SEM/TEM FRACTOGRAPHY HANDBOOK

McDonnell Douglas Astronautics Company

Huntington Beach, California
Sponsored by
Air Force Materials Laboratory
Air Force Wright Aeronautical Laboratories
Air Force Systems Command
Wright-Patterson Air Force Base, Ohio
AFML-TR-75-159
DECEMBER 1975
METALS AND CERAMICS INFORMATION CENTER
A Department of Defense Information Analysis Center
Approved for public release; distribution unlimited.
MCICHB-06
ACKNOWLEDGEMENT
This document was prepared by the Metals and Ceramics Information Center (MCIC), Battelle's
Columbus Laboratories, 505 King Avenue, Columbus, Ohio 43201. MCIC's objective is to provide
a comprehensive current resource of technical information on the development and utilization
of advanced metal- or ceramic-base materials.
The Center is operated by Battelle-Columbus under Contract Number DSA900-75-C-1803 for the
U.S. Defense Supply Agency; technical aspects of MCIC operations are monitored by the Army
Materials and Mechanics Research Center. The support of these sponsor organizations is gratefully
acknowledged.
This document was prepared under the sponsorship of the Department of Defense. Neither
the United States Government nor any person acting on behalf of the United States Government
assumes any liability resulting from the use or publication of the information contained in this
document or warrants that such use or publication will be free from privately owned rights.
Approved for public release; distribution unlimited.
All rights reserved. This document, or parts thereof, may not be reproduced in any form
without written permission of the Metals and Ceramics Information Center.
ii
FOREWORD
This final Technical Report was prepared by the McDonnell Douglas Astro-
nautics Company, Huntington Beach, California under Contract No. F336l5-74-C-
5004. The time period covered by the contract was from 15 October 1973 to
15 June 1975. The work was conducted under the direction of the Air Force
Materials Laboratory, with Mr. R. Henderson (AFML/MXA) as Project Engineer.
This contract was initiated under Task No. 738103.
The program at McDonnell Douglas was under the direction of Mr. G. F.
Pittinato, Principal investigator, with Mr. V. Kerlins, Mr. A. Phillips, and
Mr. M. A. Russo as coinvestigators. Appreciation is expressed to Mr. H.
Taketani and Mr. R. A. Rawe for their assistance in this study. The metal-
lographic work was accomplished by Mr. J. L. Evans and Mr. L. Hodde. The SEM
fractographs of the fatigue samples were taken by Mr. R. R. Wilcox.
The authors would like to thank the following for their contribution of
either material or test specimens.
Dr. C. P. Sullivan
Pratt & Whitney Aircraft
East Hartford, Connecticut
Mr. H. A. Williams
General Electric Company
Cincinnati, Ohio
Mr. R. A. Lauchner
Northrop Corporation
Hawthorne, California
Mr. D. E. Lane
Martin Marietta Aluminum, Inc.
Torrance, California
iii
Mr. J. Moore
Pratt & Whitney Aircraft
West Palm Beach, Florida
Mr. L. J. Barker
Kaiser Aluminum & Chemical Corp.
Pleasanton, California
Mr. P. W. Kroger
ALCOA
Vernon, California
Mr. R. V. Turley
Douglas Aircraft Company
Long Beach, California
NOTICES
When Government drawings, specifications, or other data are used for any
purpose other than in connection with a definitely related Government procurement
operation, the United States Government thereby incurs no responsibility nor any
obligation whatsoever; and the fact that the Government may have formulated,
furnished, or in any way supplied the said drawings, specifications, or other
data, is not to be regarded by implication or otherwise as in any manner licensing
the holder or any other person or corporation, or conveying any rights or per-
mission to manufacture, use or sell any patented invention that may in any way
be related thereto.
This technical report has been reviewed and is approved for publication by
the Public Information Office (PIO) and is releasable to the National Technical
Information Service (NTIS).
RUSSELL L. HENDERSON
Project Engineer
Corrosion Control & Failure Analysis
FOR THE COMMANDER
T ~ P E ~
Copies of this report should not be returned unless return is required by
security considerations, contractual obligations, or notice on a specific docu-
ment.
iv
TABLE OF CONTENTS
INTRODUCTION . . . .. 1
TECHNIQUES - SECTION I . . . . . 3
Introduction . . . . . .
Care and Handling of Fractures
Cleaning Fractures
Visual Examination .
SEM Sample Preparation
TEM Sample Preparation
Special Techniques
FRACTURE MODES - SECTION II
Introduction .
