You are on page 1of 10

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/257338458

Interpretation and significance of reverse chevron-shaped markings on


fracture surfaces of API X100 pipeline steels

Article  in  Materials Science and Engineering A · October 2013


DOI: 10.1016/j.msea.2012.04.108

CITATIONS READS

15 10,677

3 authors, including:

Jeffrey W. Sowards Elizabeth S Drexler


NASA National Institute of Standards and Technology
48 PUBLICATIONS   785 CITATIONS    64 PUBLICATIONS   937 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Hydrogen-assisted fatigue View project

Mechanics of pulmonary hypertension View project

All content following this page was uploaded by Elizabeth S Drexler on 11 October 2017.

The user has requested enhancement of the downloaded file.


Materials Science and Engineering A 551 (2012) 140–148

Contents lists available at SciVerse ScienceDirect

Materials Science and Engineering A


journal homepage: www.elsevier.com/locate/msea

Interpretation and significance of reverse chevron-shaped markings on fracture


surfaces of API X100 pipeline steels夽
Jeffrey W. Sowards ∗ , Chris N. McCowan, Elizabeth S. Drexler
National Institute of Standards and Technology, Materials Reliability Division, 325 Broadway, Boulder, CO 80305, United States

a r t i c l e i n f o a b s t r a c t

Article history: Fracture surfaces of X100 pipeline steels were examined with optical and electron microscopy after crack
Received 29 August 2011 tip opening angle fracture testing. Some fracture surfaces exhibited chevron-shaped fracture patterns that
Received in revised form 19 April 2012 are markedly different from classic chevron fracture. The chevron-shaped markings on the X100 fracture
Accepted 30 April 2012
surfaces point in the direction of crack growth, rather than towards the location of fracture initiation, as
Available online 10 May 2012
observed in classic cases of chevron fracture. Existing models, predicting formation of chevron fracture
patterns, do not explain the fracture behavior observed for X100 steel. A mechanism is proposed where
Keywords:
reverse chevron-shaped patterns are developed due to the shape of the crack front itself. The chevron
Banding
Chevron
shape forms as a result of crack tunneling, and the overall pattern is developed on the fracture surface
CTOA due to intermittent crack growth, resulting in alternating regions (bands) of fast fracture and slower,
Dynamic loading more ductile fracture. The contrast between these bands of alternating fracture defines the chevron. Care
Fracture should be taken during interpretation of intermittent chevron markings on fractures of ductile materials,
Pipeline steel as they may point away from rather than towards the origin of fracture.
Pop-in Published by Elsevier B.V.
Quasistatic loading
X100

1. Introduction • Chevron fracture surfaces are generally flat and perpendicular


to the principle stress and contain brittle and ductile fracture
Chevron markings (sometimes called herringbone patterns) features [9].
have been observed on surfaces of brittle fracture and were • The chevrons are orthogonal to the crack front, which is parabolic
described as early as 1914 [1]. Such markings have since been in shape [2,3,5].
observed on fractures of both crystalline and noncrystalline mate- • Crack propagation is discontinuous [2,6], occurring in alternating
rials. Although the chevron markings are typically associated with crystalline and fibrous appearance, where the ratio of these two
brittle fracture of steel plates [2–5], they have been observed on areas on the fracture surface was controlled by the brittleness of
polymer [6,7] and Al-alloy [8] fractures as well. Based on the back- the material [3].
ground provided in these works, there is a common recognition • Along both edges of the flat fracture plane, there are shear lips
and established interpretation of chevron markings. A schematic oriented at an angle of approximately 45◦ relative to the flat por-
of the classic chevron fracture markings typically observed for tion. These shear regions vary in width and diminish in size as
steel is shown in Fig. 1. A list of typical characteristics is provided the material becomes more brittle [2,3].
below: • Chevron patterns typically occur on fracture surfaces of thick
plates [5] where the plane strain condition is satisfied [8].

