Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Materials Characterization 189 (2022) 111940

Contents lists available at ScienceDirect

Materials Characterization
journal homepage: www.elsevier.com/locate/matchar

Strain localization in titanium investigated via in situ digital image


correlation with multiscale speckles
X.H. Gong a, Z.D. Feng b, D. Fan b, L. Lu a, *, S.N. Luo a
a
Key Laboratory of Advanced Technologies of Materials, Ministry of Education, Southwest Jiaotong University, Chengdu, Sichuan, PR China
b
The Peac Institute of Multiscale Sciences, Chengdu, Sichuan, PR China

A R T I C L E I N F O A B S T R A C T

Keywords: Strain localization in commercially pure titanium under uniaxial tension is investigated with in situ digital image
Strain localization correlation (DIC) at different length scales. Speckles in three different sizes are prepared on the surface of ti­
Digital image correlation tanium specimens, referred to as microscale, mesoscale and macroscale speckles. The microscale and mesoscale
Multiscale
speckles for scanning electron microscopy DIC analysis are fabricated by different electro-polishing procedures,
EBSD
Titanium
and spray painting is used to prepare the macroscale speckles for the optical DIC measurement. Electron back
scattering diffraction (EBSD) characterization is conducted for the microscale speckle case. The strain fields
obtained at different length scales combined with the microscale EBSD examination establish the relationship
between microscale strain localization and macroscale strain concentration bands. Strain localization tends to
form at grain boundaries and inside deformation twins, especially at the large-angle grain boundaries, and the
amplitude and rate of strain accumulation at grain boundaries are higher than inside deformation twins. The
microscale strain-localized regions dispersed at grain boundaries and within deformation twins present them­
selves as the strain concentration bands observed with the macroscale speckles, and both the strain localization
regions and strain concentration bands are at ±45∘ with the tensile direction.

1. Introduction deformation measurement method, which can obtain displacement and


strain by correlation between digital images of the undeformed and
Strain localization occurs across a wide variety of materials, loading deformed configurations. The DIC method was widely used to investi­
conditions and application scenarios, and usually plays a decisive role in gate the strain localization in metallic materials such as steel [9–11],
mechanical behavior, thereby influencing the suitability of a given nickel-based superalloy [12,13], aluminum [14], magnesium [15–20],
material for the intended structural application [1]. Many investigations zirconium [21] and titanium [22–29]. Such technique provides more
over different length scales ranging from the microscale to the macro­ information to better understand deformation, and it is important to
scale have been conducted to identify early precursors of strain locali­ choose appropriate speckle size and DIC parameters to match the spatial
zation and establish their relation with larger-scale material failure scale corresponding to a specific deformation mechanism.
[2–4]. Although damage can occur in the sample interior, surface Titanium and its alloy have attracted increasing attention in
deformation analysis can still serve as an efficient means to help build biomedical, aircraft and automobile applications due to their remark­
the link between strain localization and fracture in a specimen under able combination of low density, high biocompatibility, high toughness,
cyclic or sustained loading, and is commonly used for thin specimens excellent heat resistance and corrosion resistance [30–34]. With a hex­
under tension. Strain field measurements over different length scales agonal close-packed (HCP) crystal structure at room temperature, the
ranging from the microscale to the macroscale is conducive to under­ plastic deformation of titanium is accommodated by both dislocation
stand the microstructural origin of strain localization, predict material slip and deformation twinning [35–37]. Such a deformation feature
failure on the macroscale, and promote structure development for gives rise to more complex strain localization than those of conventional
fracture resistance in particular for the cases where damage nucleates on metals with a cubic crystal structure [38,39]. For instance, strain
a surface. localization occurs preferentially at both grain boundaries and twin
Digital image correlation (DIC) technique [5–8] is a contactless boundaries in magnesium with an HCP structure, as demonstrated by

* Corresponding author.
E-mail address: llu@swjtu.edu.cn (L. Lu).

https://doi.org/10.1016/j.matchar.2022.111940
Received 27 December 2021; Received in revised form 3 April 2022; Accepted 1 May 2022
Available online 10 May 2022
1044-5803/© 2022 Elsevier Inc. All rights reserved.
X.H. Gong et al. Materials Characterization 189 (2022) 111940