Dimple Rupture
Cleavage .
Fatigue
Decohesive Rupture
Miscellaneous Fracture Surface Features
ATLAS OF FRACTOGRAPHS - SECTION III
Introduction
Mechanical Testing .
Fractographic EXamination
Index of Atlas Fractographs
5
5
5
7
7
9
17
23
25
25
13
39
49
54
59
61
61
64
67
REFERENCES . 687
v
INTRODUCTION
The use of electron fractography has become a standard practice in service
failure analysis. As an aid to interpreting electron fractographs, a trans-
mission electron microscope (TEM) Fractography Handbook (ML-TDR-64-4l6) was
prepared in 1965. This handbook, which characterizes the fracture appearance
of various aircraft materials fractured under known conditions, has proven to
be an invaluable aid in the identification of fracture modes in service fail-
ures. However, since the publication of the TEM Fractography Handbook, new
alloys for aircraft and engine applications have been developed and there has
been increasing use of the scanning electron microscope (SEM) for examining
fracture surfaces.
The SEM provides a rapid means for the direct examination of fracture surfaces,
thus permitting fracture modes to be determined more rapidly than they could
be with the TEM. However, at equivalent magnifications, the TEM fractographs
exhibit sharper detail. It has also been found that identical fracture features
appear somewhat different when viewed in the SEM as compared to the TEM. Con-
sequently, there is a need for SEM characterization of fractures obtained
under known conditions as well as a pictorial comparison of fracture modes
obtained using both the SEM and TEM.
The present handbook contains both SEM and TEM fractographs and can be used
as an effective reference handbook to aid investigators in fracture surface
analyses. Basic specimen preparation techniques and the interpretation of
electron fractographs are discussed in detail. However, it is assumed that the
reader is a materials engineer, is familiar with the operation of an electron
microscope, and has had experience in failure analysis, since no effort is
made to expound on these subjects. The use of the electron microscope for
fracture analysis adds one more tool to assist the failure investigator in
his analysis, and augments, rather than replaces, well established failure
analysis techniques.
TECHNIQUES
SECTION 1
3
INTRODUCTION
The diversity of problems associated with fracture surface analysis precludes
the use of fixed rules or techniques for examining a fracture. Instead, the
investigator must decide what specific information is required from the frac-
ture and what techniques are available for obtaining this information. This
section of the handbook covers the basic procedures and techniques used in
electron fractography. Some of these procedures and techniques apply
regardless of the type of electron microscope used in the examination, while
others are specifically designed for scanning (SEM) or transmission (TEM)
electron fractography.
CARE AND HANDLING OF FRACTURES
When a fracture requires laboratory examination, both mating surfaces should
be preserved either by the application of a protective coating*, by placing in
a desiccator, or by sealing in a plastic bag containing a desiccant to prevent
any accumulation of undue moisture until the examination can be made. The
coating used should be soluble in an organic or other mild solvent so that it
can be completely removed prior to examination. Touching the fracture surface
with fingers, rubbing, or fitting the fractures together can cause serious
damage. Picking at the fracture with a sharp instrument should also be
avoided. Rough treatment or the formation of corrosion products on the fracture
can obscure vital information. Education in the proper handling of specimens
prior to any fractographic examination is strongly recommended for anyone deal-
ing in fractures either in the field or in the laboratory.
CLEANING FRACTURES
The fracture appearance should be documented by photographing or the taking of
notes before any cleaning is attempted. Also, it must be ascertained whether
identification of foreign products on the fracture will aid in the failure
analysis. Identification of these products can be quite useful in pinpointing
adverse environmental conditions that contributed to the fracture. Hasty clean-
ing can remove important evidence. The problem of cleaning the fracture
surface should be approached with caution and common sense.
* Krylon Crystal Clear Spray No. 1302, Borden, Inc. Department CP, N.Y.,N.Y.
5
It is difficult to present detailed cleaning procedures which would apply to
all fracture surfaces, since different metals are involved, and varying degrees
of surface contamination are encountered. As a general rule, the mildest,
least damaging cleaning procedure should be used. In most cases, repeated
stripping of a plastic replica is sufficient to clean a fracture. If a
cleaning solution is required, one should be chosen that will not attack the
fracture surface, but yet remove the undesirable contamination. In the case
of oil or grease, organic cleaning solutions such as acetone or trichloroethy-
lene may be used. If an immersion treatment is not sufficient, vapor degreasing
or ultrasonic procedures may be employed. It is not recommended that a metal
brush or other harsh mechanical means be used to remove contaminants; however,
light cleaning with a soft organic fiber brush is permissible.
