09 Microscale Fracture Surface Morphologies
09 Microscale Fracture Surface Morphologies
C.J. Schroeder, R.J. Parrington, J.O. Maciejewski, J.F. Lane, editors All rights reserved
https://doi.org/10.31399/asm.hb.v12.a0006843 www.asminternational.org
WHEN STRUCTURES or components fail lenses, digital microscopes (5 to 200 magnifi- on microscope observations. The key micro-
by fracture, analysis of the new surfaces result- cation), and low-magnification scanning elec- scopic features of fracture surfaces are described
ing from fracture can be used to understand their tron microscopes (<200). Many microscopes and illustrated for the important types of fracture
root cause. One of the most important sources of allow scanning of the fracture surface to create mechanisms.
information relating to the cause of mechanical a montage image assembled from numerous
failures is the fracture surface itself. A fracture adjacent images.
surface is a detailed record of the failure history Imaging microscopic features of the fracture Introduction to Microfractography
of the part (Ref 1, 2). It contains evidence of surface (microfractography) provides more
the loading history, environmental effects, and mechanistic information about the nature or Early research on the fracture of metallic
material quality. Fracture surfaces may be evalu- mode of the fracture. Most of the mechanisms materials focused on the fracture of monocrys-
ated at the macroscopic level (imaging up to proposed to explain the various microscopic talline samples. In such studies, the specimens
200) and the microscopic level. (microscale) features of fracture are often either fractured along a specific (and appropri-
Macrofractography focuses on large-scale based on dislocation generation and interac- ately aligned) crystallographic plane or cleaved
fracture features. As treated in the article “Mac- tion, involving complex slip and crystallo- through the crystal. Classic studies demon-
roscale Fracture Appearance” in this Volume, graphic relationships. The discussion of strated that in uniaxial tension, deformation
the macroscopic fracture features can reveal sub- mechanisms in this article does not include occurred discretely along crystallographic planes.
stantial information about the fracture process, detailed dislocation models or their treatment However, most engineering metals and alloys
including the fracture origin as well as the direc- with complex mathematical models, but this are polycrystalline in nature. Fracture in
tion of crack propagation and thus the failure article does present the mechanisms in more wrought polycrystalline engineering alloys
mode of the component (Ref 3). Typically, general terms in order to impart a practical can occur by a transgranular (through the
macrofractographic documentation is performed understanding as well as an ability to identify grains) or an intergranular (along the grain
using cameras with macro or macro-zoom the basic fracture mechanisms correctly based boundaries) fracture path (Fig. 1). There also
Fig. 1 A crack can propagate through wrought polycrystalline engineering alloys by a transcrystalline (through the grains) or an intercrystalline (along the grain boundaries)
fracture path. (a) Equiaxed polycrystalline grain structure. (b) Transcrystalline fracture through an equiaxed polycrystalline grain structure (transgranular).
(c) Intercrystalline fracture through an equiaxed polycrystalline grain structure (intergranular). (d) Intercrystalline (intergranular) fracture through an equiaxed polycrystalline
grain structure with grain-boundary segregation (indicated as shaded region along grain boundaries). (e) Intercrystalline fracture through a polycrystalline material with flat,
elongated grains
*In accordance with Title 5 Code of Federal Regulations Section §2635.807(b)(2), the views expressed in this article do not necessarily represent the views of the National Transportation
Safety Board or the United States.
are essentially four principal microscopic Cleaning of the fracture surfaces is a key product layers, while retaining the stripped layers
mechanisms of crack initiation and growth in step in microfractography of fractured compo- within the cellulose triacetate films. Hence, dur-
metallic materials (Table 1). Three of the nents, which are often corroded or contaminated. ing a microfractographic evaluation, cleaning
mechanisms occur with transgranular cracking, Often performed after suitable documentation of and microimaging are performed in an iterative
while fatigue or intergranular decohesion are the fracture surface in the as-received condi- manner. Sometimes, the key fractographic
associated with cracking along grain bound- tion, sequential cleaning and microimaging of features of interest are revealed within the cor-
aries. Each of these modes has a characteristic the fracture surface is performed. This work rosion product layers. If the corrosion product
fracture-surface appearance and a mechanism is slow and methodical because, as contamina- was completely stripped, these key fracture
or mechanisms by which the fracture propa- tion and corrosion products are removed, new features would be lost (Ref 6–9).
gates for a given metallic alloy and the envi- information is revealed, while irreversible Throughout this article, the stresses at the
ronmental and loading scenario under which changes in the sample ensue. One technique crack tip are referred to in the traditional frac-
the fracture occurred. It also is noted that the is to use cellulose triacetate replicas to “peel ture mechanics terminology (Ref 10, 11), as
fracture features in as-cast metals may not be off” layers of contamination and loose corrosion indicated in Fig. 2 and Table 2. As a practice,
as clearly delineated or interpreted compared
with wrought polycrystalline metals.
Typically, microfractography is performed y y
using various types of scanning electron micro- F
scopes (SEMs) (Ref 4), transmission electron
microscopes (using carbon-shadowed replicas),
and other high-resolution instruments such as
atomic force microscopes. In all practicality,
x x
most fractography is performed with SEMs F
(tungsten, LaB6, or field-emission gun fila-
ments). In some instances, such as documenting
high-cycle fatigue fracture, features as small as
10 to 100 nm may need to be imaged (typically
requiring a field-emission SEM and careful sam- z z
ple preparation and cleaning). F
For most industrially significant fractures,
F
the fracture surface or portions of the fracture
surface must be removed or excised from the
Mode I Mode II
fractured component to render it small enough
Crack opening In-plane
- shear
to fit in the electron microscope chamber. In
some instances, cellulose triacetate replicas or (a) (b)
silicone replicating compounds are used to col-
lect an inverse copy of the fracture area of y
interest rather than excising a fracture sample F
from the component. For example, in emer- y
gency situations, such as aircraft safety-of-
flight investigations, a replica of the fracture
is often acquired from the component, so that F
microfractography on the replica can be con- x
ducted in parallel with other work, such as x
three-dimensional analysis of the component.
The reader is referred to the literature for such
methods (Ref 5).
z
F z
Table 1 Microscale fracture mechanisms
in polycrystalline metals
Crack path at Fracture mechanism (fracture F
microstructural level micromode)
Mode I Opening (tensile mode) Normal stress Perpendicular to the plane of the crack
Mode II Sliding (in-plane shear mode) Shear stress in the plane of the crack In the plane of the crack perpendicular to the crack leading edge
Mode III Tearing (antiplane shear mode) Shear stress out of the plane of the crack In the plane of the crack parallel to the crack leading edge
Uniaxial tension Tension Uniaxial tension Perpendicular to the plane of the crack
Microfractography in Stress
Modeling
region. A finite-element stress model also con- changes in loading events, and the direction increased to the point that the component breaks,
firmed that a stress concentration created by of crack propagation. See also the section this mechanical scenario is referred to as mono-
geometric features of the void was sufficient “Transcrystalline Fracture: Fatigue” in this tonically increasing load to fracture (monotonic
to develop fatigue cracks at the void surface. article and the article “Quantitative Fractogra- ductile rupture process or monotonic fracture).
The fuel line and train axle examples dem- phy Applications for Fatigue Fracture Sur- In the field of failure analysis, such a macro-
onstrate how fractography, in conjunction with faces” in this Volume. scopic fracture mode is referred to as overload
stress modeling, can create a complete picture or overstress or monotonic fracture.
of the fracture process to support the develop- In overload fracture (in modes I, II, and III,
ment of a root cause of the failure. Another Transcrystalline Fracture— see Fig. 2), polycrystalline materials typically
important application of microfractography is Microvoid Coalescence fracture by microvoid coalescence (MVC),
the measurement of fractographic features on cleavage, or intergranular decohesion (when
fatigue crack surfaces to estimate fracture pro- If a component is mechanically loaded in a creep processes are not activated). When over-
cess, such as rate of fatigue crack growth, manner whereby the applied force is progressively load is the principal cause of fracture, most
common structural alloys fail by MVC. These
microvoids nucleate at regions of localized
strain discontinuity, such as those associated
with second-phase or constituent particles,
inclusions, grain boundaries, and dislocation
pileups and locks. As the strain in the material
increases, the microvoids grow, coalesce, and
eventually form a continuous fracture surface,
as depicted in Fig. 5. This type of fracture
exhibits numerous cuplike depressions that
are the direct result of the growing microvoids
Fig. 4 Fractured portion of a train axle. (a) Macroscopic image of the fracture surface with central void. (b) Digital
microscope image of a portion of the origin region. (c) Secondary electron SEM fractograph of fatigue
striations near the origin region. (d) Digital microscope image showing the demarcation between the progressive
crack growth region and the terminus to overstress fracture. (e) Three-dimensional free body diagram of the Fig. 5 Conceptual sequence revealing the formation
assembled wheels and axle, with loads superposed. (f) Digitally reconstructed cross section of the axle showing the and coalescence of microvoids as a crack
shape and location of the internal void. (g) Finite element stress map in the vicinity of the internal void, showing opens in a ductile matrix with a dispersion of particulates
high stress within a cusp in the void. such as inclusions or constituent particles
during coalescence. The cuplike depressions microvoids early in the loading cycle produce and one end of each dimple is open; that is,
are referred to as dimples, and the fracture a fracture surface that exhibits various dimple the dimples are not completely surrounded by
mode is known synonymously as dimpled rup- sizes (Fig. 10) (Ref 20). When microvoids a rim. In the case of a tear fracture, the elon-
ture or MVC. Figure 6 shows an example of a nucleate at grain boundaries (Fig. 11), inter- gated dimples on both fracture faces are ori-
fracture surface exhibiting MVC (Ref 18). The granular dimpled rupture results (Ref 21). ented in the same direction, and the closed
microvoids are uniform and generally nondirec- Dimple shape is governed by the state of ends point to the fracture origin. This charac-
tional, consistent with fracture in uniaxial ten- stress within the material as the microvoids teristic of the tear dimples can be used to
sion (facing the reader). An MVC fracture form and coalesce (Fig. 12). Fracture under establish the fracture propagation direction
surface can be quite rough and tortuous, as conditions of uniaxial tensile load results in (Ref 26) in a thin sheet that ruptures by a
revealed in the three-dimensional anaglyph in the formation of essentially equiaxed dimples full-slant fracture (by combined modes I and
Fig. 7. Figure 8 is a commonly encountered rep- bounded by a lip or rim. When fracture occurs III). This fracture consists entirely of a shear
resentation of an MVC fracture surface (Ref 19). by a combination of modes I and II, examina- lip and exhibits no macroscopic fracture direc-
As depicted in Fig. 9, the size of the dimples tion of the dimples on mating fracture surfaces tion indicators, such as chevron marks. A shear
on a fracture surface is governed by the num- can reveal the local fracture direction (Ref 22). fracture, however, exhibits elongated dimples
ber, size, and distribution of microvoids that Depending on the microstructure and plasticity
are nucleated. When the nucleation sites are of the material, the dimples can exhibit a very
few and widely spaced, the microvoids grow deep, conical shape, or they can be quite shal- F
to a large size before coalescing, and the result low. The formation of shallow dimples may
is a fracture surface that contains large dim- involve the joining of microvoids by shear
ples. Small dimples are formed when numer- along slip bands (Ref 23).