• Classic chevron markings point toward the location of fracture


initiation (opposite of the crack growth direction). Several models have been proposed and successfully applied
• The crack front leads at the plate center and lags at the surface. to describe the formation of the classic chevron fracture patterns.
• Chevrons are typically symmetric about the plate centerline [3]. Boyd first modeled this fracture type based on a work-of-fracture
approach, where the rate at which work was done to promote frac-
ture was directly proportional to the rate of increase of exposed
fracture area [5]. The model predicted that crack tunneling occurred
夽 Contribution of NIST, an agency of the US government. Not subject to copyright. in the running crack through the center of a plate and that the
∗ Corresponding author. Tel.: +1 303 497 7960; fax: +1 303 4975030. shape of the crack front was parabolic (see Fig. 1). Chevron markings
E-mail address: jeffrey.sowards@nist.gov (J.W. Sowards). formed orthogonal to the crack front and pointed toward the origin

0921-5093/$ – see front matter. Published by Elsevier B.V.


http://dx.doi.org/10.1016/j.msea.2012.04.108
J.W. Sowards et al. / Materials Science and Engineering A 551 (2012) 140–148 141

Fig. 1. Schematic illustration of “classic” chevron fracture appearance adapted from Boyd’s model;  denotes stress [5].

of fracture. As a special case, Kies et al. [6] suggested that reverse pipeline. The specimen design had a long ligament length to allow
chevron patterns could form where the apex pointed towards crack crack growth lengths of approximately 50–100 mm beyond the ini-
growth direction if cracking initiated from the sides of the plate tial notch. The specimen thickness in the gage section was 8 mm.
rather than near the center. Lee et al. [7] proposed a fracture model Chevron notches were machined into the specimens to a depth
similar to Boyd’s model [5], but for the case of reverse chevron frac- of approximately 60 mm from the specimen edge. The depths of
ture formation in polymer materials. In Lee’s model, the chevron the machined notches were then extended by either pre-cracking
apex also pointed in the same direction as the growing crack front. or electro-discharge machining another 5 mm or 10 mm in length,
The fundamental difference of this model from Boyd’s was that respectively, prior to testing. In regard to chevron-shaped mark-
crack initiation occurred at the plate surface, followed by radial ings, general fracture mode, and even maximum force at fracture
growth into the plate thickness, where cracks coalesced near the initiation, the method of pre-cracking made no difference.
plate center. The modified double cantilevered beam specimens were
As more widespread implementation of high-strength, high- secured in loading fixtures, and then tested with tensile load-
toughness pipeline steels (such as X100) occurs, interpretation of ing under actuator displacement control. Quasistatic and dynamic
fracture surfaces and consideration of the fracture mechanism is crack tip opening angle (CTOA) testing was conducted under
of interest. The present study describes chevron-shaped markings ambient laboratory conditions with uniaxial servohydraulic test
on the fracture surfaces of high-strength X100 pipeline steel that equipment. The quasistatic tests were performed with a 250 kN
have not yet been described in detail in the literature. The chevron force capacity load frame, and the dynamic tests with a 500 kN force
markings in X100 are found to be quite different from the classic capacity load frame. Actuator tensile displacement rates between
chevron markings, and the earlier models for chevron formation approximately 0.002 mm s−1 and 30,000 mm s−1 were possible
fail to predict the observed patterns. However, because chevron with these equipment configurations, resulting in both quasistatic
markings are a simple geometric pattern, it might be expected that and dynamic crack extension. Detailed equipment setup, testing
fracture markings of similar morphology could arise from a differ- procedures, and actual measurements of CTOA are discussed and
ent mechanism. Differences from and similarities to classic chevron described in detail elsewhere [10–13]. Displacement lengths were
markings in steel are discussed. A mechanism is proposed to explain sufficient that crack growth continued to near the end of specimens.
the development of chevron patterns in X100 steels. The test was interrupted on some specimens to heat-tint the crack
tip region in a furnace with oxidizing atmosphere.
2. Material and experimental procedures Microhardness values were determined after placing Vickers
indents on polished specimen surfaces with a 200 g load on the
Three X100 alloys were evaluated; the occurrence of chevron indenter tip. Hardness values are reported on the Vickers hard-
markings was not specific to one alloy composition or producer. ness number (VHN) scale. Mechanical properties were determined
Compositions of the three X100 steels evaluated during this inves- according to ASTM E08 standard methods for tension testing of
tigation are shown in Table 1. metallic materials.
Specimens were sectioned from pipes and machined into mod- Metallography was performed on specimens prepared with
ified double cantilevered beam test specimens. Note that modified standard polishing techniques for steel, including wet grinding
double cantilevered beam specimens are often referred to as “plate” with SiC paper followed by final polishing with 1 ␮m diamond
in this work, although they were machined from pipeline mate- suspension. Microstructure was revealed by etching in a 5% nital
rial. The modified double cantilevered beam specimen design and (5% nitric acid/95% methanol) solution. The post-test fracture sur-
dimensions are shown in Fig. 2. Crack growth was in the longi- faces from the modified double cantilevered beam specimens
tudinal direction of the plates to simulate a running crack in a were ultrasonically cleaned in methanol for several minutes, and
142 J.W. Sowards et al. / Materials Science and Engineering A 551 (2012) 140–148

