particularly visible. Preparation-free areas and homogeneous zones, on the other hand, appear dark. To achieve this effect, special condensers or ring diaphragms such as those used in the phase contrast method are employed. In this microscopy method, however, it should be remembered that light reflections shown in the image do not correspond to the actual object size but are disproportionately enlarged due to irradiation effect. 5.1.4 Polarization method in transmitted light Polarizing microscopy is used to study materials that behave anisotropically, i.e. that are birefringent and Fig. 9: Principle of dark field microscopy

not strongly light-absorbing. Birefringence is frequently associated with crystalline molecular structure or with structural arrangement of non-molecular orders of magnitude. In this method, crystals and oriented zones appear bright with suitable orientation, while isotropic objects are dark. A bright field microscope is converted to a polarizing microscope by fitting a polarizer below the object and an analyzer above it. By inserting a lambda plate between the object and the analyzer, the optical path difference is increased by 1 , thereby causing the images to become colored. Polarization aligns light beams into a uniform vibration direction. Beams that already lie exactly in the so-called transmitting direction pass through with only a slight loss of intensity. The remaining beams are completely stopped. Crossed polarization filters, also referred to as polarizer and analyzer, completely block the passage of light. 5.2 Reflected light microscopy Reflected light microscopy is mainly used to assess surface structures of molded parts. It is also used, however, to study the internal morphology of fiber-

Fig. 10: Principle of the polarization method in transmitted light

5
Objective lens

Analyzer plate
Object

Condenser Diaphragm ring

Polarizer

23

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

reinforced components, which due to their structure do not lend themselves to thin-section techniques. In addition to assessing fracture processes and surface defects such as the record groove effect, streaks, warpage, the diesel effect, etc., it is also possible to verify the morphologies of component interiors using the polishing technique. In this technique, the molded part is embedded in a cold-curing reactive resin and then ground and polished. Such preparations clearly show fiber alignment, fiber distribution and fiber orientation. Information is also provided on voids and weld line morphology. Preferred microscopic methods include dark field, differential interference contrast, polarization and oblique illumination. The principle of these microscopic methods in reflected light largely corresponds to that for transmitted light microscopy, except that the beam path is different. For this reason, only the differential interference contrast method is described. 5.2.1 Differential interference contrast method in reflected light In the differential interference contrast method, contrast is once again achieved by the interference between two wave trains with a phase shift (optical path difference). The optical path difference in this method is produced by the different distances covered by two wave trains due to height differences in the preparation. Linear polarized light is used for differential interference microscopy. The polarized beam is split by a Wollastan or Normarski prism into two polarized, diverging beams of equal intensity perpendicular to each other. The size of the beam split is deliberately chosen to be smaller than the resolution of the microscope. In this way, no visible double image can occur. The two beams impinge on the object at two closely adjacent points and their phase is influenced by the different heights of the specimen. The beams reflected from the object are recombined by the Wollastan or Normarski prism, brought into a common vibration direction in the analyzer and can interfere with each other (Fig. 11). The image appears in relief and therefore yields a three-dimensional effect similar to that obtained in oblique bright field illumination. This method is especially well-suited to objects with small height differentials (>10 nm).
24

Fig. 11: Principle of differential interference contrast microscopy

polarizer

analyzer

Normarski prism

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

6. Sample preparation
In failure analysis of unreinforced plastic components, microscopic structural analysis is a fast and often conclusive method of determining probable causes of failure by verifying the influence of injection molding process parameters on semi-crystalline structures. Such analyses require light-transmitting specimens produced with commercially available microtomes (rotary microtome). Reinforced polymer structures are best assessed in polished sections or microsections, since because of the cutting process these materials tend to create artifacts when prepared by microtoming. Techniques for preparing polished sections and microsections are described in more detail in the literature (e.g. [4]). To produce high-quality, conclusive micrographic images, the objects to be analyzed must be specially prepared. High-quality, artifact-free sample preparation is the basic prerequisite for interpretation of micrographs. For this reason, selection of the sampling point on the failed part is extremely important. It should be borne in mind that the material property giving rise to the failure might well decrease with distance from the origin of the failure. A classic example of this is the cold slug. To produce thin sections, microtomes with high mechanical stability are required. The cutting knives used are also very important. Hard metal or diamond knives have proven especially useful in preparing polymer thin sections. The disadvantage of diamond knives is their extremely limited cutting surface. The most important microtome knife angles are known as the clearance angle, wedge angle and rake angle, based on DIN standard tool terminology. The rake angle and clearance angle are variable and together with the wedge angle always comprise 90. To minimize the load on the knife and achieve good cutting results, the clearance angle should be between 3 and 5. The microtome knife must be clamped firmly without any play. The same applies to the sample, which is aligned in the cutting direction and must also be clamped firmly in the sample holder. To avoid internal vibrations in the sample, it should not project more than about 3 mm from the sample holder. If it is not possible to clamp the sample because of its size, it can be embedded in a suitable resin but, here again, care must be taken to ensure that the sample projects no more than about 3 mm from the embedding mold. The best cutting results are obtained by uniform cutting, i.e. guiding the object steadily past the knife at low cutting speed. Additional factors that influence the cutting result are the setting angle and the slice thickness. The setting angle is the angle of the blade to the object. To minimize the cutting forces on the blade, the setting angle is increased in proportion to the width of the surface being cut and to the strength of the sample material. To achieve perfectly sharp image quality, the thin sections should always be thinner than the smallest object being analyzed. As a rule, the thin section thickness is between 2 m and 10 m. Generally speaking, the thinner the thin section, the better the image quality, since outlines can then be clearly distinguished. Depending on the specific properties of the object, the microtome section can be cut at room temperature or below. Cooling is carried out with liquid nitrogen. The thickness of the microtome section depends on the type of material, its composition and the problem being analyzed. Many years experience with polymers and good material knowledge are of considerable advantage in selecting the right thickness. Thin sections from the cutting operation are then transferred with fine tweezers onto a glass slide, where they are embedded in cedarwood oil with the same refractive index as the glass and protected with a cover glass. Embedding the thin sections in cedarwood oil serves to suppress any interfering refraction effects. Thin sections so prepared can then be used for microscopic analysis.