ous nucleating sites are activated and adjacent Fracture surfaces that result from tear (mode
microvoids join (coalesce) before they have I) or shear (modes II and II) loading conditions
an opportunity to grow to a larger size. For exhibit elongated dimples (Ref 24, 25). The
example, extremely small dimples are often characteristics of elongated dimples are that
found in alloys with numerous fine second-phase they are, as the name implies, elongated (one
particles or in oxide-dispersion-strengthened axis of the dimple is longer than the other),
materials. F
The distribution of the microvoid nucleation
(a)
sites can significantly influence the fracture-
surface appearance. In some alloys, the non-
uniform distribution of nucleating particles F
and the nucleation and growth of isolated
10 Pm
F
Fig. 7 Three-dimensional secondary electron SEM
fractograph anaglyph of a microvoid (b)
coalescence fracture surface on type 6061-T6 aluminum
alloy. The anaglyph, which reveals the height variation
and tortuosity of the fracture surface, was created from
F
stereo pairs. Attributed to E. Komarnicki, National
Transportation Safety Board, Washington, D.C.
F
(c)
10 Pm
10 Pm
Transcrystalline Fracture—
Cleavage
Crystals often slip or twin when subjected to
a monotonically increasing stress process.
When resistance to twinning and slip are
sufficiently high, sudden separation along spe-
cific crystallographic planes under normal
stress can occur. This fracture process is called
cleavage (Ref 42, 43). Transcrystalline cleav-
age is a low-energy fracture process that pro-
pagates through grains in polycrystalline
materials along well-defined, low-index crys-
tallographic planes known as cleavage planes
(Ref 44) (Fig. 1b). Such low-energy fractures
are generally associated with lower fracture
toughness. Theoretically, a cleavage fracture
in a perfect single crystal should have match-
ing faces and should be completely flat and Fig. 16 Secondary electron SEM fractograph of a
400-series aluminum casting alloy locking
featureless. However, engineering alloys are Fig. 15 Secondary electron SEM fractographs of tab on a wire cable pulley assembly that fractured during
polycrystalline, and they contain grain and fracture in 90 lb railroad track comprising a helicopter electrical cable pulling operation. (a) Lower-
ferrite-pearlite steel. The monotonic ductile rupture magnification micrograph. Original magnification: 490.
subgrain boundaries, inclusions, dislocations, process progressed by (a) transcrystalline cleavage (b) Higher-magnification micrograph showing void
and other imperfections that affect a propagat- through pearlite colonies, revealing (b) long, parallel formation at constituent particles in the casting. Original
ing cleavage fracture; thus, true, featureless dimples along the pearlite lamellae. magnification: 1510
cleavage is seldom observed. These imperfec- axis perpendicular to the boundary, a twist and higher steps. Thus, when viewing a cleavage
tions and changes in crystal lattice orientation boundary results (Fig. 18b). Because of the fracture that propagates across a twist boundary,
(i.e., grain texture), such as possible mismatch significant misalignment of cleavage planes at the cleavage steps do not cross but rather initiate
of the low-index planes across grain or sub- the boundary, the propagating fracture reinitiates new steps at the boundary (Fig. 18b). Most
grain boundaries, produce distinct cleavage at the boundary as a series of parallel cleavage boundaries, rather than being simple tilt or twist,
fracture-surface features. The distinct frac- fractures connected by small (low) cleavage are a combination of both types and are referred
ture-surface features include cleavage steps, steps. As the fracture propagates away from the to as tilt-twist boundaries. A cleavage fracture
river patterns, feather markings, chevron (her- boundary, the numerous cleavage planes join, exhibiting twist and tilt boundaries is shown in
ringbone) patterns, and tongues (Ref 45, 46). resulting in fewer individual cleavage planes Fig. 19.
A typical cleavage fracture viewed at differ-
ent magnifications is presented in Fig. 17
(Ref 47). As shown schematically in Fig. 18, Cleavage
cleavage fractures frequently initiate on many planes Fracture
parallel cleavage planes. As the fracture advances, direction
Tilt
however, the number of active planes decreases
by a joining process that forms progressively
higher cleavage steps. At the microscale, this
network of cleavage steps is known as a river
pattern (there are also river patterns at the
macroscopic scale). Because the branches of
the river pattern join in the direction of crack Grain or
propagation, these markings can be used to Cleavage steps subgrain boundary
establish the local fracture direction.
A tilt boundary exists when principal cleav- (a)
age planes form a small angle with respect to
one another as a result of a slight rotation
River pattern
about a common axis parallel to the intersec-
tion (Fig. 18a). In the case of a tilt boundary, Cleavage feathers Fracture
Twist direction
the cleavage fracture path is virtually uninter-
rupted, and the cleavage planes and steps prop- Cleavage
agate across the boundary. However, when the plane
principal cleavage planes are rotated about an
Grain or
Cleavage step subgrain boundary
(b)
Fig. 18 Schematic of cleavage fracture formation showing the effect of subgrain boundaries. (a) Tilt boundary.
(b) Twist boundary
Feather markings are a fan-shaped array of the general macroscopic propagation direction fractures were replicated in the laboratory
fine cleavage steps on a large cleavage facet of the crack. Such information can be useful (known as reference laboratory fractures).
(Fig. 19); they may also be called fine river for identifying the origin region of the fracture. In another example, transcrystalline cleavage
patterns. The apex of the fan points back to Cleavage fracture can occur in materials of fracture features were documented in a medium-
the fracture origin. Large cleavage steps are varying microstructures under a variety of load- low carbon steel casting that fractured in over-
shown in Fig. 19. ing conditions and environments. For example, stress, with cracks initiating at an internal casting
To support failure analysis, local crack cleavagelike fractures interspersed with inter- void, as demonstrated in Fig. 21 (Ref 49).
direction may be ascertained by studying the granular separations were observed on fractured
river pattern on each cleavage fracture facet. ultrahigh-strength alloy steel pins (Ref 48) due to
Crack direction mapping at the microscopic hydrogen-assisted cracking (Fig. 20, Ref 20). To Transcrystalline Fracture—Fatigue
level can, in aggregate, be used to ascertain confirm the fracture morphology, reference
A fracture that is the result of repetitive or
cyclic loading is known as a fatigue fracture
(Ref 50), as depicted in Fig. 22. Fatigue crack
initiation and growth in structures can often
occur at loads considered acceptable or safe
for the design and end use (Ref 51–53). Local
stresses on the component that concentrate at
singularities such as scratches, pores, cavities,
pits (corrosion), and nonmetallic inclusions
can influence the initiation of fatigue cracks.
The fatigue fracture diagram in Fig. 22 depicts
the three stages of fatigue fracture in a
Fig. 20 Secondary electron SEM images at original magnification of 870 (top) and 750 (bottom) of a helicopter
rotor control system drive pin revealing cleavagelike fractures interspersed with intergranular separations
and small regions of ductile dimples, consistent with hydrogen embrittlement. AerMet100 alloy, quenched and aged
to 53 HRC, exhibited a hydrogen content of approximately 9 ppm.
5 mm
5 mm
Fig. 22 (a) Conceptual example of a structural
Fig. 21 Secondary electron SEM images of the fracture surface of a failed snatch block hook revealing component that fractured after a fatigue
transcrystalline cleavage fracture features initiating at a casting void. The fracture in the 0.16 wt% C crack propagated to a critical size. (b) The relative
steel casting was due to overstress (63 HRBW average hardness—pearlite and proeutectoid ferrite microstructure). locations of stage 1, stage 2, and stage 3 fractures are
The inset image is a low-magnification montage image of the overall fracture surface of the hook. shown in the image.
cyclically loaded mechanical component: it cracks are faceted and sometimes resemble of 0.1 to 1.0 nm/cycle, suggesting that because
incubates and initiates during stage 1, propagates cleavage; they do not exhibit fatigue striation this distance is less than one atomic distance,
for most of its length during stage 2, and pro- features themselves. There are no clearly defined the crack does not grow with each stress rever-
ceeds to catastrophic overstress fracture during features on stage 1 fracture surfaces, but ridges sal (Ref 65). Stage 1 incubation and crack
stage 3, as evidenced by microfractography can occur with torsional loading or under corro- development are affected by the environment,
(Ref 54, 55). An example of a fractured recipro- sion fatigue conditions (Ref 52). The crack tends mean stresses (including residual surface
cating engine crankshaft exhibiting the three to follow crystallographic planes but changes stress), loading frequency, surface feature
stages of fatigue fracture is shown in Fig. 23 direction at discontinuities, such as grain bound- dimensions, and crystallographic and micro-
(Ref 56). Different cyclic load profiles may be aries. Stage 1 fatigue is normally observed on structure features.