the DIC analysis from micrographs obtained with scanning electron CP-Ti specimen and a copper rod are used as electrodes. The electro-
microscopy (SEM) and electron backscatter diffraction (EBSD) mea­ polishing voltage and current are controlled by a direct-current power
surements [40]. The localized plastic strain in pre-twinned commer­ supply, and liquid nitrogen is used to cool the mixed solution. In order to
cially pure titanium (CP-Ti) during detwinning deformation tends to obtain micron-scale speckles for DIC analysis at the microscale (high-
accumulate around the {1012} extension twins [41]. Nevertheless, the magnification SEM-DIC) and reduce the interference with EBSD char­
effects of deformation twins and grain boundaries on strain localization acterization, the electro-polishing voltage, temperature and duration are
in CP-Ti during tension are rarely studied. Furthermore, ex situ optical set to 30 V, − 30∘C, and 40 s, respectively. Furthermore, the electro-
DIC measurements with different magnifications ranging from 3.2× to polishing temperature and duration are adjusted to − 20∘C and 60 s,
50×, were conducted to characterize the spatial distribution of residual respectively, to acquire larger speckles for DIC analysis at the mesoscale
deformation in titanium; the strain fields at the highest optical magni­ (low-magnification SEM-DIC) with an electro-polishing voltage of 30 V.
fication revealed deformation patterns that are not detectable at lower The electro-polished CP-Ti specimen is gripped on a homemade
magnifications, and the length estimation of a representative volume miniature tensile device placed inside the chamber of a Thermo Fisher
element based on the standard deviation of the average residual strain Scientific Quanta 250 SEM, as illustrated in Fig. 2b. The dog-bone sha­
with different optical magnification presents a clear difference [42]. ped CP-Ti specimen is tilted by 70∘ with the horizontal plane for the
However, in situ and finer-scale DIC measurements are still lacking on EBSD examination. The incident electron beam irradiates the CP-Ti
CP-Ti. specimen, and excites secondary and backscattered electrons used for
In this work, the electro-polishing and spray painting methods are SEM imaging and EBSD measurements, respectively. The miniature
used to prepare speckles of three different characteristic sizes on the tensile device is equipped with a load cell to obtain stress− strain curves.
surface of a CP-Ti sample. The strain fields at different length scales are The bulk strain is calculated by measuring the axial displacement
obtained via in situ SEM DIC and optical DIC measurements on the CP-Ti generated by the stepper motor, and corrected by the rigid deformation
specimens with different-sized speckles. EBSD characterization is per­ of the miniature tensile device. The strain rate is roughly 10− 3 s− 1,
formed to reveal microstructure evolution. Such examinations realize in controlled by the speed of stepper motor. The CP-Ti specimen is suc­
situ investigation of strain localization over a wide range of length scales. cessively subjected to uniaxial tension to bulk strains of 0.03, 0.06, 0.09,
The microscale strain-localized regions are dispersed at grain bound­ and 0.12. When the specimen deforms at one of the above strains, both
aries and within deformation twins, forming the strain concentration the SEM imaging and EBSD examinations are performed with the
bands on the macroscale. loading. The time to acquire SEM and EBSD images is 10 s and 20 min,
respectively. The SEM accelerating voltage and working distance are 20
2. Materials and experiments keV and 15 mm, respectively, and the scanning step size for the EBSD
scan is 1 μm.
The CP-Ti studied here has a nominal chemical composition (wt%) of For even larger speckles, spray painting instead of electro-polishing
Fe 0.15, O 0.15, Si 0.10, C 0.03, N 0.02, H 0.02, and Ti balance. Fig. 1 is applied, since higher electro-polishing temperature/longer duration
shows the microstructure characterization of the as-received CP-Ti. As damage the specimen surfaces. As displayed in Fig. 3a, a thin layer of
revealed by the EBSD analysis, the CP-Ti is equiaxed with an average black paint is first spread on the whole surface of a precision-polished
grain size of ~40 μm. The CP-Ti has a basal transverse texture, and the specimen. By spraying white paint at a distance, tiny white paint
deviation of the average grain orientation from ND is about 40 degrees. droplets attach to the specimen surface, forming speckles for optical
Dog-bone shaped specimens are harvested form the CP-Ti plate. The DIC. Then the speckled CP-Ti sample is subjected to tensile loading with
gauge dimensions are 4 mm long, 2 mm wide, and 0.4 mm thick. The a material testing system and illuminated by a light source, and the
electro-polishing and spray painting methods are used to prepare deformation is recorded by a charge-coupled device camera. The im­
speckles with different speckle sizes on the surfaces of three different aging exposure time is 1 ms, and the frame interval is 5 s. The corre­
dog-bone shaped specimens. Three different speckle sizes are fabricated sponding strain rate is 10− 3 s− 1.
on three CP-Ti specimens. Then the specimens are subjected to in situ
tensile loading along with SEM/optical imaging. 3. Results and discussions
Fig. 2a presents the electro-polishing method for preparing SEM
speckles. A dog-bone shaped specimen is first ground and mechanically SEM micrographs and optical images of the CP-Ti specimens after
polished. Then the polished specimen is immersed in a solution of 10 vol electro-polishing and spray painting together with the histograms of
% perchloric acid and 90 vol% methanol for electropolishing, and a pixel intensity are presented in Fig. 4. Both SEM and optical imaging
magnetic stirrer is placed in the solution to mix the solution evenly. The regions are located in the middle of the specimen gauge section, so the
deformation histories described by different speckles are the same. The
field of view increases with the decrease of imaging magnification. A
(a) (b) {0002}
large number of artificial speckles distribute randomly in all the three
RD
images with different speckle size scales. The average speckle sizes for
the low- and high-temperature electro-polishing and spray painting are
approximately 1.0 μm, 5.4 μm and 57.9 μm, referred to as the micro­
TD
scale, mesoscale and macroscale speckles. According to the pixel in­
tensity histograms, the ranges of grayscale values for all three images are
large, and the microscale and mesoscale speckles exhibit a better quality
-
TD RD {1120} than the macroscale speckles due to the considerably higher
homogeneity.
In order to evaluate the overall quality of microscale, mesoscale and
TD macroscale speckles, we introduce the mean intensity gradient [43] of a
speckle pattern, defined as
9
0001
--
2110
H ⃒
⃒ ( )⃒
50 μm ∑W ∑
∇g xi , yj ⃒
(1)
0
-
1010
δg = .
i=1 j=1
WH
Fig. 1. EBSD characterization of the CP-Ti. (a) Inverse pole figure map. (b) Pole
Here W and H are the width and height (in unit of pixels) of the
figures. RD: rolling direction; TD: transverse direction; ND: normal direction.