Table 1
Chemical compositions (wt %) of X100 pipeline steels employed in this study.

X100 Pipe Al C Cr Cu Fe Mn Mo N Nb Ni P S Si Ti V

Alloy A 0.030 0.076 0.200 0.15 Balance 1.73 0.22 0.007 0.045 0.28 0.010 0.007 0.300 0.016 0.006
Alloy B 0.012 0.064 0.023 0.28 Balance 1.87 0.23 0.003 0.017 0.47 0.009 0.001 0.099 0.017 0.002
Alloy C 0.025 0.084 0.021 0.29 Balance 2.09 0.13 0.005 0.041 0.50 0.010 0.002 0.108 0.007 0.006

Fig. 2. Schematic drawing of a modified double cantilevered beam specimen and relevant dimensions.

then examined with conventional optical stereoscopic imaging and tests are reported for each steel. Standard deviation of yield
scanning electron microscopy (SEM). SEM imaging was performed strength and ultimate tensile strength were ±9 MPa from the
at 10 kV with secondary electrons. average or better, in both longitudinal and transverse orientations.
The ratio of yield strength to ultimate tensile strength (Y/T) is
3. Results included in Table 2. The Y/T ratio was quite high in this class of
materials, where values were typically 0.85 or greater. Materials
3.1. Microstructure and mechanical properties with high values of Y/T in general have less capacity for inelastic
deformation than those with low values [14], and tend to concen-
A representative optical micrograph of X100 pipeline steel trate deformation to the vicinity of a crack tip, rather than the plate
microstructure is shown in Fig. 3. The microstructure was a fine surrounding the crack tip [15]. The high ratio of Y/T presented
bainitic-ferrite with lath and polygonal morphology and exhibited here may influence the formation of reverse chevron-shaped
some banding along the rolling direction. Blocky TiN inclusions patterns, as discussed below. Examination of CTOA fractures of
were observed in unetched specimens and varied between approx- materials with lower Y/T (such as X65 steel), which is not reported
imately 1 ␮m and 5 ␮m in size. All three of the alloys investigated in this work, did not exhibit the reverse chevron-shaped patterns
in this study exhibited similar microstructural features to those reported here for X100 steels.
shown in Fig. 3. Hardness was determined in the bainitic-ferrite microstructure
Yield strength and ultimate tensile strength of the three X100 shown in Fig. 3. The average hardness (of ten measurements) in this
steels are presented in Table 2. Average properties of four tension region was 288 ± 6 VHN. Such hardness values are typical of this
microstructure in each of the three materials. Significantly higher
peak hardness values (up to approximately 490 VHN) were mea-
sured in regions of banded microstructure that were biased towards
the center of plate thickness.

3.2. Effect of displacement rate

Test displacement rate had a discernible effect on the fracture


mode, as shown in Fig. 4a. Quasistatic testing rates (≤300 mm s−1 )
generally resulted in flat fracture, slant fracture, or double-slant
fracture. Flat-fracture surfaces were approximately perpendicular

Table 2
Average tensile mechanical properties of X100 pipeline steels in transverse and
longitudinal orientations including 0.2% offset yield strength (Y) and ultimate tensile
strength (T).