used during fatigue testing, ranging from con- high-cycle, low-stress fractures and is frequently From a practical perspective, it has been
stant-amplitude tests to more complex spectrum absent in low-cycle, high-stress fatigue. At large generally observed that a stage 1 crack grows
(variable-amplitude) tests or block loading at plastic-strain amplitudes, fatigue cracks may ini- for approximately one to two grains (as
two or more loading amplitudes. tiate at grain boundaries (Ref 63). As much as demonstrated in Fig. 26 and 27). As the ratio
90% of the cumulative loading cycles prior to of tensile stress to shear stress increases, the
Stage 1—Fatigue Crack Incubation and failure may be consumed in initiating a viable crack may change direction (sometimes at a
Initiation fatigue crack (Ref 64). grain-boundary triple point) and follow along
Fundamental studies of the stage 1 initiation the maximum principal normal stress at the
Metal alloys used for technical and engi- process suggest that below approximately crack tip. At this transition, the crack is said
neering applications largely comprise poly- 100 Å, fatigue striations do not form; rather, to be stage 2 or crack propagation. A stage
crystalline grains. One can consider a single crack growth is nonuniform. Incremental crack 2 fracture surface may exhibit ridges asso-
crystal, in the form of one of the grains of a growth may not occur with every stress rever- ciated with crack opening and subsequent clos-
structure, that intersects the surface of a mechan- sal along the fracture front. It is believed that ing, known as striations. Generally, there is a
ical or structural component (Fig. 24b). Under the crack growth rate in stage 1 is on the order direct relationship between striation formation
dynamic loading conditions, fine dislocation slip
occurs along slip bands in the crystal, resulting in
Stage 2
ridge and groove formation, undulations, and Progressive
Cyclic
force
crack Stage 3
ultimately slip band intrusions and extrusions propagation
F (T )
Region
(Ref 54, 57–59) (Fig. 25). Through this process, of final
overstress
a crack develops at the slip band extrusion, Stage 1 fracture
resulting in a fatigue crack nucleus (Fig. 26). Crack
initiation
Crack
progression
Often, the active slip planes are those most region
marks
closely aligned with the maximum shearing
stress. With stress reversals, the crack will grow
(a) F (T )
in length along this slip plane during stage 1 ini- Cyclic
force
tiation. An example of stage 1 fatigue features is
shown in Fig. 27 (Ref 60, 61). Typically, stage 1
Part Part
surface Stage 3 surface
Monotonic
Stage 1 Stage 2 fracture
Incubation Progressive crack
Final
and initiation propagation
overstress
Intrusion Striation fracture
and extrusion formation
localized
plastic slip
steps
Fracture
direction
(b)
Stage 1 Stage 2 Stage 3
Fracture
Fracture
Component
Slip bands
surface
Force Force
Grain
boundaries
(a)
Slip steps
Fig. 26 Using the concept of Wood’s model, an incipient stage 1 fatigue crack can
develop to a length of one to two grains over n force reversals. Numerous
incipient stage 1 cracks can develop concurrently. On the part surface, the extrusions
can be discontinuous, tonguelike or continuous, ribbonlike.
Force Force
Slip bands
(b)
Extrusion Intrusion
(c)
100 Pm
Cyclic force Cyclic force
over n cycles over n cycles Fig. 27 Backscattered electron SEM image of a stage 1 fracture feature in a ŽS6K
Slip bands nickel-base superalloy. The crack initiation site is flat and faceted.
Ratchet marks and crack-arrest marks were observed emanating from the facet. The
(d) expanding pattern of ratchet marks was consistent with fatigue crack propagation
from the crack initiation site. Prior to SEM examination, the sample was cleaned of
deposits using a boiling aqueous potassium permanganate (KMnO4) and sodium
Fig. 25 Variation in Wood’s model to reveal how intrusions and extrusions of slip hydroxide (NaOH) solution. The arrows show the crack growth directions. Alloy
bands can result in sites for stage 1 fatigue crack initiation. Numerous sites ŽS6K is a nickel-base superalloy, manufactured in the former USSR, similar in
for stage 1 cracks may develop concurrently as force reversals progress. composition to Nimonic 80A. Source: Ref 62
and a single loading cycle (Ref 66). As the stage propagating crack. The size, type, and frequency
fatigue fracture that initiates at a subsurface
2 crack continues to grow with peak load of the artifact affect the stress-intensity factor.
anomaly or defect such as a massive inclusion.
cycling, there is a point at which the crack is of As an example, an aircraft high-pressure turbine
critical length. With one additional advance of stage 2 disk fractured because of low-cycle Stage 2—Fatigue Crack Propagation
the crack front, the strength of the remaining lig- fatigue cracks that initiated from an internal
ament is no longer sufficient, and the component subsurface manufacturing anomaly. In this The largest portion of a fatigue fracture con-
ruptures instantaneously in overstress (typically case, very close to the origin, the striation den- sists of stage 2 crack growth, which generally
exhibiting dimpled rupture or cleavage). sity was very high (6300 striations/mm), and occurs by transgranular fracture (Fig. 1b). Stage
Because of damage that frequently occurs at the the far-field striation density fell to approxi- 2 is more influenced by the magnitude of the
fatigue crack origin, it is often difficult to micro- mately 780 striations/mm, typical for higher alternating stress than by either the mean stress
scopically image features in the stage 1 region. alternating stress, low-cycle fatigue crack or material microstructure (Ref 74, 75). Fatigue
As an example, Fig. 27 (Ref 60) shows a flat, fac- propagation mechanisms (Ref 67). fracture surfaces generated during stage 2 fatigue
eted fatigue crack initiation site in alloy ŽS6K The work of Wanhill et al. suggests that usually exhibit striations as a result of the open-
(nickel-base superalloy). This feature was consis- fatigue crack initiation most often occurs at ing, advancing, and closing of the crack during
tent with stage 1 fatigue crack growth typical after singularities such as an artifact, flaw, or each stress cycle (Ref 76) (Fig. 31, Ref 77;
fatigue crack initiation in nickel-base superalloys. defect (Fig. 28, Ref 68; Fig. 29, Ref 69) on Fig 32, Ref 78). Striations are a visual record of
Ratchet marks and crack-arrest marks were the component surface, such as scratches, the position of the fatigue crack front during
observed emanating from the facet. The expanding pores, cavities, pits (corrosion) (Fig. 29), crack propagation through the material. Striation
pattern of radial marks was consistent with fatigue and nonmetallic inclusions (Fig. 30, Ref 70) spacing, measured from center to center of each
crack propagation from the crack initiation site. (Ref 71–73). Within this artifact, an intru- striation, can range from 0.001 to approximately
Perhaps one of the more opportunistic ways sion-extrusion mechanism occurs along active 1 mm under high-cycle fatigue spectrum-loading
to observe stage 1 fracture features is from a slip planes, ultimately resulting in a conditions (Fig. 33, Ref 14).
1 mm
Fig. 28 Secondary electron SEM fractograph of a fatigue-initiated fracture in a tempered steel valve spring from a
reciprocating internal combustion engine. The fatigue fracture initiated at a pit on the surface of the steel
wire on the inside diameter of the spring.
Fig. 30 Secondary electron SEM fractographs of fatigue crack initiation at a nonmetallic inclusion in an airplane
wing spar fabricated from aluminum alloy 6061-T651
the mean stress or the cyclic frequency. Increas- another (Fig. 34). A plateau that has a concave closure (Ref 35, 82–84) (Fig. 36). This loading
ing the magnitude of the alternating stress pro- surface curvature exhibits a convex contour on scenario results in a compressive stress at the
duces a corresponding increase of the striation the mating fracture face (Ref 81). The plateaus crack tip due to the relaxation of the residual
spacing. An increase in mean stress can also are joined either by tear ridges or walls that elastic tensile stresses induced in the uncracked
increase the striation spacing, although not to contain fatigue striations. Fatigue striations portion of the material during the rising load
the extent a numerically equivalent increase in often bow outward in the direction of crack cycle (Fig. 36). The closing crack does not
the alternating stress would. Within reasonable propagation and generally tend to align perpen- reweld, because the new slip surfaces created
limits, the cyclic frequency has the least effect dicular to the principal (macroscopic) crack during the crack-opening displacement are
on striation spacing. In some cases, fatigue stria- propagation direction. However, variations in instantaneously oxidized (Ref 85), which makes
tion spacing can change significantly over a very local stresses and microstructure can change complete slip reversal unlikely in atmosphere.
short distance. This is due in part to changes in the orientation of the plane of fracture and alter This model also accounts for conditions where
local stress conditions as the crack propagates. the direction of striation alignment (Fig. 35). slip may not occur precisely at the crack tip due
Typically, each striation is the result of one The way in which the crack closes can affect to the presence of lattice or microstructural
load cycle and marks the position of the the appearance of the striation features. Some imperfections (Ref 83, 86, 87).
fatigue crack front at the time the striation of the more notable crack-closure scenarios The essential absence of striations on fatigue
was formed. However, when there is a sudden include: fracture surfaces of metals tested in vacuum
decrease in the applied load, the crack can tends to support the assumption that oxidation
No closure
temporarily stop propagating, and no striations reduces slip reversal during crack closure,
Plasticity-induced closure
are formed. The crack resumes propagation which results in the formation of striations
Roughness-induced (asperity-induced) closure
only after a certain number of cycles are (Ref 79, 88, 89). The lack of oxidation in high
Mode-2-induced closure
applied at the lower stress (Ref 26, 76, 77, vacuum promotes a more complete slip rever-
Oxide-induced closure
79, 80). This phenomenon of intermittent crack sal (Ref 90), which results in a smooth and
Phase-transformation-induced closure
arrest is believed to be due to the presence of a relatively featureless fatigue fracture surface.
residual compressive-stress field within the Some fracture surfaces containing widely
crack tip plastic zone produced after the last In ductile metals, stage 2 fatigue crack spaced fatigue striations exhibit slip traces on
high-stress fatigue cycle (Ref 76, 77). growth propagates by the incremental genera- the leading edges of the striation and relatively
tion of a new fracture surface during crack smooth trailing edges, as predicted by the
Characteristics of Fractures with Fatigue opening and blunting and then resharpening model in Fig. 36. However, not all fatigue
Striation Features during crack closure. There have been variants striations exhibit distinct slip traces, as sug-
of this general concept, but they are based on gested by Fig. 36, which is a simplified repre-
During Stage 2 fatigue fracture, the crack localized plastic flow crystallographic slip dur- sentation of the fatigue process.
often propagates on multiple plateaus that are ing crack opening and subsequent reversed Cracks that propagate by stage 2 fatigue
at different elevations with respect to one shearing of the activated slip bands during crack fracture often produce striated features on the
fracture surface as a result of the opening,
advancing, and closing of the crack during
each stress cycle (Ref 73). The striation
spacing, related to the crack propagation rate
(da/dN), can be correlated with the amplitude
of the stress cycle and the length of the crack
through the stress-intensity factor range (DK)
(Fig. 37). As shown in Fig. 37, each stage of
a fatigue fracture can be represented on a plot
of da/dN versus DK. However, stage 2 can
most reliably be correlated to DK through an
exponential relationship, generally referred to
as the Paris-Erdogan region (Ref 91). If the
stress amplitude is approximately the same
10 Pm 2 Pm
(a) (b)
Macroscopic
fracture Secondary crack Fatigue striations
5
propagation
direction 6 10
11
12 13
3
9
4
2
7
8
1
Secondary
Ridges crack
(c)
Fig. 35 Secondary electron SEM fractograph of the
Fig. 34 (a) and (b) Secondary electron SEM fractographs of a fatigue fracture in an M-flange in an aircraft high- turbine engine diffuser case in Fig. 34.
pressure turbine case that propagated by cyclic fatigue (wrought nickel-base superalloy, such as IN- Variations in local stresses and microstructure can
718). Each striation was associated with the change in burner pressure during engine startup, takeoff, and engine change the orientation of the plane of fracture and alter
shutdown, resulting in one striation per flight cycle. Note that the striations are on plateaus. (c) Schematic example the direction of striation alignment. The yellow arrows
of fatigue fracture plateaus with concomitant striations indicate the local direction of crack propagation.