2
X.H. Gong et al. Materials Characterization 189 (2022) 111940

(a) (b)
Thermometer

Electron beam
Load cell

V
Secondary
Copper Sample + A
- electron
detector
Stir bar Power supply
MTS
Magnetic stirrer Timer EBSD
detector

Fig. 2. (a) Illustration of speckle preparation via electro-polishing. (b) Schematic of in situ tensile test together with SEM and EBSD measurements.

(a) (b)

Light
Sample
CCD camera

Spray black paint Spray white paint


Tensile loading

Fig. 3. (a) Illustration of speckle preparation via spray painting. (b) Schematic of in situ tensile test along with optical imaging.

region of interest in an image, respectively. x and y are coordinates. g is In addition to the EBSD characterization, SEM secondary electron
( ) [ ( ) ]12 imaging and optical imaging are used to record the deformation of CP-Ti
the pixel intensity function. ∣∇g xi , yj ∣ = gx2 xi , yj + gy2 (xi , yj ) is the
specimen and map strain fields by performing the DIC analysis. By
modulus of the intensity gradient at pixel (xi, yj), and gx(xi, yj) and gy(xi, calculating the correlation coefficients of two images before and after
yj) are the x- and y-components of the intensity gradient at coordinate tensile deformation, the randomly distributed speckles are “tracked,”
(xi, yj), respectively. The mean intensity gradients for the microscale, and the displacement (u) field is obtained, from which the Euler-Almansi
mesoscale and macroscale speckles are 28.1, 32.7 and 11.3, respectively. strain (eij) is calculated as
A larger mean intensity gradient indicates a higher correlation coeffi­
1( )
cient of speckles between two images, i.e., better speckle quality. The eij = ui,j + uj,i − uk,i uk,j , (2)
mean intensity gradient values of the microscale and mesoscale speckles 2
are much higher, consistent with the histograms of pixel intensity. Fig. 5 where i, j, k = x, y.
shows the tensile stress− strain curve of the CP-Ti specimen. The tensile For DIC analysis, the uncertainties in strain for the microscale,
strain rate is about 10− 3 s− 1. Upon yield, the CP-Ti specimen undergoes mesoscale and macroscale speckles are analyzed at different subset radii
a short period of strain hardening, and then the flow stress gradually [44,45]. DIC analysis is performed on two undeformed images with
decreases until the specimen fails. The yield stress and tensile strength different subset radii (10, 15, ⋯, 60 pixels), yielding the Euler-Almansi
are 275 MPa and 332 MPa, respectively. Four states on the stress− strain strain fields, from which the average strain components are obtained.
curve are selected for the SEM, EBSD and imaging measurements (the The average values and standard deviations of the Euler-Almansi strain
squares), representing the undeformed, strain hardening, maximum for the microscale, mesoscale and macroscale speckles are obtained as a
tensile stress and strain softening stages. For the microscale speckle case, function of subset radius (Fig. 7). The step size of grid is 1 pixel, and the
stress relaxation is observed during EBSD measurement. subset overlap for different subset radii is over 90%. At a given subset
The tension-induced microstructure evolution of the CP-Ti is pre­ radius, the average strain increases as the speckle size increases, and the
sented in Fig. 6, and the corresponding undeformed microstructure is standard deviations for the micro- and mesoscale speckles are similar
displayed Fig. 1. The {1012} extension twinning dominates the plastic but considerably lower than for the macroscale speckles in particular at
deformation. At a strain of 0.03, a few {1122} extension twins and more small subset radii. With the increase of subset radius, the average strain
{1012} extension twins are observed. As the tension progresses, both the and standard deviation all decrease, and the differences among the three
number and the area of the {1012} extension twins increases signifi­ speckle types become smaller. For a subset radius of 45 pixels, both the
cantly, while the {1122} extension twins only widen slightly. Such average strain and the standard deviation for microscale, mesoscale and
differences are attributed to a lower critical resolved shear stresses for macroscale speckles are similar. Therefore, a subset radius of 45 pixels is
activating the {1012} twin than the {1122} twin. selected for the DIC analysis.
By performing DIC analysis of the microscale, mesoscale and