Oxidation, corrosion, or other chemical reaction products are usually more
difficult to remove. In these instances, mild acid or alkaline solutions such
as acetic acid, orthophosphoric acid, or sodium hydroxide, heated if required,
may be employed. Commercial ultrasonic cleaning equipment manufacturers
supply special inhibited cleaning solutions which remove oxides from metal
surfaces. It should be remembered that chemical reactions such as oxidation
and corrosion consume the base metal. Therefore, part of the fracture is
inherently destroyed and removing this oxide layer will not restore the frac-
ture to its original condition.
The following are cleaning solutions which are used for specific applications:
Removal of oxide from aluminum alloys:
70 cc orthophosphoric acid (85%)
32 g chromic acid
130 cc water
Solution may be warmed.
Removal of rust from steel:
1. Orthophosphoric acid. Use concentrated or diluted up to 50% with water.
Solution may be warmed.
2. 100 cc 6N HC1 inhibited with 0.2 g hexamethylene-tetramine.
Use at ambient temperature.
6
Removal of residual salt (NaCl) deposits:
Immerse specimen in glycerin.
After cleaning by any of the above solutions, the specimen must be thoroughly
rinsed separately in water and alcohol and then dried.
VISUAL EXAMINATION
A fracture must be considered in its entirety because the examination of only
one small area may lead to an inaccurate interpretation of the fracture mode.
It is essential, therefore, to examine the fracture origin as well as adjacent
areas.
The initial step in the examination of a fracture is to determine the location
of the fracture origin, and subsequently, the exact areas for examination in
the electron microscope. Visually or by using a stereoscopic light microscope,
it is usually possible to determine the origin of a fracture by such features
as chevron marks, hackle marks, fLip-flops, texture changes, radial fracture
traces, or by the absence of shear lips along the edges. These methods are
discussed in detail in Reference 1. A fractographic method for locating the
fracture origin is discussed later in this section under Special Techniques.
SEM SA}fPLE PREPARATION
Viewing a fracture in the SEM requires that the sample be cut and subsequently
mounted on a relatively small sample holder. In mounting the sample, it is
absolutely essential that a conductive path (ground) exists between the point
where the electron beam strikes the sample and the holder. For metallic
specimens which have a clean surface and are electrically conductive, the
sample is simply grounded to the holder by using a commercially available
conductive paint such as Television Tube Koat.* For optimum grounding, the
area where the conductive paint contacts the sample and the holder should be
clean and free of oxide coatings. This can be accomplished by lightly sanding
the contact surface of the sample, as well as the holder, and wiping the sanded
*Television Tube Koat is manufactured by G. C. Electronics, Rockford, Illinois.
7
areas with a solvent. If a specimen is cut, the clean, cut surface serves as
a good contact area for grounding. Sample holders which are repeatedly used
generally accumulate fingerprints or debris and are normally lightly oxidized.
Since oily deposits and oxides (especially aluminum oxide) are insulators,
the cleanliness of the sample holder is essential, and often overlooked, in
obtaining a proper ground.
Nonconductive surfaces on the sample must be coated with a thin conductive
material to prevent them from accumulating an electrical charge from the elec-
tron beam, Figure 1. In practice, this is accomplished by grounding the sample
to the holder and then vacuum vapor depositing or sputtering a thin conductive
coating such as gold, gold-palladium, or carbon on its surfaces. Rotating the
sample during vapor deposition ensures a uniform conductive coating and pre-
vents the formation of shadows, Figure 2. For most applications, a 1.5 inch
(3.8 cm) length of 0.008 inch (0.020 cm) diameter gold wire evaporated on a
rotating sample placed approximately 2.5 inches (6.4 cm) from the gold source
(basket) will provide a satisfactory coating. These coatings can also be
S14685 sm 2000X
Figure 1 Charged particle (arrow)
on a fracture surface.
8
S14586 8m 1600X
Figure 2 Dark area (arrow) resulting
from an uneven distribution of vapor-
deposited gold.
S14686 SEM (A) l800X S14687 SEM (B) 2200X
Figure 3 SEM fractographs of a lightly oxidized fracture showing the effect
of a poorly conductive surface. (A) as oxidized, (B) gold coated.
applied to metal surfaces to improve their image quality, Figure 3. Some con-
ductive spray coatings are available, however, these are inferior to vapor-
deposited metals and are generally unsatisfactory for fracture analysis.