Increment
Formation of of crack Stage 1 Stage 2 Stage 3
a striation growth due Crack initiation— Progressive crack propagation—
to one slip plane crack striated crack growth
dN
da
Mating striations Crack tip loading
cycle 1 loading cycle 1 1 loading cycle = 1 striation
Closing crack Fig. 37 Hypothetical graph showing the relationship between the fatigue crack
propagation rate as a function of the cyclic stress-intensity factor
Loading cycle, n
Very-
Closed crack High- high-
Low- cycle cycle
New striation Ultra- cycle fatigue fatigue
low- fatigue
Fig. 36 Sketches showing the progressive mechanism of fatigue crack propagation cycle
by alternate slip at the crack tip and the formation of a mating pair of fatigue
striations. Da, crack growth increment
Fig. 38 Relative range of ultralow-cycle fatigue, low-
cycle fatigue, high-cycle fatigue, and very-
aircraft can be well described with the GAG high-cycle fatigue
from cycle to cycle, the spacing of consecutive
striations should be nominally uniform (Ref profile, those of helicopters and military air-
35, 74, 92) (i.e., within the field of view of craft are subject to applied load spectra that
an SEM image). When performing crack are complex and highly variable. Traditional
growth rate studies, it is often assumed that fatigue-prediction tools do not always perform
one striation feature on the fracture surface is well in calculating the fatigue lives of systems
correlated with one peak loading cycle. As an under these conditions (Ref 101). Fatigue stri-
example, for airplane turbine engine compo- ation examination of spectrum-loaded frac-
nents that fracture in fatigue, each stage 2 stria- tures often undercounts the total number of
tion feature generally correlates to one takeoff striations, leading to an overestimate of the
cycle, where the engine is producing maxi- part fatigue life. Care and conservative safety
mum power (Ref 93). In another example, factors should be used if designing components
fatigue fracture striation features in pressurized that may be subjected to spectrum-loading
airplane fuselage structures often correlate to profiles and additional load amplitudes.
one ground-air-ground (GAG) cycle (fuselage The number of cycles that a component fab-
is pressurized at altitude and depressurized on ricated from a given material can endure until
landing). There are many articles that explain final fracture is directly related to the magni- Fig. 39 Secondary electron SEM fractograph of a
low-cycle fatigue fracture surface on a
how to develop fatigue crack growth rates tude of the stress reversal and the mean stress. fractured aircraft turbine engine fan blade fabricated
based on quantitative fractography and stria- Such fatigue master curves have been devel- from wrought alloy Ti-8Al-1Mo-1V
tion spacing data; hence, it is often the focal oped for design purposes. Conceptually, high
point of fatigue fracture fractography (Ref 61, stress amplitude and a zero mean stress (stress
94–100). ratio R = –1) can result in a low number of features such as ductile tear ridges or patches
It should be noted that while fatigue analysis cycles to final fracture, referred to as low-cycle of microvoid coalescence. In LCF, if the stri-
using observed striations can be useful, many fatigue (LCF). As shown in Fig. 38, the num- ation features can be resolved by SEM, a plot
components and systems do not undergo a sim- ber of cycles to final fracture can be classified of striation spacing as a function of distance
ple uniformly variable (e.g., sinusoidal) load- over various ranges. The fracture surfaces from the origin region will generally become
ing profile. These components may be subject can yield clues about the severity of the load- progressively larger. Conversely, high-cycle
to additional underlying lower-amplitude ing. Low-cycle fatigue fracture surfaces, for fatigue can have relatively smaller striation
cyclic stresses, such as from vibrations, exter- example, will exhibit striation features with spacing and less ductile tearing, as shown in
nal winds, and natural resonances, as well as relatively long spacing distance, such as Fig. 31 to 33.
erratic spectrum-loading profiles. While load approximately 2 mm striation spacing, as In components that are thermally cycled from
cycles, for example, of commercial civilian demonstrated in Fig. 39 (Ref 102), and other ambient to high temperatures, such as turbine
engine blades and surrounding components and Intercrystalline (Intergranular) deformation). Creep fracture is an exception
boiler tubes in energy production, thermal Fracture to this observation, because it often exhibits
strains may compressively yield the metal at both intergranular fracture and substantial
high temperature. Upon cooling and contraction deformation.
to original dimensions, the component may Polycrystalline metals and alloys can frac- There are numerous mechanisms that can
experience tensile stress. Repetitive heating and ture in a manner whereby the crack propagates occur under suitable environments, sensitive
cooling can lead to a failure mode known as ther- along grain boundaries (Fig. 1c). Such a frac- metallurgical conditions, or stress conditions
mal fatigue (TF) or thermomechanical fatigue ture phenomenon is synonymously called that result in the decohesion of the grain
(TMF), if ancillary mechanical loads are present intercrystalline fracture, intergranular fracture, boundaries under mechanical stress (Table 3).
(Ref 103). Thermal fatigue and TMF often prop- intergranular decohesion, or grain-boundary Some metals, intermetallics, and alloys inher-
agate as LCF. It is possible that an initial TF or decohesion. The influence of microstructure ently have a limited number of slip systems,
TMF crack can act as a precrack for high-cycle can affect the appearance of intercrystalline and intercrystalline fracture is a normal frac-
fatigue crack propagation in the component. fracture surfaces. For example, when an inter- ture mode. A number of metals and alloys
Often there is concomitant oxidation scale on crystalline fracture occurs along flattened, elon- have a propensity to form films or precipitates
the TF or TMF fracture surface. gated grains that form nearly uninterrupted along grain boundaries. These films can form
For components with precision surface treat- planes through the material (as in severely as a result of heat treating processes or per-
ments and highly controlled microstructures extruded alloys, deep-drawn parts, and along haps with aging at elevated temperatures.
(with consideration for the distribution and the parting planes of some forgings), a relatively Regardless, the films and precipitates at the
size of second-phase particles and inclusions), smooth, featureless fracture results (Fig. 1e). grain boundaries create a favorable crack
fatigue life can approach 108 to 1010 cycles, Intercrystalline fracture has occurred under path. Examples of grain-boundary films and
known as very-high-cycle fatigue (Ref 104). loading conditions such as: precipitates include sensitization in austenitic
In high-strength steels, fatigue crack initiation Monotonically increasing load to fracture stainless steels or improperly aged aluminum
often occurs at subsurface inclusions in a (overstress fracture) alloys (Fig. 1d).
region termed the fine granular area (FGA). Impact or high strain rate Regardless of the mechanism, the fracture
The formation mechanism of the FGA is still Creep (constant or varying load at elevated surface will exhibit some degree of faceted
under technical investigation and has not been temperature > 0.4Tm over time) planes along the boundaries between grains.
settled (Ref 105). Steady-state stress substantially below the Figure 41 (Ref 19) shows an intercrystalline
material failure limit (environmentally fracture surface. In this case, a medium-carbon
Stage 3—Final Fracture assisted fracture) grade 8 bolt fractured due to hydrogen-
Alternating stress (fatigue fracture) assisted cracking. Because the steel is
Stage 3 is the terminal propagation phase of a quenched and tempered to 38 HRC, the inter-
fatigue crack in which the striation-forming In failure analysis, intercrystalline fracture crystalline fracture follows along the prior-
mode is progressively displaced by monotonic is typically indicative of a low-toughness or austenitic grain boundaries of the tempered
fracture modes, such as microvoid coalescence, low-energy fracture process. In most instances, martensite microstructure.
transcrystalline cleavage, or intercrystalline components that fracture in an intercrystalline Intercrystalline fracture is typically an indica-
decohesion, depending on the alloy and the cir- manner exhibit little, if any, macroductility tion of an environmental reaction or processing
cumstances of the separation. The rate of crack (i.e., such as necking or gross mechanical issue. Decohesion may occur along grain
growth increases during the transition to stage
3 until the fatigue crack becomes unstable and
the part separates or otherwise fails. Because Table 3 Common metallurgical mechanisms of intercrystalline fracture
the crack propagation is increasingly dominated Inherent low number of slip systems, material naturally fractures intergranularly
by the monotonic fracture mode, stage 3 fatigue Grain-boundary films and precipitates (such as sensitization in austenitic stainless steels or improperly aged aluminum
is sensitive to both microstructure and mean alloys)
Metallurgical embrittlement
stress (Ref 75, 106). A micrograph showing the
a. Strain aging in steels
transition from stage 2 to stage 3 is shown in b. Quench-hardened steel
Fig. 40 (Ref 107). i. Blue brittleness
ii. Temper embrittlement
iii. Quench cracking
iv. Carbonitrided and carburized cases
c. Stainless steels
i. Thermal embrittlement in maraging steels
ii. 475 C embrittlement in ferritic stainless steels
iii. Sigma-phase embrittlement
iv. 350 C embrittlement in martensitic stainless steels
Nitrided and ferritic nitrocarburized steels (surface effect)
Topologically close-packed phase embrittlement (nickel-base alloys)
Service environmental factors
Environmentally assisted sustained-peak low-cycle fatigue
Hydrogen-assisted cracking
Nitrogen embrittlement
Liquid metal embrittlement, liquid-metal-assisted cracking
Neutron irradiation embrittlement
Stress-corrosion cracking
Intergranular corrosion (such as in sensitized austenitic stainless steels)
Temperature (low-temperature impact)
Rupture of melted and resolidified grain-boundary constituents (as in overheated aluminum alloys)
Fig. 40 Secondary electron SEM fractograph showing Grain-boundary liquation
the transition from stage 2 fatigue fracture to Fractures > 0.4Tm (creep)
stage 3 monotonic fracture (overstress) in a turbine Hot cracking, separation of melted material in the boundaries before it solidifies, as in cracking at the heat-affected zones
engine diffuser case wall (fabricated from a nickel-base of welds
high-temperature superalloy). Original magnification: Dezincification and other dealloying processes
200
Fig. 41 Secondary electron SEM fractograph showing intercrystalline fracture along prior-austenitic grain Fig. 43 Secondary electron SEM fractograph of
boundaries due to hydrogen-assisted cracking in a grade 8 zinc-electroplated crankshaft bolt. The bolt intercrystalline fracture due to the
was quenched-and-tempered martensite with a core hardness of 38 HRC. (a) Fracture surface along prior-austenitic application of force under conditions of incipient melting
grain facets. (b) Fracture surface showing mixed microvoid coalescence and intercrystalline fracture. (c) in aluminum alloy 2024-T6 after exposure to a fire
Intercrystalline fracture adjacent to the fractured zinc-electroplated coating
boundaries for several reasons: second-phase This includes creep deformation and rupture reference fatigue life in a controlled environ-
particles or films at grain boundaries, constituent from high-temperature exposure and various ment such as dry air or vacuum) is designated,
composition segregation at grain boundaries types of environmentally assisted cracking, such in practice, as environmentally assisted fatigue
(Ref 108), and/or environmental influence. as hydrogen embrittlement, stress-corrosion fracture or corrosion fatigue. The specific
Figure 42 shows an intercrystalline fracture with cracking, sulfide-stress cracking, liquid-metal- mechanisms at play are highly dependent on
microvoid coalescence ornamentation on the induced embrittlement, corrosion fatigue, the variables at hand (Ref 111).