3
X.H. Gong et al. Materials Characterization 189 (2022) 111940

Fig. 4. Comparison of speckles prepared via electropolishing (the top and middle rows) and spray painting (the bottom row). Left column: (a) and (d) SEM mi­
crographs with magnifications of 1000× and 200×, respectively, (g) optical image with a magnifications of 10×. Middle column: histograms of pixel intensity. Right
column: speckle size (equivalent diameter) distribution and cumulative counts.

overlaid on the strain fields. For the microscale speckles, (Fig. 8a–8c),
the grain boundaries are overlaid on the strain field maps. At a bulk
strain of 0.03, significant strain concentration can be observed, and the
regions with high strain values are mainly located near the general grain
boundaries and inside the deformation twins. The high strains close to
grain boundaries were observed previously, and attributed to intense
local slip activity, grain boundary sliding or kink bands formation [24].
Moreover, moderate strain concentration also appears in the interiors of
some grains, and exhibits the characteristics of strain infiltration from
grain boundary to grain interior. The activation of the pyramidal 〈a〉 slip
systems in polycrystalline aggregates occurs primarily due to the large
stresses generated in grain-boundary regions because of the misorien­
tation between neighboring grains [46]. The strain fields and corre­
sponding Schmid factor mapping for the {1011}〈1120〉 pyramidal 〈a〉
slip are presented in Fig. 9a and b. The Schmid factors of the grains with
high intracrystalline strain are larger than those of the grains with low
intracrystalline strain. In other words, the strain localization within
grains may be related to the {1011}〈1120〉 pyramidal 〈a〉 slip. As the
tensile strain increases, strain accumulates in the regions where strain
Fig. 5. Stress–strain curve for the CP-Ti specimen. Squares: the states for the
SEM, EBSD and optical imaging measurements. concentration occurs in the early deformation stage, and the accumu­
lation rate inside deformation twins is lower than the rate nearby the
general grain boundaries. Strain concentration is nearly absent in the
macroscale speckles, the evolution of the Euler-Almansi strain fields at
other regions. It is worth noting that the strain concentrated regions are
different length scales for the CP-Ti is shown in Fig. 8. For the high-
roughly oriented at ±45∘ with respect to the tensile loading direction,
magnification SEM-DIC case, DIC analysis can realize strain field map­
and are relatively scattered. Similar ±45∘ orientations are are observed
ping with the SEM micrographs, and the SEM and EBSD images are
in CP-Ti subjected to cyclic tensile loading via optical DIC analysis.
measured in the same region. Therefore, the spatial correlation between
These concentrated bands appeared to correspond to the grain
EBSD and DIC information is established, and the grain boundaries are

4
X.H. Gong et al. Materials Characterization 189 (2022) 111940

Fig. 6. (a) Inverse pole figure maps (the


(a) 0.03 (b) 0.06 (c) 0.09 upper row) and corresponding image qual­
ity maps overlaid with twin boundaries (the
lower row) for the CP-Ti deformed at
different tensile strains (numbers in the
upper right corner). The types of deforma­
tion twins are identified by the angle be­
tween neighboring grains along a specific
crystallographic orientation. For instance, if
the angle between the 〈1120〉 orientations
(d) 0.03 (e) 0.06 (f) 0.09 of two neighboring grains is 85.0±5∘, then
the twin is a {1012} extension twin. The
blue, red and green lines refer to the {1012}
extension, {1122} contraction and {1121}
extension twin boundaries, respectively.
(a)–(c) refer to approximately the same re­
gion of interest. (For interpretation of the
references to colour in this figure legend,
RD 0001
--
2110
- -
1210 85.0±5° {1012} extension 100 μm the reader is referred to the web version of
- -
1010 64.4±5° {1122} contraction
- - this article.)
1010 35.0±5° {1121} extension
ND -
1010