Aside from a poorly conductive surface, any sample which is even slightly
magnetic will yield poor image quality due to a defocusing effect.(l) There-
fore, it is a good practice to demagnetize (degauss) samples of materials
that can be magnetized because such operations as magnetic particle inspection
or cutting can result in residual magnetism. Small inexpensive demagnetizing
coils are commercially available.
TEM SAMPLE PREPARATION
The actual fracture surface can not be viewed in the TEM because the image is
formed by an electron beam which must pass through the sample. For this
reason, various methods have been developed for replicating the fracture
9
detail on very thin, shadowed carbon films that are transparent to the electron
beam. There are a number of techniques that can be used to replicate the
(2 to 5)
fracture surface ; however, only the plastic-carbon method will be
discussed in detail. This technique is convenient to use, does not destroy
the fracture surface, has good resolution, and satisfactory results
for all routine fractographic examinations.
The plastic-carbon replication technique is shown schematically in Figure 4.
This technique involves replicating the fracture surface with plastic, deposit-
ing a metal (shadowing) and carbon on the plastic, and finally dissolving the
plastic away from the shadowed carbon replica. Each of these steps is
examined in detail in the following discussion.
Plastic Replication
Cellulose acetate tapes ranging in thickness from 0.001 to 0.005 inch (0.003
to 0.013 cm) are used to replicate fracture surfaces. A general rule for
selecting a tape thickness is the more jagged the fracture, the thicker the
tape. However, a thin tape should be used in preference to a thicker one
whenever possible. In many instances, rather than USing a single thickness
of the 0.005 inch (0.013 cm) tape, a double thickness of the 0.001 inch
(0.003 cm) tape may be preferred. The plastic used should be dipped in a
cleaning solution, such as DuPont Freon PCA, to remove any surface contamina-
tion. The plastic can also be cleaned by wiping with lens tissue.
Two different methods can be used to obtain a plastic replica of a fracture
surface. The thickness of the tape that is being used establishes the method
of application. When using cellulose acetate tapes that are less than approx-
imately 0.003 inch (0.008 cm) thick, an acetone solution is used to partially
soften the plastic prior to placing it on the fracture surface. The tape is
softened only to the extent that it will follow the contours of the
fracture. Too soft a plastic film or excessive use of acetone may result in
the formation of vapor bubbles at the plastic-metal interface. The applied
tape is then permitted to dry in place at least 10 minutes without any
application of pressure or heat. When the tape has thoroughly dried and
hardened, tweezers are used to remove it from the fracture. This "negative
replica" of the fracture surface, trinnned to the desired size, is now ready
10
y--
CHROMIUM COATING
............ --
CHROMIUM......... ......... - - -
VAPORIZED .................. - -_.
IN TUNGSTEN
SPIRAL
SIDE
SPHERE
-
-
-
-
METAL SURFACE
WITH SCRATCH
PLASTIC F ILIvi
PLASTIC :FILM
STRIPPED, IhvERTED,
AND SHADOWED WITH
CHROMIUM.
-
CARBON ARC
DEPOSITION OF CARBON
ELECTRON BEAM
v
o
Figure 4 Plastic-carbon replication method.
11
CELLULOSE ACETATE
REMOVED IN ACETONE
FILM,INVERTED,AND
EXAMINED IN
MICROSCOPE.
for shadowing and carbon deposition. Arrow points may be cut on the plastic
replica to orient it with respect to the fracture surface.
If the fracture surface is rough or jagged and the use of a double-thickness
thin tape is not satisfactory, a 0.005 inch (0.013 cm) thick cellulose acetate
tape should be employed. Because of the thickness of the tape, even if it is
softened in acetone, it has little tendency to follow the rough surface
contours. For this reason, some liquid replicating solution (cellulose ace-
tate dissolved in acetone) is applied to one surface of the tape. The
replicating solution is allowed to partially soften the tape (usually 2 to 3
minutes). Just prior to replication, a coating of the replicating solution is
applied to the fracture. The solution-covered side of the tape is then
pressed on to the wetted fracture surface. A firm pressure is exerted on
the tape for about 1 to 2 minutes. When the acetone has evaporated (usually
30-45 minutes), the cellulose acetate from the solution and the tape form a
continuous replica which is then removed from the fracture surface as one unit.