grain facets, indicating that, in some instances, and thermomechanical fatigue. Environmen- The objective of microfractography is to
localized plasticity is involved rather than just tally assisted cracking can be mechanistically identify fractographic features that correspond
decohesion (Ref 109). complex, depending on the material, loading to a specific fracture micromode. With
In some instances, fracture due to the appli- conditions, environment, and microstructural environmentally assisted fatigue fracture, often
cation of force under conditions of incipient conditions. More details on environmentally the damage from a corrosive environment
melting can result in an intercrystalline frac- assisted crack initiation and growth are provided removes the fracture microfeatures (Ref 112).
ture surface, as shown in Fig. 43 (Ref 110). in the subsequent articles “Fracture Mechan- Hence, the evidence of interest may be
Sometimes this morphology can be used to isms” and “Factors Influencing Fractures and contained within the corrosion products or
confirm a minimum temperature that compo- Fracture Appearance” in this Volume. may be lost due to dissolution or sloughing
nents reached in a fire. of the corrosion product. Often, other circum-
In most instances, testing and the consider- Environmentally Assisted Fatigue stantial data must be considered for a diagnosis
ation of other factors leading to the fracture
are necessary to confirm the specific cause of
Fracture of environmentally assisted fatigue fracture.
Figure 44(a) (Ref 113) shows subtle fatigue
intergranular decohesion. For example, in liq- The fatigue fracture incubation and initia- fracture striation features with a thin surface
uid metal embrittlement, a thin film of a lower- tion processes are very sensitive to the compo- layer of corrosion product. After removing
melting-point metal is typically present on the nent surface condition and the enveloping the corrosion product with a chelating rust
fractured grain facets (generally confirmed by environment. Stage 1 extrusion-intrusion pro- remover, the remnants of fine striations can
microscopic chemical-analytical methods such cesses, demonstrated in Fig. 25 and 26, can be observed (Fig. 44b). Along with the micro-
as energy-dispersive spectroscopy). be affected by environmental exposure. For a fractography, compositional analysis of the
given surface condition, the service environ- scale, metallurgical cross-sectional analysis
ment can alter the number of cumulative load- at the origin, and other macrofracture mor-
Environmental Factors ing cycles that a component can withstand phology considerations were required to iden-
before final fracture. The reduction in cumula- tify the environmentally assisted fatigue
The environment can have important effects tive loading cycles to failure as a result of fracture mode. In some instances, striation
on crack initiation and crack growth behavior. environmental exposure (compared to a features are replaced with other features, as
Fig. 45 Secondary electron SEM fractograph of an Fig. 46 (a) Secondary electron SEM fractograph of a creep-induced fracture surface in an Inconel 100 compressor
environmentally assisted fatigue crack in an turbine engine blade, revealing intercrystalline and interdendritic fracture features. Original magnification:
aluminum alloy propeller that exhibited flutinglike 14.5. (b) Optical metallographic cross section through the fracture surface, revealing surface oxide scale and a
features. (a) Original magnification: 150 (b) Original reaction zone (outlined by the blue dashed line). Bright-field illumination; electrolytic phosphoric acid etch. Original
magnification: 500 magnification: 230. (c) Higher-magnification optical micrograph of reaction zone in (b). Original magnification: 1200
Hydrogen-Assisted Cracking In some hydrogen-embrittled steels, the corroded as the SCC process progresses
fracture micromode can change from dimpled (Fig. 50, Ref 147; Fig. 51, Ref 148, 149). Often
When certain body-centered cubic, body- rupture to mixed-mode cleavage, cleavage, or the presence of corrosion products on the frac-
centered tetragonal, and hexagonal close- intergranular decohesion (most common). ture surfaces, along with the analysis of corro-
packed metals or alloys of such elements as These changes in fracture mode or appearance dent salts and residue, provides the basis for the
iron (steel), nickel, titanium, vanadium, tanta- may not occur over the entire fracture surface designation of an SCC mechanism (Fig. 52,
lum, niobium, zirconium, and hafnium are and are usually more evident in the region of Ref 150). Fractography is typically accompanied
exposed to hydrogen, they are susceptible to the fracture origin. Figure 48 shows an exam- by cross-sectional metallography to identify an
a type of failure known as hydrogen-assisted ple of a hydrogen-embrittled AISI 4340 steel intergranular or transgranular fracture morphol-
cracking or hydrogen embrittlement (Ref 121). that exhibits cleavage fracture and secondary ogy. There is consensus among the technical
Although the face-centered cubic metals and intercrystalline decohesion. community that there are a variety of different
alloys are generally considered to have good mechanisms for different material systems and
resistance to hydrogen embrittlement, it has been Stress-Corrosion Cracking environments. Such mechanisms include stress-
shown that the type 300-series austenitic stain- assisted directed dissolution, selective dissolution
less steels (Ref 122–125) and certain type The concomitant effects of stress and envi- vacancy creep, film-induced cleavage, corro-
2000- and 7000-series high-strength aluminum ronment on a susceptible material can lead to sion-enhanced localized plasticity, hydrogen-
alloys are also embrittled by hydrogen crack incubation (Ref 143), initiation, and enhanced localized plasticity (and other hydro-
(Ref 126–132). Although the result of hydrogen propagation in a brittle manner at stresses gen effects), and adsorption-induced dislocation
embrittlement is generally perceived to be a cat- below the ultimate strength of the metal. This emission (Ref 141, 142, 151–153).
astrophic fracture that occurs well below the ulti- failure mode is regarded as stress-corrosion
mate strength of the material and exhibits little cracking (SCC) or environmentally assisted Liquid Metal Embrittlement
ductility, the effects of hydrogen can be quite cracking (EAC). Such a failure mode was first
varied (Ref 133). They can range from a slight chronicled in approximately 1875 (Ref 121) In the case of liquid metal embrittlement
decrease in the percent reduction of area at frac- and later with cracking in copper alloys (LME), many university studies have been
ture to premature rupture that exhibits no ductil- (Ref 144). The SCC and EAC failures generally
ity (plastic deformation) and occurs at a manifest with brittle fracture features such as
relatively low applied stress compared to the intercrystalline or transcrystalline cleavage in
failure strength. materials that otherwise would exhibit ductility
The source of hydrogen may be a processing upon fracture. Generally, cracks progress
operation, such as electroplating (Fig. 41) or through the material, driven by stresses (applied
acid pickling, or the hydrogen may be acquired or residual stresses) below the ultimate strength
from the environment in which the part oper- until a critical crack length is reached and other
ates (for example, a corrosive environment or fracture modes, such as overstress or fatigue,
under cathodic protection) (Ref 134). take over (Ref 145) (Fig. 49, Ref 146). Macro-
Some of the prevailing mechanisms for scopically, SCC can exhibit crack-arrest marks,
hydrogen-assisted cracking in steels include making it similar in appearance to other progres-
hydrogen-induced decohesion (Ref 135, 136), sive cracking fracture surfaces, such as fatigue.
hydrogen-enhanced local plasticity (Ref 137), However, microscopically the fracture features
hydrogen-induced phase transformation (Ref 138), are distinct. It is a common observation that
hydrogen-enhanced strain-induced vacancy for- SCC fractures exhibit branched cracking, which
mation (Ref 139), and adsorption-induced dislo- may be documented on the fracture surface or in
cation emission (Ref 140). Probably no one metallurgical cross sections through the fracture. Fig. 49 Secondary electron SEM fractograph of a
fractured commercial airplane landing
mechanism is applicable to all metals, and sev- Depending on the SCC mechanism, the gear. The gear is fabricated from quenched-and-
eral mechanisms may operate simultaneously branched cracks may be intergranular or trans- tempered type 300M steel. The fractograph shows the
to embrittle a metal (Ref 141, 142). Whatever granular (see the article “Fracture Mechanisms” transition from stress-corrosion cracking (SCC) to
the mechanism, the end result is an adverse fatigue to final overstress fracture once a critical crack
in this Volume for more details). length was obtained. The SCC region shows well-
effect on the mechanical properties or load-car- The challenge of SCC fractography is that, defined intercrystalline facets with some microvoid
rying capacity of the metal. in some instances, the fracture surface is coalescence decoration.