X100 Steel Longitudinal Transverse

Y (MPa) T (MPa) Y/T Y (MPa) T (MPa) Y/T

Alloy A 722 855 0.844 912 916 0.996


Alloy B 729 838 0.870 833 868 0.960
Fig. 3. Representative microstructure of X100 pipeline steels investigated in this
Alloy C 732 806 0.908 798 827 0.965
study.
J.W. Sowards et al. / Materials Science and Engineering A 551 (2012) 140–148 143

Fig. 4. (a) Cross sections of X100 modified double cantilevered beam specimens tested at different crosshead velocities, (b) macrograph of the fracture surface from a
quasi-statically tested specimen, and (c) macrograph of the fracture surface from a dynamically tested specimen. The black arrow indicates crack growth direction.

to the loading direction. Slant and double-slant fracture surfaces the specimen centers that contained high-hardness microstruc-
were oriented approximately 45◦ to the loading direction. Dynamic tural banding. In both cases, the crack front was led preferentially
testing displacement rates (>300 mm s−1 ) always resulted in full- by the banded region, and the tunneling occurred with relatively
slant fractures. The most distinctive chevron-shaped markings we straight crack flanks that formed a wedge or chevron-shaped crack
present in this report occurred on the slant fractures that were front. Note that the chevron apexes point in the direction of crack
tested at a displacement rate of 0.02 mm s−1 . Faint chevron marks extension, unlike the classical case shown in Fig. 1. As previously
were visible on fracture surfaces tested at other displacement rates, noted, the chevron-shaped markings are much more distinct on the
but these less-distinctive markings may simply reflect the shape of quasistatic fracture. This is because the chevron shapes on the slant
the crack front, as shown in Fig. 4b and c. fractures produced at quasistatic displacement rates were a result
Chevron-shaped markings are highlighted for clarification on of intermittent crack growth. At dynamic displacement rates, the
the macrographs in Fig. 4. It appears in both the slant fractures crack growth was not discernibly intermittent.
from specimens tested at quasistatic and dynamic rates that the
crack tunneled from the initiation point (flat-fracture region). In 3.3. Fractography
both of the examples in Fig. 4, the fracture orientation changed
to slant after initiation. The slant-fracture morphology was pri- A chevron-shaped pattern observed on a slant-fracture surface
marily ductile tearing, although brittle fracture was observed near of a quasistatic test is shown in Fig. 5a. The direction of crack growth
144 J.W. Sowards et al. / Materials Science and Engineering A 551 (2012) 140–148

Fig. 5. Fracture surface from a quasistatic test showing (a) light microscopy macrograph (b–e) low- and high-magnification SEM images of chevron-shaped pattern regions
containing stair-shape ridges, (f) SEM image of region of quasi-cleavage fracture, and (g) SEM image of ductile fracture along ridges.

was from left to right, as indicated on the image. The chevron pat- (discussed below), and because these crack extensions occurred on
tern is distinct, due to the contrast between alternating bands of a fracture plane that was slightly tilted with respect to the plane of
unstable (pop-in) and stable fracture. The regions of unstable crack unstable crack extension.
extension are the lighter bands in Fig. 5a (labeled “U”), and contain Details of the linear features and the interface between the
linear step or ridge-like features. The regions of stable crack exten- regions of unstable and stable crack growth are shown in Fig. 5b
sion (labeled “S” in Fig. 5a) are darker in shading. These regions and c, respectively. The linear features served as steps up the slant-
have darker contrast because of their microscopic fracture features fracture surface and occurred in only the regions of unstable crack
J.W. Sowards et al. / Materials Science and Engineering A 551 (2012) 140–148 145

Fig. 6. Fracture surface from a dynamic test showing SEM images of (a) low-magnification brittle region near the centerline of plate and chevron-shaped crack front markings,
(b) high magnification of chevron marking running on diagonal from upper left to lower right, (c) region of centerline fracture, and (d) terraced cracking at centerline.