If difficulty is experienced in stripping, it may be found advantageous to
heat the replica and fracture in an oven at 200F (93C) for approximately
15 minutes and cool in air to room temperature. The heating process thoroughly
dries the plastic permitting easier stripping with a minimum of replica distor-
tion. The replicas are then placed on a glass slide with the impression side
up in preparation for shadowing and carbon deposition.
Shadowing Techniques and Carbon Deposition
In order to increase the contrast and give a replica a three-dimensional
effect, a process known as shadowing is used. Shadowing is an operation
whereby a heavy metal is deposited at an oblique angle to the surface by
evaporating it from an incandescent filament or an arc in a vacuum chamber,
Figure 4. The shadowing material is deposited at an angle of approximately
45 (smooth surfaces require lower angles) in such a way as to relate with
a known direction such as the macroscopic fracture direction. The vaporized
metal atoms travel in essentially straight lines from the filament and strike
the replica at an oblique angle. Upon contact, the metal condenses where
it strikes. Certain favorably oriented surface features receive a thicker
metal deposit than others, and in fact, some areas adjacent to surface
12
protrusions or depressions receive no metal deposit at all. The difference
in deposited metal thickness between the front and back side of a surface
feature produces a difference in contrast due to an increase in electron
scattering by the metal coated areas. Since electron scattering is a function
of the atomic number and mass density of the element, it would require less
deposition of a heavy metal to produce a desired contrast. Commonly used
shadowing materials are chromium, germanium, palladium, palladium-platinum,
platinum-carbon, and gold-palladium. The minimum amount of shadowing material
necessary to produce contrast should be employed when maximum resolution is
required.
There are several methods used to judge the thickness or amount of shadowing
material deposited. One method is to use a drop of silicone oil on a glass
slide which is placed over a white piece of paper. The difference in contrast
between the paper under the oil drop and the surrounding area gives an indication
of the amount of metal deposited. A finger print on a white piece of paper
or a small solid object placed on a slide may be used to observe the development
of an actual shadow and thus provide a way for estimating the amount of metal
deposited. A good indication of a sufficient amount of deposited material is
the inception of a change in contrast between the shadowed and unshadowed regions.
An insufficient amount of shadow material will result in poor contrast, while
too much material may obscure surface detail and give accentuated shadows.
To avoid granulation of the shadow material, it is necessary to maintain a
high vacuum condition (about 10-
4
to 10-
5
torr) during the shadowing process.
After shadowing, granulation can also result from exposure to too high an intensity
electron beam. Gold shadowing is particularly sensitive to this phenomenon.
After the shadowing operation, carbon is deposited on the replica. It is this
carbon film with the shadowed metal that is ultimately examined in the TEM.
Carbon is deposited either normal to or at a slight angle to the surface of the
replica. In order to ensure a uniform deposition of carbon, a rotating stage
should be used. The complete evaporation of a 0.040 inch (0.1 em) diameter
0.40 inch (1 cm) long carbon rod located about 5 inches (12.7 cm) above the
13
replica is usually sufficient to yield a sound replica. For optimum results,
the vacuum should be less than 5 x 10-
4
torr while depositing the carbon.
Dissolution of Plastic
The replica is immersed in acetone to dissolve the cellulose acetate. A
gentle periodic agitation of the acetone is recommended to facilitate the
dissolving of the plastic. Due to the expansion of the cellulose acetate
during dissolution in acetone, the carbon film may distort and fragment into
pieces so small as to render the replica useless. To prevent this undesirable
condition, the following procedures may be employed:
(1) Use a thinner strip of plastic.
(Z) Use a solution of acetone diluted to 50% by ethyl alcohol or distilled
water. After soaking the replica in this diluted acetone solution until
most of the plastic has dissolved, place it in pure acetone to dissolve
the remaining cellulose acetate.
(3) Use a warm solution of acetone or vapors of acetone to dissolve the
plastic.
(4) Place the trimmed replicas, carbon side up, on a piece of filter paper
in a Petri dish. Using an eyedropper, add a 60 to 70% solution of
acetone in water to the dish by allowing the drops of solution to fall
near the edge of the dish well away from the replicas. The objective is
to float the replicas without getting the carbon surface wet. Allow
the replicas to float for several hours and then add straight acetone to
remove the remaining plastic.
(5) Use paraffin to strengthen the carbon film during the dissolution of
the plastic. The paraffin is removed by soaking the replica in benzene.
After the plastic film is dissolved, the carbon replica is picked up on a 75
to ZOO mesh screen. Screens of various grid configurations are available. The
screen holding the replica is then placed on filter paper in a covered container
until it is inserted in the TEM.
14