Fig. 48 Cleavage fracture with secondary intercrystalline Fig. 50 Secondary electron SEM fractograph showing
Fig. 47 Secondary electron SEM fractograph of intercrystalline fracture features along with
intergranular fissures in an Inconel 100 cracks due to delayed hydrogen-induced microvoid coalescence features typical of stress-corrosion
compressor turbine engine blade that fractured under cracking in AerMet100 alloy quench hardened and cracking of a type 1095 steel spiral spring from an
creep conditions tempered to 53 HRC airplane magneto exposed to condensed moisture
Transgranular
Thermomechanical Fatigue (Ref 159) master rod is shown in Fig. 54 (Ref 160). In a fracture surface from an American Associa-
instances where both halves of the fracture have tion of Railroads (AAR) TC-128 grade B rail-
Thermomechanical fatigue refers to fatigue been retained, it is useful to document the macro- road tank car steel that exhibits elongated
under simultaneous changes in temperature fractographic features on both halves. Some- dimples associated with long manganese sul-
and mechanical strain. Thermomechanical times one half will retain features that have fide stringers present in the steel fracture.
fatigue encompasses several mechanisms in been damaged on the other half. However, in Microstructural artifacts such as pearlite
addition to “pure” fatigue damage, including failure analysis, often only one half of the frac- lamellae can be mistaken for fatigue striations
high-temperature creep and oxidation, which ture is retained or recovered. For example, when or dimpled rupture features, as demonstrated
directly contribute to damage. These mechan- the tip of an airplane propeller fractures, centrip- in Fig. 15 and 56.
isms differ, depending on the strain-temperature etal forces send the tip half flying into the atmo-
history. They are different from those predicted sphere, such that only one half of the fracture
by creep tests (with no reversals) and by stress- will be retained. Corrosion Damage
free (or constant-stress) oxidation tests. Micro- In fatigue, short, tight cracks can have frac-
structural degradation can occur under TMF in ture faces that remain in close contact over Corrosion damage and subsequent removal
the form of: many stress cycles, particularly in instances of corrosion product from fracture surfaces
of low-stress, high-cycle fatigue. In such cases, can obscure or remove the original fracto-
Overaging, such as coarsening of precipi- graphic features. Sometimes vestiges of the
the fracture faces can be quite polished, and
tates or lamellae fine fracture features are no longer present. original fracture features can be faintly
Strain aging in the case of solute-hardened observed during a careful stepwise or iterative
These scenarios are the plight of the fractogra-
systems pher, motivating them to perform fracture path corrosion product removal process. Figure 44
Precipitation of second-phase particles shows an example of fatigue striation that
mapping and computer modeling to help
Phase transformation within the tempera- had been obscured under layers of corrosion
understand the root cause of the failure.
ture limits of the cycle product. In another example, fractographic
analysis of a crude oil pipeline fracture
Local stresses and cracking also may result Microstructural Influences on revealed that the fracture initiated as environ-
from variations in mechanical properties or Microfracture Mechanisms mentally assisted cracks, which, upon blunting
thermal expansion coefficients between the just below the outside surface, transitioned to
matrix and strengthening particles. Inclusions, grain flow, and other attributes corrosion fatigue cracks that grew and coa-
Crack initiation within nickel-base superalloys of a microstructure can influence the fracture lesced (Ref 161, 162). The crude oil pipeline
can occur intergranularly at oxidized surface- micromorphology. For example, Fig. 55 shows fracture surface in Fig. 57 (Ref 163) shows
exposed grain boundaries or transgranularly.
Transgranular initiation can be caused by hetero-
geneous planar slip, which produces initiation
along persistent slip bands at free surfaces. Trans-
granular initiation can also occur at pores, in-
clusions, and carbides. Transgranular crack
initiation is more prominent at low temperatures Pearlite
and high frequencies. This is because contribu- colonies
tions from the creep and environmental compo-
nents of damage are minimal. The interaction of
creep damage and environmental attack and their
Fine dimple rupture
effects on crack propagation under TMF condi-
tions can be very complicated. In general, trans- 2 mm
granular crack propagation occurs with out-of-
phase creep-fatigue loading, while intergranular Fig. 56 Secondary electron SEM fractograph of a
Fig. 54 Secondary electron SEM fractograph of a steel fracture showing fracture features
fracture is observed for in-phase TMF loading fractured connecting rod from an aircraft through pearlite colonies compared with fine dimpled
within representative service temperatures at reciprocating engine. The rod is fabricated from rupture
moderate strain rates (10–3 to 10–6/s). quenched-and-tempered alloy steel, and the fracture
surface exhibits mechanical damage due to smearing.
and F. Zakar, The Efficacy of a Chelating Fractography, STP 436, American of a Single Crystal: Scanning Tunneling
Agent-Based Oxide Remover to Clean Society for Testing and Materials, Microscopy and Spectroscopy Study, Surf.
Corrosion Products from Ferrous Frac- 1968, p 151–178 Sci., Vol 490, 2001, p L573–578
ture Surfaces, J. Fail. Anal. Prevent., 27. C.D. Beachem and D.A. Meyn, in Elec- 44. Electron Fractography Handbook, MCIC-
Vol 19, 2019, p 1860–1865 tron Fractography, STP 436, American HB-06, Metals and Ceramics Information
10. H.M. Westergaard, Bearing Pressures Society for Testing and Materials, 1968, Center, Battelle, Columbus, OH, 1975
and Cracks, J. Appl. Mech., Vol 61, p 59 45. J. Friedel, in Fracture: Proceedings of
1939, p A49–A53 28. S.P. Lynch, Ductile and Brittle Crack the Swampscott Conference, MIT Press,
11. G.R. Irwin, Handbuch der Physik, Vol 6, Growth: Fractography, Mechanisms and 1959, p 498
Flugge Ed., Springer, Berlin, 1958, Criteria, Mater. Forum, Vol 11, 1988, 46. C.D. Beachem, Microscopic Fracture
p 558–590 p 268–283 Processes, Fracture, An Advanced Trea-
12. S.A. Silling, Reformulation of Elasticity 29. D.A. Meyn and E.J. Brooks, Microstruc- tise: Microscopic and Macroscopic Fun-
Theory for Discontinuities and Long- tural Origin of Flutes and Their Use in damentals, Vol 1, H. Liebowitz, Ed.,
Range Forces, J. Mech. Phys. Solids, Distinguishing Striationless Fatigue from Academic Press, New York, 1968
Vol 48, 2000, p 175–209 Stress-Corrosion Cracking in Titanium 47. J. Epperson, “Accident Docket DCA1
13. E. Madenci and E. Oterkus, Peridynamic Alloys, Fractography and Materials Sci- 6SH002,” Materials Laboratory Factual
Theory and Its Applications, Springer, ence, STP 733, L.N. Gilbertson and R.D. Report 17-001, National Transportation
New York, NY, 2014 Zipp, Ed., American Society for Testing Safety Board, Washington, D.C., 2017
14. M.K. Budinski, “Accident Docket WPR1 and Materials, 1981, p 5–31 48. P.M. Novotny, An Aging Study of Car-
6FA055,” Materials Laboratory Factual 30. I. Aitchison and B. Cox, Corrosion, Vol penter AerMet 100 Alloy, SPEICH Sym-
Report 16-023, National Transportation 28, 1972, p 83 posium Proceedings, 1992
Safety Board, Washington, D.C., 2016 31. J. Spurrier and J.C. Scully, Corrosion, 49. M. Budinski, “Accident Docket DCA18S
15. E. Mueller, “Accident Docket DCA14 Vol 28, 1972, p 453 M016,” Materials Laboratory Factual
MR004,” Materials Laboratory Fac- 32. D.B. Knorr and R.M. Pelloux, Metall. Report 18-058, National Transportation
tual Report 14-055, National Trans- Trans. A, Vol 13, 1975, p 73 Safety Board, Washington, D.C., 2018
portation Safety Board, Washington, 33. R.J.H. Wanhill, Corrosion, Vol 32, 1976, 50. J. Glynn, On the Causes of Fracture of
D.C., 2014 p 163 the Axles of Railway Carriages, Minutes
16. X. Liu, “Accident Docket DCA14 34. H. Hänninen and T. Hakkarainen, Metall. Proc. Inst. Civ. Eng., Vol 3, London,
MR004,” Finite Element Modeling Study Trans. A, Vol 10, 1979, p 1196–1199 1844, p 202–203
Report, National Transportation Safety 35. C. Laird, The Influence of Metallurgical 51. H.J. Gough, The Fatigue of Metals, Ern-
Board, Washington, D.C., 2014 Structure on the Mechanisms of Fatigue est Benn Ltd., London, 1926
17. E. Mueller and X. Liu, Failure Analysis Crack Propagation, Fatigue Crack Prop- 52. R. Cazaud, Fatigue of Metals, 1948;
and Finite Element Modelling of a Rail agation, J. Grosskreutz, Ed., American Translated by A.J. Fenner, Chapman &
Axle Fatigue Fracture, Int. J. Rail. Tech- Society for Testing and Materials, West Hall, Ltd., London, 1953
nol., Vol 7 (No. 1), 2018 Conshohocken, PA, 1967, p 131–180 53. J.A. Collins, Failure of Materials in
18. E. Mueller, “Accident Docket ERA20 36. R.D. Buchheit and D. Broek, “Microstruc- Mechanical Design: Analysis, Predic-
LA055,” Materials Laboratory Factual tural and Fractographic Characterizations tion, Prevention, John Wiley and Sons,
Report 20-010, National Transportation of Fatigue Tested Rail Steels,” Final 1981, p 164–499
Safety Board, Washington, D.C., 2020 Report DOT-TSC, Oct 1977, p 48–80 54. P.J.E. Forsyth, Fatigue Damage and
19. W.R. Rossey, Jr., “Accident Docket IAD02 37. U. Lindborg, Morphology of Brittle Frac- Crack Growth in Aluminum Alloys, Acta
FA091,” Materials Laboratory Factual ture in Pearlite, Trans. ASM, Vol 3, Sept Metall., Vol 11, July 1963, p 703–714
Report 03-016, National Transportation 1964, p 500–504 55. C.A. Zapffe and C.O. Worden, Fracto-
Safety Board, Washington, D.C., 2003 38. E. Mueller, “Accident Docket DCA17 graphic Registrations of Fatigue, Vol
20. J. Epperson, “Accident Docket CEN10 MR007,” Materials Laboratory Factual 43, Trans. ASM, 1951, p 958–969
FA291,” Materials Laboratory Factual Report 17-050, National Transportation 56. M. Fox, “Accident Docket CEN19
Report 11-013, National Transportation Safety Board, Washington, D.C., 2017 LA055,” Materials Laboratory Factual
Safety Board, Washington, D.C., 2010 39. D.C. Grundy, “Fatigue and Fracture of a Report 20-013, National Transportation
21. C. Schultheisz, “Accident Docket Railway Wheel Steel,” Master of Science Safety Board, Washington, D.C., 2009
DFW06 LA094,” Materials Laboratory thesis, Massachusetts Institute of Tech- 57. W.A. Wood and A.K. Head, J. Inst. Met.,
Factual Report 07-019, National Trans- nology, May 1994, p 74 Vol 79, 1950, p 89
portation Safety Board, Washington, D. 40. S. Hamada, D. Sasaki, M. Ueda, and H. 58. W.A. Wood, in Fracture: Proceedings of
C., 2007 Noguchi, Fatigue Limit Evaluation Con- the Swampscott Conference, MIT Press,
22. C.D. Beachem, Metall. Trans. A, Vol 6, sidering Crack Initiation for Lamellar 1959, p 412
1975, p 377–383 Pearlitic Steel, Proc. Eng., Vol 10, 59. C. Laird and G.C. Smith, Philos. Mag.,
23. I. Le May, Metallography in Failure 2011, p 1467–1472 Vol 7, 1962, p 847
Analysis, J.L. McCall and P.M. French, 41. E. Mueller, “Accident Docket ERA18 60. E. Mueller, “Accident Docket WPR1
Ed., American Society for Metals, 1977 FA122,” Materials Laboratory Factual 1LA386,” Materials Laboratory Factual
24. F.P. McClintock and G.R. Irwin, in Fracture Report 18-084, National Transportation Report 14-049, National Transportation
Toughness Testing and Its Applications, Safety Board, Washington, D.C., 2010 Safety Board, Washington, D.C., 2014
STP 381, American Society for Testing 42. C. Crussard, R. Borione, J. Plateau, Y. 61. J. Belan, Influence of Cooling Rate on g0
and Materials, 1965, p 84–113 Morillon, and F. Maratray, A Study of Morphology in Cast Ni-Base Superalloy,
25. P.C. Paris and G.C. Sih, in Fracture Impact Tests and the Mechanism of Brit- Acta Metall Slovaca, Vol 17 (No. 1),
Toughness Testing and Its Applications, tle Fracture, J. Iron Steel Inst., Vol 183, 2011, p 38–44
STP 381, American Society for Testing 1956, p 146–177 62. J. Belan, Materiálove´inžinierstvo, Vol
and Materials, 1965, p 30–81 43. O.Y. Kolesnychenko, R. de Kort, and H. 18, 2011, p 121–128
26. B.V. Whiteson, A. Phillips, V. Kerlins, van Kempen, Atomically Flat Ultra-Clean 63. C.V. Cooper and M.E. Fine, Metall. Trans.