extension. In Fig. 5c the interface is evident where the unstable fracture occurring at the outside surface of the specimens. Evidence
crack extension was arrested and a region of fully ductile stable of 45◦ shear lips at the outside surface of the specimens mark
crack extension began. Here we see the sides of the steps (in the where the final fracture occurred, and support this sequence of
plane of plate) appear less ductile than the region of stable crack events.
extension on the right side, which is composed solely of ductile- Dynamic fracture surfaces were highly ductile, as evident by the
dimple morphologies. extensive microvoid coalescence shown in Fig. 6. Fig. 6a and b show
At higher magnifications, details of chevron regions reveal a no evidence of intermittent crack growth like that observed at the
complex morphology of alternating ductile and brittle features. The quasistatic test displacement rates. Even in regions where chevron-
intersection of stable and unstable fracture results in the chevron- shaped crack-front markings are evident, such as in Fig. 6b, regions
shaped marking, as shown in Fig. 5d, where it runs diagonally on both sides of the chevron ridge are fully ductile. A change in
through the upper right corner of the image. The ridges (see label ductile-dimple directionality was sometimes observed from one
in Fig. 5d) that form the chevron-shaped markings have a fully side of the chevron ridge to the other. This may be the cause of
ductile-dimpled appearance and were elevated from other areas of the macro scale contrast that makes the chevrons visible on the
fracture. Secondary ridge patterns, or steps, extend back towards dynamic fractures.
the fracture origin and have both ductile and quasi-cleavage fea- In both dynamic and quasistatic tests, the crack fronts were
tures associated with them. These secondary ridges are parallel to typically led by a region of brittle microstructure near the center
the crack growth direction, as shown at higher magnification in of pipe-wall thickness, as shown in Fig. 6c. These regions showed
Fig. 5e. Fracture in the region labeled “1” in Fig. 5f occurs on planes low ductility compared with surrounding regions, and secondary
parallel to or within 45◦ of the applied loading and has a charac- cracking was often observed (Fig. 6d). These more brittle regions
teristic quasi-cleavage fracture appearance. The ridges, indicated are shown to lead or strongly influence the symmetry of the crack
as region “2,” are roughly perpendicular to the applied loading front, because when the apex of the crack front is slightly (Fig. 6a) or
and exhibit a ductile-dimple fracture appearance (Fig. 5g). This significantly (Fig. 7) off center, it was observed that the centerline
particular fracture occurred on a roughly 45◦ slant with respect of the pipe wall thickness was also off the sample centerline. This
to the specimen surface, and the alternating patterns of ridges is because the modified double cantilevered beam specimens were
seem to indicate the movement of the fracture as it ascended and machined from curved pipe sections [16], not from flattened pipe
descended the slant fracture surface (outwards from the centerline) or plate material, and the mid-thickness of the pipe-wall was not
towards the plate surfaces. Ductile-dimple directionality supports necessarily at the centerline of the modified double cantilevered
the observation that fractures moved from the plate centerline out- beam specimen. The banded microstructure in the region along
wards. the pipe centerline has high hardness, as shown in Fig. 7. These
Apexes of the chevrons occurred near the middle of the regions fractured at angles of approximately 45◦ to the orientation
plate thickness, coinciding with the high-hardness microstructural of the modified double cantilevered beam specimen (and the prin-
banding that was observed in polished specimen cross-sections. cipal stress axis), and secondary cracks extended from the fracture
Fracture likely initiated at these interior banded regions, with final surface into the hard band.
146 J.W. Sowards et al. / Materials Science and Engineering A 551 (2012) 140–148