speckle size (Fig. 8d–8f), the strain concentrated regions are still ori­
(a) ented approximately at ±45∘ with the tensile direction. However, the
angle for some concentrated regions is much smaller or larger than 45∘.
Such features may be attributed to the effects of microstructure and
boundaries on local deformation, causing deviation from the idealized
45∘. Compared with the case of microscale speckles, the evolution of
strain field for the case of mesoscale speckles is similar, while the dis­
tribution of the strain concentrated regions is not as dispersed as the
microscale case. In the optical imaging case (macroscale speckles), the
strain fields show several main strain concentrated bands oriented at
±45∘ with the tensile direction. The “multiscale” strain fields indicate
that the strain localized regions at different spatial scales are oriented at
±45∘ with the tensile direction, and the discrete strain concentrated
regions at the microscale form large strain concentrated bands on the
macroscale as a result of spatial averaging.
Furthermore, additional characterization at the microscale is per­
formed on a CP-Ti specimen with a tensile strain of 0.03. The image
quality and Euler-Almansi strain field maps overlaid with grain
boundaries are shown in Fig. 9c and d. Similar to Fig. 8a, strain parti­
(b) tioning within deformation twins and grain boundaries are also
observed in Fig. 9c and d. Nevertheless, the strain levels within the
{1122} contraction twins are much lower, indicating that the defor­
mation twin type affects the degree of strain localization.
In order to understand the mechanism of strain localization of the
CP-Ti specimen under uniaxial tension, the misorientation angles for the
strain concentrated grain boundaries are calculated in Fig. 10b given the
strain field and EBSD results. The strain localized grain boundaries are
those with the strains around them twice the global strain of the CP-Ti
specimen. The distribution of the misorientation angles in the unde­
formed CP-Ti is approximately uniform (Fig. 10a), and different from
that of the deformed sample. For the deformed sample, the grain
boundaries with a high misorientation angle (>80∘) have the highest
proportion, indicating that strain concentration tends to form at large-
angle grain boundaries (Fig. 10b); such a biased distribution of misori­
entation angle is a result of strain localization.
In addition, the angles (θ) between the loading direction and the
strain-localized twin boundaries/general grain boundaries are presented
in Fig. 10c and d. The proportions of θ = 45∘ for both the twin boundaries
Fig. 7. (a) Average strain and (b) standard deviation of strain as a function of and the general grain boundaries are the largest, and the more θ deviates
subset radius for speckled images.
from 45 degrees, the lower its proportion. Such deviations may be
caused by the coordination of microstructure on local deformation.
boundaries inclined ±45∘ with respect to the loading direction, and the Furthermore, the contributions to strain concentration of twin bound­
bands sometimes extended into neighboring grain interiors [42]. aries and general grain boundaries are compared. The average strains
With the decrease of SEM imaging magnification and the increase of for the strain-localized deformation twins, the general grain boundaries

5
X.H. Gong et al. Materials Characterization 189 (2022) 111940

(a) 0.03 (b) 0.06 (c) 0.09

45°

45°

0 0.18 50 μm
(d) 0.03 (e) 0.06 (f) 0.09

45°

45°

0 0.18 200 μm
(g) 0.03 (h) 0.06 (i) 0.09

45°
45°

ND RD
0 0.18 1 mm
(j) 0.03 (k) 0.06 (l) 0.09

Fig. 8. Euler-Almansi strain field maps of three different CP-Ti specimens acquired with high-magnification SEM-DIC (the top row), low-magnification SEM-DIC (the
middle row), and optical DIC (the bottom row), respectively, at different tensile strains as noted. The high-magnification SEM maps (the upper row) are overlaid with
twin boundaries (gray curves) and general grain boundaries. The loading direction is horizontal. The dashed lines are at ±45∘ with the loading direction.

(a) 0.03 (c) 0.03

0 0.16 50 μm 0 0.16 50 μm

(b) 0.03 (d) 0.03

- - -
0 0.5 {1012} extension {1122} contraction {1121} extension

Fig. 9. (a) Euler-Almansi strain field maps shown in Fig. 8a and b corresponding Schmid factor mapping for the {1011}〈1120〉 pyramidal 〈a〉 slip. (c) Image quality
and (d) Euler-Almansi strain field maps overlaid with twin boundaries (gray curves) and general grain boundaries for another CP-Ti specimens at a strain of 0.03.

and the high angle grain boundaries are shown in Fig. 11. The spatial average strain for high angle grain boundaries is larger than that of
cutoff used to calculate the mean strain is 3 pixels around the grain general grain boundaries. Such results suggest that the accumulation of
boundary. Both the average value at the same tensile strain and the localized plastic strain during tension tends to occur preferentially at
increase rate of average strain for the strain-concentrated grain general grain boundaries, especially at the large angle grain boundaries.
boundaries are larger than those for the strain-concentrated twins. The A probable reason is that the pre-existing grain boundaries are prone to

6
X.H. Gong et al. Materials Characterization 189 (2022) 111940

(a) (b)

(c) (d)
45°

45°

Fig. 10. Neighbor-to-neighbor misorientation angle distributions for (a) the as-received CP-Ti (corresponding to Fig. 1a) and (b) the strain-localized grain boundaries
in deformed CP-Ti (Fig. 6a), and histograms of the angle between strain-localized (c) twin boundaries and (d) general grain boundaries with the loading direction.

conclusions are as follows.