and R.A. Rawe, Ed., in Electron Cr (001) Surfaces Produced by Cleavage A, Vol 16 (No. 4), 1985, p 641–649
64. O. Buck, W.L. Morris, and M.R. James, 79. R. Koterazawa, M. Mori, T. Matsni, and D. 99. “Aircraft Accident Report 1/2005, Report
Fracture and Failure: Analyses, Mechanisms Shimo, J. Eng. Mater. Technol. (Trans. on the Accident to Sikorsky S-76A+, G-
and Applications, P.P. Tung, S.P. Agrawal, ASME), Vol 95 (No. 4), 1973, p 202 BJVX near the Leman 49/26 Foxtrot Plat-
A. Kumar, and M. Katcher, Ed., Proceed- 80. R.W. Hertzberg, Fatigue Fracture Sur- form in the North Sea, 16 July 2002,”
ings of the ASM Fracture and Failure Ses- face Appearance, in Fatigue Crack Prop- Annex P, Air Accidents Investigation
sions at the 1980 Western Metal and Tool agation, STP 415, American Society for Branch, Department for Transport, United
Exposition and Conference (Los Angeles, Testing and Material, 1967, p 205 Kingdom, 2005
CA), American Society for Metals, 1981 81. C.D. Beachem, Trans. ASM, Vol 60 (No. 100. “Cycle Counting in Fatigue Analy-
65. D. Broek, Elementary Engineering Frac- 3), 1967, p 325 sis,” E 1049-85, American Society for
ture Mechanics, 3rd ed., Martinus Nijh- 82. R. Pippan, C. Zelger, E. Gach, C. Bich- Testing and Materials, West Consho-
off Publishers, The Hague, 1982 ler, and H. Weinhandl, On the Mecha- hocken, PA
66. P.J.E. Forsyth and D.A. Ryder, Fatigue nism of Fatigue Crack Propagation in 101. L. Molent, M. McDonald, S. Barter, and
Fracture: Some Results Derived from Ductile Metallic Materials, Fatigue R. Jones, Evaluation of Spectrum Fatigue
the Microscopic Examination of Crack Fract. Eng. Mater. Struct., 2011 Crack Growth Using Variable Amplitude
Surfaces, Aircr. Eng. Aerosp. Technol., 83. R.M.N. Pelloux, Mechanisms of Forma- Data, Int. J. Fatigue, Vol 30 (No. 1),
Vol 32 (No. 4), p 96–99 tion of Ductile Striations, Trans. ASM, 2008, p 119–137
67. D. Kramer, “Accident Docket DCA17 Vol 62, 1969, p 281–285 102. D. Kramer, “Accident Docket ENG17I
FA021,” Materials Laboratory Factual 84. R.M.N. Pelloux, Crack Extension by A008,” Materials Laboratory Factual
Report 17-034, National Transportation Alternating Shear, Eng. Fract. Mech., Report 17-039, National Transportation
Safety Board, Washington, D.C., 2017 Vol 7, 1970, p 235–247 Safety Board, Washington, D.C., 2017
68. M. Budinski, “Accident Docket WPR18F 85. F.E. Fujita, Acta Metall., Vol 6, 1958, p 543 103. N.E. Cherolis, D.J. Benac, and W.D. Pri-
A119,” Materials Laboratory Factual 86. D. Broek and G.O. Bowles, Int. J. Fract. demore, Fatigue in Rotating Equipment:
Report 18-078, National Transportation Mech., Vol 6, 1970, p 321–322 Is It HCF or LCF?, J. Fail. Anal. Pre-
Safety Board, Washington, D.C., 2018 87. P. Neumann, Acta Metall., Vol 22, 1974, vent., Vol 16 (No. 4), 2016, p 828–841
69. M. Fox, “Accident Docket ANC1 p 1155–1178 104. K. Wei and B.L. He, Failure Mechanism of
2IA085,” Materials Laboratory Factual 88. D.A. Meyn, Trans. ASM, Vol 61 (No. 1), Very High Cycle Fatigue for High Strength
Report 12-148, National Transportation 1968, p 42 Steels, Key Eng. Mater., Vol 664, Trans
Safety Board, Washington, D.C., 2012 89. D.L. Davidson and J. Lankford, Metall. Tech Publications, Ltd., Sept 2015,
70. pt?>E. Mueller, “Accident Docket Trans. A, Vol 15, 1984, p 1931–1940 p 275–281
CEN15 LA399,” Materials Laboratory 90. R.D. Carter, E.W. Lee, E.A. Starke, Jr., 105. Y. Lia et al., On the Formation Mechan-
Factual Report 16-016, National Trans- and C.J. Beevers, Metall. Trans. A, Vol isms of Fine Granular Area (FGA) on the
portation Safety Board, Washington, D.C., 15, 1984, p 555–563 Fracture Surface for High Strength Steels
2016 91. P.C. Paris and F. Erdogan, A Critical Anal- in the VHCF Regime, Int. J. Fatigue, Vol
71. R. Wanhill, S. Barter, and L. Molent, ysis of Crack Propagation Laws, J. Basic 82, 2016, p 402–410
Fatigue Crack Growth Failure and Lifing Eng. (Trans. ASME) D, Vol 85, 1963, p 528 106. C.E. Richards and T.C. Lindley, Eng.
Analyses for Metallic Aircraft Structures 92. H.H. Johnson and P.C. Paris, Eng. Fract. Fract. Mech., Vol 4, 1972, p 951
and Components, Springer, Netherlands, Mech., Vol 1, 1968, p 3 107. D. Kramer, “Accident Docket ENG12
2019 93. “Guidance Material for Aircraft Engine IA001,” Materials Laboratory Factual
72. S.A. Barter, L. Molent, and R.J.H. Wan- Life-Limited Parts Requirements,” FAA Report 12-028, National Transportation
hill, Typical Fatigue-Initiating Discontinu- Advisory Circular 33.70-1, Federal Avia- Safety Board, Washington, D.C., 2012
ities in Metallic Aircraft Structures, Int. tion Administration, Washington, D.C., 108. E.W. Hart, Two-Dimensional Phase
J. Fatigue, Vol 41, Aug 2012, p 11–22 July 31, 2009 Transformations in Grain Boundaries,
73. L. Molent and B. Dixon, Airframe Metal 94. P.H. DeVries, K.T. Ruth, and D.P. Den- Scr. Metall., Vol 2, 1968, p 179–180
Fatigue Revisited, Int. J. Fatigue, Vol nies, Counting on Fatigue: Striations 109. S.P. Lynch, B.C. Muddle, and T. Pasang,
131, Feb 2020 and Their Measure, J. Fail. Anal. Pre- Mechanisms of Brittle Intergranular
74. R.O. Ritchie, in Environment-Sensitive vent., Vol 10, 2010, p 120–137 Fracture in Al-Li Alloys and Comparison
Fracture of Engineering Materials, Z.A. 95. M. Bowry, N. Glover, D. Rugg, and P. with Other Alloys, Philos. Mag. A, Vol
Foroulis, Ed., The Metallurgical Society, Bowen, “Quantitative Assessment of 82, 2002, p 3361–3373
1979, p 538–564 Fatigue Crack Growth Path on Propagation 110. N. McAtee, “Accident Docket WPR16
75. R.O. Ritchie and J.F. Knott, Mater. Sci. Cycle Estimation,” 11th International FA153,” Materials Laboratory Factual
Eng., Vol 14, 1974, p 7 Conference on Fracture (ICF11), 2005 Report 17-067, National Transportation
76. R.W. Hertzberg, Deformation and Frac- 96. D. Hull, Crack Dynamic Effects, Chap. 9, Safety Board, Washington, D.C., 2017
ture Mechanics of Engineering Materi- Fractography: Observing, Measuring and 111. S. Lynch, Mechanisms of Fatigue and
als, 3rd ed., John Wiley and Sons, New Interpreting Fracture Surface Topogra- Environmentally Assisted Fatigue, Fatigue
York, 1989 phy, Cambridge University Press, 1999, Mechanisms, J. Fong, Ed., American Soci-
77. F. Zakar, “Accident Docket DCA09F p 259–291 ety for Testing and Materials, West Con-
A065,” Materials Laboratory Factual 97. R.O. Ritchie, Mechanisms of Fatigue- shohocken, PA, 1979, p 174–213
Report 09-060, National Transportation Crack Propagation in Ductile and Brittle 112. M.J. Mullen, A.H. Griebel, and J.M. Tar-
Safety Board, Washington, D.C., 2009 Solids, Int. J. Fract., Vol 100, 1999, taglia, Effects of Inhibited Acid Cleaning
78. F. Zakar, “Accident Docket DCA11M p 55–83 of Steel Fracture Surfaces for Failure
A039,” Materials Laboratory Factual 98. Y.N. Lenets and R.S. Bellows, Crack Prop- Analyses, J. Fail. Anal. Prevent., Vol 8,
Report 11-092, National Transporta- agation Life Prediction for Ti-6Al-4V 2008, p 12–19
tion Safety Board, Washington, D.C., Based on Striation Spacing Measurements, 113. E. Mueller, “Accident Docket DCA15M
2011 Int. J. Fatigue, Vol 22, 2000, p 521–529 P002,” Materials Laboratory Factual
Report 16-010, National Transportation 134. Appendix K: Threaded Fastener Failure 148. J. Epperson, “Accident Docket MIA0
Safety Board, Washington, D.C., 2016 Modes, High-Performance Bolting Tech- 3IA168,” Materials Laboratory Factual
114. “Occurrence A15W0023,” Engineering nology for Offshore Oil and Natural Gas Report 04-025, National Transportation
Laboratory Report LP0582015, Trans- Operations, National Academies of Safety Board, Washington, D.C., 2004