constituents associated with the centerline of the plate, which are


likely high-hardness martensite and upper bainite that formed
due to alloy segregation at the centerline [18]. One possible result
of brittle fracture near the centerline of the specimens is that the
faster rate associated with this brittle fracture might tend to outrun
the more ductile fracture on the flanks, resulting in less of the
curvature normally associated with fully ductile fracture fronts.
The flanks of the crack front trail behind the apex and form
the chevron-shaped crack front. The linear features that serve as
steps up the slant fracture surface (see Fig. 5) occur in only the
unstable crack extension regions. The unstable crack extension is
arrested at the start of the narrower banded regions, due to the
drop in force associated with the sudden crack extension (in a
displacement-controlled test). Then, as the force increases again as
test displacement continues to increase, the region of stable crack
extension occurs, due to ductile tearing associated with re-blunting
of the arrested crack front. As the force again becomes sufficiently
high, another unstable crack extension event occurs, which arrests,
and the cycle begins again.
Fig. 7. Macrograph of a pipeline cross section after dynamic test and inset showing It is expected that the crack front spreads out and climbs up
the brittle microstructural region that guided fracture near the plate centerline. and down the 45◦ slant fracture plane, leaving the linear features
Arrows indicate a high-hardness region, and values of hardness are in the VHN200
scale.
that are the most ductile regions of the step-like morphology (per-
pendicular to applied loading). Whether the fracture initiates near
the center of the plate for each cycle of unstable crack extension
4. Discussion and moves out to the plate surface (as indicated in the Fig. 8) is
not known. However, to some extent, a fracture sequence initiated
The X100 modified double cantilevered beam specimens tested from the interior of the plate is expected, because of the continu-
at both quasistatic and dynamic displacement rates exhibited ous presence of a small shear lip at the surface of the plate that is
chevron-shaped crack fronts. However, the fracture surface pat- indicative of final fracture.
terns on dynamically tested specimens were not as well defined. In The high Y/T ratio inherent to X100 materials might also influ-
comparison, quasistatic fractures (0.02 mm s−1 displacement rate) ence the development of the chevron-shaped patterns reported
of the X100 modified double cantilevered beam specimens had here since they have a limited strain-hardening capacity. Through-
well-defined chevron-like shapes, due to the high contrast between thickness plastic zones develop in thin-plate high-Y/T steels due
alternating bands of crack extensions formed by different fracture to low work hardening and low resistance to ductile tearing [15].
modes. Tensile prestrain (normalized by fracture strain) has been shown
The fracture behavior examined here has neither been identified by modeling to decrease critical J-integral for cleavage- and ductile-
nor discussed in detail elsewhere, but it appears to be characteris- fracture initiation toughness, and crack tip opening displacement
tic of the three different X100 alloys evaluated in this study. These [19]. Test results of high-Y/T pipeline steels give support to those
X100 alloys were produced at different times by different manu- predicted trends, as well [20]. Considering these previous studies,
facturers, and all show examples of the chevron-shaped fracture a low inelastic stain capacity of the current materials, and an open-
behavior discussed here for the range of quasistatic testing con- ing of the crack mouth during crack tunneling along the brittle
ditions evaluated. In addition, this fracture morphology has been microstructure, an effective prestrain in the plastic zone ahead of
observed during quasistatic drop weight tear test studies of X100 the crack front is likely. In support of this, Sivaprasad et al. have pos-
[17], which leads to the deduction that this pattern is not an arti- tulated that fracture toughness in ductile materials is governed by
fact specific to a particular test machine’s compliance or specimen the spread of plastic strains at a crack tip [21]. In development of the
configuration. However, it does not appear that the well-developed reverse chevron-shaped pattern, after the initial crack tunneling
chevron fracture features observed on X100 alloys tested at qua- period for a “pop-in” event, the crack dispreads towards the plate
sistatic test rates have any relationship with the classic chevron edges through the material with reduced toughness. When the
fracture mode that has been understood for many years. Below, crack front propagates completely through this prestrained mate-
the fracture mechanism leading to the formation of well-developed rial, a stable period of ductile tearing commences until available
chevron-shaped patterns in X100 is presented and compared with ductility is consumed and the process recurs.
the classic chevron fracture pattern.

4.1. Formation of the X100 chevron-shaped fracture pattern 4.2. Comparison with classic chevron fracture

Development of the chevron-shaped pattern in the X100 alloys The most striking differences between classic chevron mark-
is shown schematically in Fig. 8. The initial region of crack extension ings and the chevron-shaped markings observed on the X100 alloys
in the modified double cantilevered beam specimens is typically on include the directionality of the chevrons and the orientation of the
a plane perpendicular to the principal stress (flat), but soon turns to fracture plane (Table 3). In the classic case, chevrons form on flat
an angle near 45◦ (slant). Crack extensions on the slant surface are fractures and point toward the origin of the fracture. For the X100
shown in bands of alternating unstable and stable crack extensions. alloys the chevron shape is formed on a slant fracture (∼45◦ from
The crack front is led by and is symmetric about a plane of the applied stress) and is coincident with the crack front, point-
brittle phases aligned in the rolling direction along the midplane ing in the direction of crack extension. A similarity between the
of the plate. Typically, the central portion of the crack front tunnels two cases is that the patterns are made distinct due to the con-
ahead of the near-surface regions by about 3 mm. The apex of the trast developed between alternating bands of fully ductile, and less
crack front is symmetric about the hard, brittle microstructural ductile or brittle, crack extension.
J.W. Sowards et al. / Materials Science and Engineering A 551 (2012) 140–148 147