• The strain localization regions are mainly at ±45∘ with the tensile
direction, regardless of the resolution of strain field and speckle size, and
the deviations from ±45∘ may be related to the coordination of micro­
structure induced by local deformation.
• The strain concentration bands observed with the macroscale
speckles are composed of the microscale strain localized regions
dispersed at grain boundaries and within deformation twins.
• Strain localization tends to form at grain boundaries and inside
deformation twins, especially at the large angle grain boundaries.
• The amplitude and rate of strain accumulation at grain bound­
aries are higher than inside deformation twins, probably because the
pre-existing grain boundaries are prone to dislocation tangling.

Data availability

The raw/processed data required to reproduce these findings cannot


be shared at this time as the data also forms part of an ongoing study.

Credit authorship contribution statement


Fig. 11. Average strain values of the strain-localized twin boundaries, general
grain boundaries and high angle grain boundaries (>80∘) for the CP-Ti
deformed at different tensile strains. Insets: representative strain-localized
X.H. Gong, Z.D. Feng and D. Fan: Methodology, Investigation,
twin boundary and strain-localized general grain boundary. Conceptualization. L. Lu: Visualization, Data curation, Formal analysis,
Supervision, Writing - original draft, Resources. S.N. Luo: Writing - re­
view & editing, Validation, Funding acquisition.
dislocation nucleation and tangling, resulting in strain concentration,
compared with the deformation twin boundaries.
Declaration of Competing Interest
4. Summary
The authors declare that they have no known competing financial
In this study, speckles at three different scales are prepared to interests or personal relationships that could have appeared to influence
investigate strain localization in titanium under uniaxial tension. The the work reported in this paper.
electro-polishing is used to fabricate microscale and mesoscale speckles,
and the macroscale speckles are prepared via spray painting. The Acknowledgments
microstructure evolution and strain fields during tension are obtained by
in situ EBSD, SEM DIC and optical DIC measurements. Our main This work was sponsored in part by the National Natural Science
Foundation of China (Grant Nos. 11902274 and 11627901), the