25

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

7. Literature

[1] Ehrenstein, Prof. Dr. & Raue, Dr. (2001). Vorgehensweise bei der Kunststoffschadensanalyse (1). Kunststoffberater, 4/2001, S. 28-31. [2] Michler, Georg (1992). Kunststoff Mikromechanik. Mnchen: Hanser. [3] Patzelt, Walter J. (1974). Polarisationsmikroskopie. Wetzlar: Ernst Leitz GmbH. [4] Ehrenstein, Prof. Dr. (1992). Kunststoff Schadenanalyse. Mnchen: Hanser.

26

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

NOTICE TO USERS: To the best of our knowledge, the information contained in this publication is accurate, however we do not assume any liability whatsoever for the accuracy and completeness of such information. The information contained in this publication should not be construed as a promise or guarantee of specific properties of our products. Further, the analysis techniques included in this publication are often simplifications and, therefore, approximate in nature. More vigorous analysis techniques and prototype testing are strongly recommended to verify satisfactory part performance. Anyone intending to rely on any recommendation or to use any equipment, processing technique or material mentioned in this publication should satisfy themselves that they can meet all applicable safety and health standards. It is the sole responsibility of the users to investigate whether any existing patents are infringed by the use of the materials mentioned in this publication. Properties of molded parts can be influenced by a wide variety of factors including, but not limited to, material selection, additives, part design, processing conditions and environmental exposure. Any determination of the suitability of a particular material and part design for any use contemplated by the user is the sole responsibility of the user. The user must verify that the material, as subsequently processed, meets the requirements of the particular product or use. The user is encouraged to test prototypes or samples of the product under the harshest conditions to be encountered to determine the suitability of the materials. Material data and values included in this publication are either based on testing of laboratory test specimens and represent data that fall within the normal range of properties for natural material or were extracted from various published sources. All are believed to be representative. These values alone do not represent a sufficient basis for any part design and are not intended for use in establishing maximum, minimum, or ranges of values for specification purposes. Colorants or other additives may cause significant variations in data values.

We strongly recommend that users seek and adhere to the manufacturers current instructions for handling each material they use, and to entrust the handling of such material to adequately trained personnel only. Please call the numbers listed for additional technical information. Call Customer Services at the number listed for the appropriate Material Safety Data Sheets (MSDS) before attempting to process our products. Moreover, there is a need to reduce human exposure to many materials to the lowest practical limits in view of possible adverse effects. To the extent that any hazards may have been mentioned in this publication, we neither suggest nor guarantee that such hazards are the only ones that exist. The products mentioned herein are not intended for use in medical or dental implants.

Copyright by Ticona GmbH Edition: September 2004

7
27

Failure Analysis by means of Microscopy

Calculations Design Applications C.4.1

28

Hostaform , Celcon Celanex Impet

polyoxymethylene copolymer (POM)

thermoplastic polyester (PBT)

thermoplastic polyester (PET)

Vandar

thermoplastic polyester alloys

Riteflex Vectra

thermoplastic polyester elastomer (TPE-E)

liquid crystal polymer (LCP)

Fortron
polyphenylene sulfide (PPS)

Celstran , Compel GUR

long fiber reinforced thermoplastics (LFRT)

ultra-high molecular weight polyethylene (PE-UHMW)

Europe
Ticona GmbH Information Service Tel.: +49 (0) 180-5 84 26 62 (Germany) +49 (0) 69-30 51 62 99 (Europe) Fax: +49 (0) 180-2 02 12 02 eMail: infoservice@ticona.de Internet: www.ticona.com

Americas
Ticona LLC Product Information Service Tel.: +1-800-833-4882 Fax: +1-908-598-4306 eMail: prodinfo@ticona.com Internet: www.ticona.com