Fig. 8. Schematic of the development of a chevron-shaped fracture pattern during quasistatic testing of X100; arrows indicate the directions of crack front growth.

It has been demonstrated that classic chevron fracture can occur Key comparisons between classic chevron fractures and those
with the chevron pattern pointing in the direction of crack exten- observed in the current study are summarized in Table 3. Note that
sion if the fracture initiates at the outside surface of the specimen there are similarities in some of the geometric patterns, although
[6,7]. However, in the case of X100, the presence of shear lips on the underlying fracture mechanisms are distinct in nature. There-
the outside of the modified double cantilevered beam specimens fore, if geometric patterns with the chevron shape are observed
indicates that fracture does not initiate at the outside surface of the on fractures of highly ductile materials, careful interpretation of
specimens. This is a region of final fracture, not fracture initiation, microscopic and macroscopic features is necessary to identify frac-
which is a commonality with Boyd’s model [5] where fracture is led ture mode, as it has been demonstrated here that relying solely on
on or near the centerline. observation of chevron patterns is not adequate and may lead to
Another difference between classic chevron fracture and the misidentification of a ductile fracture as a brittle fracture.
X100 fracture is the extent of plastic deformation associated with
the two cases. Significant necking was observed in X100 modified 5. Conclusions
double cantilevered beam specimens near the vicinity of fracture,
indicating a large plastic zone size. X100 fractures are macroscopi- • Fracture surfaces exhibiting chevron-shaped patterns are char-
cally ductile (see Fig. 4a), and large plastic zones in mild steel plates acteristic of X100 materials under the quasistatic test loading
are expected to result in slant fracture rather than flat fracture [5]. conditions evaluated here.
The chevron-shaped pattern on fracture surfaces of X100 emu- • Chevron-shaped crack fronts are typical for fractures of X100
lates the shape of the crack front. This is unlike the crack front during quasistatic and dynamic crack growth conditions, but
shape of classic chevron fracture, where chevrons form in an ori- distinctive chevron patterns on fracture surfaces are due to
entation orthogonal to parabolic crack fronts [5]. This observation intermittent fracture observed only at quasistatic testing rates.
was confirmed by heat tinting crack fronts prior to fully fracturing A mechanism has been proposed to explain the formation of
the specimens. The crack fronts are chevron-shaped. chevron-shaped fractures on X100, where brittle microstructural
The distinctive chevron-shaped markings observed on the X100 banding leads the growing crack by extensive tunneling, and
alloys are due to the intermittent nature of the fracture process, intermittent pop-in causes the distinctive chevron due to alter-
with alternating unstable and stable crack-extension events. This nating bands of brittle and ductile fracture. Final fracture occurs
intermittent fracture process is also a key factor in the development at the outer surface where shear lips were evident.
of the classic chevron pattern. However, those fractures would • The evolution of chevron-shaped markings on X100 is differ-
be expected to have much higher fracture velocities [9] than the ent from the classical case of chevron fracture. Therefore, a new
chevron-shaped markings induced during quasistatic testing of the terminology should be considered that is more appropriate for
X100 alloys reported on here. describing the patterns observed in high-strength pipeline steels.
• Care should be taken in interpreting chevron-shaped markings
on fracture surfaces of materials with significant ductility as they
Table 3
may point away from, not towards the origin of fracture.
Summary and comparison of key characteristic features observed on classic chevron
fractures with X100 fractures.
Acknowledgments
Characteristic feature Classic chevron fracture Fracture of X100 steel