7
X.H. Gong et al. Materials Characterization 189 (2022) 111940

Fundamental Research Funds for the Central Universities of China commercial and open-source software packages, Mater. Charact. 163 (2020),
110271.
(Grant No. 2682021CX115) and Sichuan Science and Technology Pro­
[22] M.P. Echlin, J.C. Stinville, V.M. Miller, W.C. Lenthe, T.M. Pollock, Incipient slip
gram (Grant Nos. 2020YFG0139 and 2022YFG0033). and long range plastic strain localization in microtextured Ti-6Al-4V titanium, Acta
Mater. 114 (2016) 164–175.
References [23] T.A. Book, M.D. Sangid, Strain localization in Ti-6Al-4V widmanstätten
microstructures produced by additive manufacturing, Mater. Charact. 122 (2016)
104–112.
[1] S.D. Antolovich, R.W. Armstrong, Plastic strain localization in metals: origins and [24] B. Barkia, V. Doquet, E. Héripré, I. Guillot, Characterization and analysis of
consequences, Prog. Mater. Sci. 59 (2014) 1–160. deformation heterogeneities in commercial purity titanium, Mater. Charact. 108
[2] Y. Tanaka, K. Naito, S. Kishimoto, Y. Kagawa, Development of a pattern to measure (2015) 94–101.
multiscale deformation and strain distribution via in situ FE-SEM observations, [25] P. Baudoin, T. Hama, H. Takuda, Influence of critical resolved shear stress ratios on
Nanotechnology 22 (2011), 115704. the response of a commercially pure titanium oligocrystal: crystal plasticity
[3] C.C. Aydıner, M.A. Telemez, Multiscale deformation heterogeneity in twinning simulations and experiment, Int. J. Plast. 115 (2019) 111–131.
magnesium investigated with in situ image correlation, Int. J. Plast. 56 (2014) [26] L. Zeng, L. Wang, P. Hua, Z. He, G. Zhang, In-situ investigation of dwell fatigue
203–218. damage mechanism of pure Ti using digital image correlation technique, Mater.
[4] M. Koyama, K. Yamanouchi, Q. Wang, S. Ri, Y. Tanaka, Y. Hamano, S. Yamasaki, Charact. 181 (2021), 111466.
M. Mitsuhara, M. Ohkubo, H. Noguchi, et al., Multiscale in situ deformation [27] R. Sperry, S. Han, Z. Chen, S.H. Daly, M.A. Crimp, D.T. Fullwood, Comparison of
experiments: a sequential process from strain localization to failure in a laminated EBSD, DIC, AFM, and ECCI for active slip system identification in deformed Ti-7Al,
Ti-6Al-4V alloy, Mater. Charact. 128 (2017) 217–225. Mater. Charact. 173 (2021), 110941.
[5] T.C. Chu, W.F. Ranson, M.A. Sutton, Applications of digital-image-correlation [28] S. Hémery, J. Stinville, F. Wang, M. Charpagne, M. Emigh, T. Pollock, V. Valle,
techniques to experimental mechanics, Exp. Mech. 25 (1985) 232–244. Strain localization and fatigue crack formation at (0001) twist boundaries in
[6] J. Blaber, B. Adair, A. Antoniou, Ncorr: open-source 2D digital image correlation titanium alloys, Acta Mater. 219 (2021), 117227.
matlab software, Exp. Mech. 55 (2015) 1105–1122. [29] M.-S. Lee, M.-K. Ji, Y.-T. Hyun, E.-Y. Kim, T.-S. Jun, Effect of texture and
[7] L. Lu, D. Fan, B.X. Bie, X.X. Ran, M.L. Qi, N. Parab, J.Z. Sun, H.J. Liao, M. temperature gradient on anisotropic plastic deformation of commercially pure
C. Hudspeth, B. Claus, K. Fezzaa, T. Sun, W. Chen, X.L. Gong, S.N. Luo, Note: titanium at room and low temperatures, Mater. Charact. 172 (2021), 110834.
dynamic strain field mapping with synchrotron X-ray digital image correlation, [30] E. Ezugwu, Z. Wang, Titanium alloys and their machinability—a review, J. Mater.
Rev. Sci. Instrum. 85 (2014), 076101. Process. Technol. 68 (1997) 262–274.
[8] B. Pan, K. Qian, H. Xie, A. Asundi, Two-dimensional digital image correlation for [31] F. Lagattu, F. Bridier, P. Villechaise, J. Brillaud, In-plane strain measurements on a
in-plane displacement and strain measurement: a review, Meas. Sci. Technol. 20 microscopic scale by coupling digital image correlation and an in situ SEM
(2009), 062001. technique, Mater. Charact. 56 (2006) 10–18.
[9] V. Tarigopula, O. Hopperstad, M. Langseth, A. Clausen, F. Hild, A study of [32] F. Mathieu, F. Hild, S. Roux, Identification of a crack propagation law by digital
localisation in dual-phase high-strength steels under dynamic loading using digital image correlation, Int. J. Fatigue 36 (2012) 146–154.
image correlation and FE analysis, Int. J. Solids Struct. 45 (2008) 601–619. [33] R. Balokhonov, V. Romanova, A. Panin, M. Kazachenok, S. Martynov, Strain
[10] X.Z. Li, C. Li, L. Lu, J.Y. Huang, S.N. Luo, Interactions between slip bands and localization in titanium with a modified surface layer, Phys. Mesomech. 21 (2018)
interfaces in a compressed duplex stainless steel, Mater. Sci. Eng. A 818 (2021), 32–42.
141325. [34] L. Lu, Q.W. Xia, F. Zhao, J.C. Shi, S.N. Luo, Y. Cai, Mechanical properties and