portation Safety Board of Canada Sciences, Engineering, and Medicine, 149. J. Epperson, “Accident Docket DCA98R
115. M.F. Ashby, G. Gandhi, and D.M.R. Taplin, Washington, D.C., The National Acade- A024,” Materials Laboratory Factual
Overview No. 3: Fracture-Mechanism mies Press, 2018, DOI 17226/25032 Report 98-088, National Transportation
Maps and Their Construction for f.c.c. 135. L.B. Pfeil, The Effect of Occluded Safety Board, Washington, D.C., 1998
Metals and Alloys, Acta Metall., Vol 27 Hydrogen on the Tensile Strength of 150. M. Fox, “Accident Docket FTW0
(No. 5), 1979, p 699–729 Iron, Proc. R. Soc. (London) A, Math. 1LA005,” Materials Laboratory Factual
116. D.A. Miller and T.G. Langdon, Creep Phys. Eng. Sci., Vol 112 (No. 760), Report 01-045, National Transportation
Fracture Maps for 316 Stainless Steel, 1926, p 182–195 Safety Board, Washington, D.C., 2001
Metall. Trans. A, Vol 10 (No. 11), 1979, 136. W.W. Gerberich, R.A. Oriani, M.J. Lji, X. 151. Corrosion-Deformation Interactions, Sec-
p 1635–1641 Chen, and T. Foecke, Necessity of Both ond Int. Conf. on Corrosion-Deformation
117. T.L. Silveira and S.N. Monteiro, Creep Plasticity and Brittleness in the Fracture Interactions, T. Magnin, Ed., Institute of
Fracture Behavior of Austenitic Stainless Thresholds of Iron, Philos. Mag. A, Phys. Materials, Nice, France, 1996
Steel from 550 to 800 C, Proceedings of Condens. Matter. Struct. Defects Mech. 152. Hydrogen Effects on Materials Behavior
the Fifth International Creep Fracture, Prop., Vol 63 (No. 2), 1991, p 363–376 and Corrosion Deformation Interactions,
March 29–April 3, 1981 (Cannes, France), 137. H.K. Birnbaum and P. Sofronis, Hydrogen- Proceedings of the Int. Conf. on Hydro-
p 1595–1601 Enhanced Localized Plasticity—A Mech- gen Effects on Material Behavior and
118. S.N. Monteiro et al., Creep Fracture anism for Hydrogen-Related Fracture, Corrosion Deformation Interactions
Mechanisms and Maps in AISI Type 316 Mater. Sci. Eng. A, Vol 176 (No. 1), (Moran, WY), R.N. Moody et al., Ed.,
Austenitic Stainless Steels from Distinct 1994, p 191–202 TMS, 2003
Origins, Mater. Res., Vol 20 (Suppl. 2), 138. A. Barnoush, “Hydrogen Embrittle- 153. H. Uhlig, New Perspectives in the Stress
2017 ment,” Ph.D. thesis, Universitat Des Corrosion Problem, Physical Metallurgy
119. W. Rossey, “Accident Docket ATL02L Saarlandes, 2011 of Stress Corrosion Fracture, T. Rhodin,
A060,” Materials Laboratory Factual 139. M. Nagumo, Hydrogen Related Failure of Ed., Interscience Publishers, New York,
Report 02-107, National Transportation Steels—A New Aspect, Mater. Sci. Tech- 1959, p 1–17
Safety Board, Washington, D.C., 2002 nol., Vol 20 (No. 8), 2004, p 940–950 154. D.D. Krueger, “The Development of
120. E. Mueller, “Accident Docket ERA17L 140. S.P. Lynch, A Fractographic Study of Direct Age 718 for Gas Turbine Engine
A326,” Materials Laboratory Factual Hydrogen-Assisted Cracking and Liquid- Disk Applications,” GE Aircraft Engines,
Report 19-005, National Transportation Metal Embrittlement in Nickel, J. Mater. Cincinnati, OH, TMS, 1989
Safety Board, Washington, D.C., 2019 Sci., Vol 21 (No. 2), 1986, p 692–704 155. D. Fournier and A. Pineau, Low Cycle
121. W.H. Johnson, On the Remarkable Changes 141. S.P. Lynch, Environmentally Assisted Fatigue Behavior of Inconel 718 at
Produced in Iron and Steel by the Action of Cracking: Overview of Evidence for an 298 K and 823 K, Metall. Trans. A, Vol
Hydrogen and Acids, Proc. R. Soc. (Lon- Adsorption-Induced Localised-Slip Pro- 8, 1977, p 1095–1105
don), Vol 23, 1874–1875, p 168–180 cess, Acta Metall., Vol 36, 1988, p 2639– 156. H. Ghonem and D. Zheng, “AAFOSR-
122. M.B. Whiteman and A.R. Troiano, Cor- 2661, and references therein 89-0285: Mechanisms and Modelling of
rosion, Vol 21, 1965, p 53 142. S.P. Lynch, Metallographic Contribu- Environment-Dependent Fatigue Crack
123. R. Langenborg, J. Iron Steel Inst., Vol tions to Understanding Mechanisms of Growth in a Nickel Based Superalloy,”
207, 1967, p 363 Environmentally Assisted Cracking, Met- prepared for Department of Air Force,
124. M.L. Holtzworth, Corrosion, Vol 25, allography, Vol 23, 1989, p 147–171 Air Force Office of Scientific Research,
1969, p 107 143. J. Gao, “Experiments to Explore the Bolling Air Force Base, Washington,
125. S. Singh and C. Altstetter, Metall. Trans. Mechanisms of Stress Corrosion Crack- D.C., 1992
A, Vol 13, 1982, p 1799 ing,” Thesis, University of Rochester, 157. H. Ghonem and D. Zheng, Depth of
126. D.A. Hardwick, M. Taheri, A.W. 2011 Intergranular Oxygen Diffusion during
Thompson, and I.M. Bernstein, Metall. 144. H. Moore, S. Beckinsale, and C.E. Mal- Environment-Dependent Fatigue Crack
Trans. A, Vol 13, 1982, p 235 linson, The Season Cracking of Brass Growth in Alloy 718, Mater. Sci. Eng.
127. R.J. Gest and A.R. Troiano, Corrosion, and Other Copper Alloys, J. Inst. Met., A, Vol 150, 1992, p 151–160
Vol 30, 1974, p 274 Vol 25, 1921, p 35–152 158. D. Kramer, “Accident Docket DCA15F
128. D.G. Chakrapani and E.N. Pugh, Metall. 145. R.J.H. Wanhill, R.T. Byrnes, and C.L. A185,” Materials Laboratory Factual
Trans. A, Vol 7, 1976, p 173 Smith, “Stress Corrosion Cracking in Report 15-131, National Transportation
129. J. Albrecht, B.J. McTiernan, I.M. Bern- Aerospace Vehicles,” Unclassified Report Safety Board, Washington, D.C., 2003
stein, and A.W. Thompson, Scr. Metall., NLR-TP-2010-538m, National Aerospace 159. H. Sehitoglu, Thermal and Thermome-
Vol 11, 1977, p 393 Laboratory, Amsterdam, The Netherlands, chanical Fatigue of Structural Alloys,
130. M. Taheri, J. Albrecht, I.M. Bernstein, Oct 2010 Fatigue and Fracture, Vol 19, ASM
and A.W. Thompson, Scr. Metall., Vol 146. “AAIB Investigation to Boeing 737-476 Handbook, ASM International, 1996,
13, 1975, p 871 (SF), EI-STD Left Main Landing Gear p 527–556, https://doi.org/10.31399/asm.
131. J. Albrecht, A.W. Thompson, and I.M. Failure during Landing Rollout, East hb.v19.a0002391
Bernstein, Metall. Trans. A, Vol 10, 1979, Midlands Airport, 29 April 2014,” Air 160. E. Mueller, “Accident Docket WPR18
p 1759 Accidents Investigation Branch Bulletin, LA248,” Materials Laboratory Factual
132. J. Albrecht, I.M. Bernstein, and A.W. 4/2015 EI-STD EW/C2014/04/03, April Report 19-059, National Transportation
Thompson, Metall. Trans. A, Vol 13, 9, 2015 Safety Board, Washington, D.C., 2019
1982, p 811 147. J. Epperson, “Accident Docket LAX02F 161. W. Zheng, D. Bibby, J. Li, G. Williams,
133. D. Warren, Hydrogen Effects on Steel, A056,” Materials Laboratory Factual W. Revie, and B. Tyson, “Stress Corro-
Mater. Perform., Vol 26 (No. 1), 1987, Report 03-052, National Transportation sion Cracking of Oil and Gas Pipelines:
p 38–48 Safety Board, Washington, D.C., 2003 New Insights on Crack Growth Behaviour
Gained from Full-Scale and Small Scale Solution,” ICF 12 International Confer- 164. E. Mueller, “Accident Docket DCA12
Tests,” ICF 12 International Conference ence on Fracture (Ottawa, Ontario, IA114,” Materials Laboratory Factual
on Fracture (Ottawa, Ontario, Canada), Canada), 2009 Report 13-088, National Transportation
2009 163. M. Fox, “Accident Docket DCA10M Safety Board, Washington, D.C., 2011
162. B. Fang, R.L. Eadie, M. Elboujdaini, and P007,” Materials Laboratory Factual 165. E. Mueller, “Accident Docket DCA12
W. Chen, “Transition from Pits to Cracks Report 11-055, National Transportation IA114,” Materials Laboratory Factual
in Pipeline Steel in Near-Neutral pH Safety Board, Washington, D.C., 2011 Report 13-087, National Transportation
Safety Board, Washington, D.C., 2011