Chevron direction Apex points toward crack Apex points in the The authors wish to thank J.R. Fekete and M.D. Richards for crit-
origin direction of crack
ical reviews of this manuscript. Work performed by J.W. Sowards
propagation
Fracture mode Flat and perpendicular to Typically slant or was supported by a National Research Council Postdoctoral Fellow-
plate double-slant fracture ship.
Deformation Very little macroscopic Significant macroscopic
plastic plastic
References
Crack front Parabolic in shape in No parabolic crack fronts
both forward and reverse were observed and the
[6] chevron patterns front appears to be knife [1] C. de Freminville, Rev. Metall.-Paris 11 (1914) 971–1056.
[2] H. de Leiris, L’analyse morphologique des Cassures, Paris, France, 1945.
wedge
[3] C.F. Tipper, Admiralty Ship Welding Committee, London, 1948.
Propagation Intermittent Intermittent pop-in
[4] C.F. Tipper, J. Iron Steel Inst. 185 (1957) 4–9.
148 J.W. Sowards et al. / Materials Science and Engineering A 551 (2012) 140–148

[5] G.M. Boyd, Engineering 175 (1953) 65–69, 100–102. [13] Ph.P. Darcis, C.N. McCowan, H. Windhoff, J.D. McColskey, T.A. Siewert, Eng.
[6] J.A. Kies, A.M. Sullivan, G.R. Irwin, J. Appl. Phys. 21 (1950) 716–720. Fract. Mech. 75 (2008) 2453–2468.
[7] M.-P. Lee, A. Hiltner, E. Baer, C.-I. Kao, J. Mater. Sci. 28 (1993) [14] Y. Fukumoto, Eng. Struct. 18 (1996) 786–791.
1491–1502. [15] A.C. Bannister, J.R. Ocejo, F. Gutierrez-Solana, Eng. Fract. Mech. 67 (2000)
[8] J.A. Francis, G.M.D. Cantin, Mat. Sci. Eng. A-Struct. 407 (2005) 547–562.
322–329. [16] Ph.P. Darcis, C.N. McCowan, J.D. McColskey, R. Fields, Fatigue Fract. Eng. M. 31
[9] ASM Handbook Committee, Metals Handbook, Vol. 12: Fractography, Ninth Ed., (2008) 1065–1078.
ASM, International, Metals Park, OH, 1987, pp. 107–112. [17] S. Xu, W.R. Tyson, R. Eagleson, C.N. McCowan, E.S. Drexler, J.D. McColskey, Ph.P.
[10] A. Shtechman, C. McCowan, R. Reuven, P. Darcis, R. Smith, J. Merritt, T. Siew- Darcis, ASME Conf. Proc., vol. 269, 2010, http://dx.doi.org/10.1115/IPC2010-
ert, E. Drexler, J.M. Treinen, J.D. McColskey, ASME Conf. Proc., vol. 273, 2008, 31076.
http://dx.doi.org/10.1115/IPC2008-64362. [18] H. Tamehiro, T. Takeda, S. Matsuda, K. Yamamoto, N.T. Okumura, Trans. Iron
[11] R. Reuven, C. McCowan, E. Drexler, A. Shtechman, Ph. Darcis, M. Treinen, R. Steel Inst. Jpn. 25 (1985) 982–988.
Smith, J. Merritt, T. Siewert, J.D. McColskey, ASME Conf. Proc., vol. 279, 2008, [19] A. Cosham, J. Offshore Mech. Arct. Eng. 123 (2001) 182–190.
http://dx.doi.org/10.1115/IPC2008-64363. [20] N. Fukuda, N. Hagiwara, T. Masuda, J. Offshore Mech. Arct. Eng. 127 (2005)
[12] E. Drexler, P. Darcis, C. McCowan, J.M. Treinen, A. Shtechman, R. Reuven, T. 263–268.
Siewert, R. Smith, J. Merritt, J.D. McColskey, ASME Conf. Proc., vol. 287, 2008, [21] S. Sivaprasad, S. Tarafder, V.R. Ranganath, K.K. Ray, Mater. Sci. Eng. A-Struct.
http://dx.doi.org/10.1115/IPC2008-64369. 284 (2000) 195–201.

View publication stats

You might also like