[11] M. Eskandari, M. Yadegari-Dehnavi, A. Zarei-Hanzaki, M. Mohtadi-Bonab, R. Basu, strengthening mechanisms of shock-modified titanium alloy, Mater. Lett. 283
J. Szpunar, In-situ strain localization analysis in low density transformation- (2021), 128744.
twinning induced plasticity steel using digital image correlation, Opt. Lasers Eng. [35] A. Akhtar, Basal slip and twinning in α-titanium single crystals, Metall. Mater.
67 (2015) 1–16. Trans. A 6 (1975) 1105.
[12] J. Liu, N. Vanderesse, J.C. Stinville, T. Pollock, P. Bocher, D. Texier, In-plane and [36] A.A. Salem, S.R. Kalidindi, R.D. Doherty, Strain hardening of titanium: role of
out-of-plane deformation at the sub-grain scale in polycrystalline materials deformation twinning, Acta Mater. 51 (2003) 4225–4237.
assessed by confocal microscopy, Acta Mater. 169 (2019) 260–274. [37] J. Tan, L. Lu, H.Y. Li, X.H. Xiao, Z. Li, S.N. Luo, Anisotropic deformation and
[13] J. Stinville, M. Echlin, D. Texier, F. Bridier, P. Bocher, T. Pollock, Sub-grain scale damage of dual-phase Ti-6Al-4V under high strain rate loading, Mater. Sci. Eng. A
digital image correlation by electron microscopy for polycrystalline materials 742 (2019) 532–539.
during elastic and plastic deformation, Exp. Mech. 56 (2016) 197–216. [38] N. Jia, P. Eisenlohr, F. Roters, D. Raabe, X. Zhao, Orientation dependence of shear
[14] M. Marzouk, M. Jain, S. Shankar, Effect of Sr-modification on the bendability of banding in face-centered-cubic single crystals, Acta Mater. 60 (2012) 3415–3434.
cast aluminum alloy A356 using digital image correlation method, Mater. Sci. Eng. [39] J. Li, Y. Li, C. Huang, T. Suo, Q. Wei, On adiabatic shear localization in
A 598 (2014) 277–287. nanostructured face-centered cubic alloys with different stacking fault energies,
[15] A. Orozco-Caballero, D. Lunt, J.-D. Robson, J.-Q. Da Fonseca, How magnesium Acta Mater. 141 (2017) 163–182.
accommodates local deformation incompatibility: a high-resolution digital image [40] X. Hong, A. Godfrey, W. Liu, A. Orozco-Caballero, J.Q. da Fonseca, Effect of pre-
correlation study, Acta Mater. 133 (2017) 367–379. existing twinning on strain localization during deformation of a magnesium alloy,
[16] L. Lu, J.W. Huang, D. Fan, B.X. Bie, T. Sun, K. Fezzaa, X.L. Gong, S.N. Luo, Mater. Lett. 209 (2017) 94–96.
Anisotropic deformation of extruded magnesium alloy AZ31 under uniaxial [41] X.X. Guan, L. Lu, S.N. Luo, D. Fan, In situ observations of detwinning and strain
compression: a study with simultaneous in situ synchrotron x-ray imaging and localization in pure titanium, Mater. Sci. Eng. A 813 (2021), 141073.
diffraction, Acta Mater. 120 (2016) 86–94. [42] C. Efstathiou, H. Sehitoglu, J. Lambros, Multiscale strain measurements of
[17] K. Hazeli, J. Cuadra, P. Vanniamparambil, A. Kontsos, In situ identification of twin- plastically deforming polycrystalline titanium: role of deformation heterogeneities,
related bands near yielding in a magnesium alloy, Scr. Mater. 68 (2013) 83–86. Int. J. Plast. 26 (2010) 93–106.
[18] B. Zhu, X. Liu, C. Xie, J. Su, P. Guo, C. Tang, W. Liu, Unveiling the underlying [43] B. Pan, Z. Lu, H. Xie, Mean intensity gradient: an effective global parameter for
mechanism of forming edge cracks upon high strain-rate rolling of magnesium quality assessment of the speckle patterns used in digital image correlation, Opt.
alloy, J. Mater. Sci. Technol. 50 (2020) 59–65. Lasers Eng. 48 (2010) 469–477.
[19] K. Frydrych, T. Libura, Z. Kowalewski, M. Maj, K. Kowalczyk-Gajewska, On the role [44] M. Bornert, F. Brémand, P. Doumalin, J.-C. Dupré, M. Fazzini, M. Grédiac, F. Hild,
of slip, twinning and detwinning in magnesium alloy AZ31B sheet, Mater. Sci. Eng. S. Mistou, J. Molimard, J.-J. Orteu, L. Robert, Y. Surrel, P. Vacher, B. Wattrisse,
A 813 (2021), 141152. Assessment of digital image correlation measurement errors: methodology and
[20] P. Guo, Q. Tang, L. Li, C. Xie, W. Liu, B. Zhu, X. Liu, The deformation mechanism results, Exp. Mech. 49 (2009) 353–370.
and adiabatic shearing behavior of extruded Mg-8.0Al-0.1Mn alloy in different [45] G. Crammond, S. Boyd, J. Dulieu-Barton, Speckle pattern quality assessment for
heat treated states under high-speed impact load, J. Mater. Res. Tech. 11 (2021) digital image correlation, Opt. Lasers Eng. 51 (2013) 1368–1378.
2195–2207. [46] S. Balasubramanian, L. Anand, Plasticity of initially textured hexagonal
[21] D. Lunt, R. Thomas, M. Roy, J. Duff, M. Atkinson, P. Frankel, M. Preuss, J.Q. Da polycrystals at high homologous temperatures: application to titanium, Acta Mater.
Fonseca, Comparison of sub-grain scale digital image correlation calculated using 50 (2002) 133–148.

You might also like