International Journal of Plasticity 74 (2015) 92e109

Contents lists available at ScienceDirect

International Journal of Plasticity

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

An experimental study of the polycrystalline plasticity of

austenitic stainless steel
~o Quinta da Fonseca a, *
Fabio Di Gioacchino a, b, Joa
a
The School of Materials, The University of Manchester, Oxford Road, Manchester, M13 9PL, UK
b
Department of Materials Science and Metallurgy, University of Cambridge, 27 Charles Babbage Rd, Cambridge CB3 0FS, UK

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

Article history: The development and validation of crystal plasticity models requires the ability to map
Received 17 February 2015 deformation at the microstructural scale. Here, a new method of high-resolution defor-
Received in revised form 10 May 2015 mation mapping is used to measure strain, material rotation and lattice rotation in
Available online 23 June 2015
austenitic stainless steel at sub-micron resolution. Electron back-scatter diffraction maps
are used to link the deformation to the microstructure. Deformation occurs in domains, in
Keywords:
which most of the plastic strain originates from the activation of a single slip system with
B: Crystal plasticity
high resolved shear stress. Within domains, slip is localized in lamellar regions that in-
B: Polycrystalline material
A: Dislocations
crease in number with strain. The deformation incompatibility between grains that de-
A: Grain boundaries velops as a consequence of this single crystal like behaviour is accommodated by either a
Deformation domains gradient in slip intensity and the consequent development of lattice curvature at the grain
boundary or the activation of an additional high stressed slip system and the consequent
formation of a complementary deformation domain within the grain. In many cases,
however, lattice curvature across grain boundaries is small because the deformation do-
mains in neighbouring grains are compatible. The implications of these observations for
continuum crystal plasticity modelling are discussed.
© 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC
BY license (http://creativecommons.org/licenses/by/4.0/).

1. Introduction

1.1. Background

The mechanical performance of engineering alloys depends on their microstructure and process history. The failure of
these materials in service is usually caused by the accumulation of damage that occurs as a consequence of localized
deformation at the microstructural scale, the details of which depend on the microstructure, including grain size and texture,
second phases and deformation substructure (Dieter, 1986). In an attempt to better understand these processes and develop
models capable of using microstructural information to predict component performance, crystal plasticity finite element
modelling (CPFEM) is being increasingly used to model deformation at the microstructural scale (Barbe et al., 2001a) (Barbe
et al., 2001b), (Diard et al., 2005), (Roters et al., 2010), and even predict crack nucleation and growth, (McDowell and Dunne,
2010), (Anahid et al., 2011), (Bache et al., 2010), (Dunne, 2014). In CPFE modelling, continuum mechanics is extended to the

* Corresponding author.
E-mail address: joao.fonseca@manchester.ac.uk (J. Quinta da Fonseca).

http://dx.doi.org/10.1016/j.ijplas.2015.05.012
0749-6419/© 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/
4.0/).
 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109 93

microstructural scale by using a finite element mesh to represent the microstructure and constitutive laws that represent the
anisotropic nature of single crystal deformation. As the length scale decreases, however, the applicability of continuum
mechanics is increasingly limited by the discrete (in time and space) nature of plastic deformation, clearly demonstrated in
recent experiments on miniature single crystals by Uchic et al. (2004), Dimiduk et al. (2006) and by the development of
dislocation patterns in deformed metals, both single and polycrystalline, forming deformation substructures with charac-
teristic length scales of a few microns, e.g. Kuhlmann-Wilsdorf and Hansen (1991), Bay et al. (1992), Hughes (1993) and
Hansen et al. (2001).
This raises the question: at what length scale are crystal plasticity continuum mechanics models no longer capable of
representing the combined behaviour of dislocations and their interactions? This is an important question because the ability
of models to capture, for example, the effects of microstructure on crack initiation presumably relies on the ability of
accurately predict stress and strains at the sub-micron scale (Bieler et al., 2009). Answering this is also important from a
practical point of view, since decisions must be made regarding mesh refinement in CPFE models. A very fine mesh is
necessary for a very realistic model but it would be counterproductive to refine the mesh beyond the limits of continuum
mechanics. Understanding these limits would also help guide the development of higher-order crystal plasticity models
(Kuroda, 2014), (Anand et al., 2015) which have been primarily used to predict natural or intrinsic length scale effects on flow
stress and hardening of small samples, but might also be useful in extending continuum mechanics to modelling stresses and
strains at the microstructural scale during polycrystalline deformation.
The obvious way to understand the limits of current models is to compare predictions of local deformation to exper-
imental measurements made at the relevant length scale. A large number of such comparisons were considered in a recent
review (Pokharel et al., 2014), which included comparisons of lattice rotation and curvature at the sub-micron scale made
using electron back-scattered diffraction and at the micron scale using synchrotron X-ray diffraction. The review concluded
that, although crystal plasticity FE and other full-field models appear to describe mean grain behaviour well, the agreement
on a grain-to-grain basis is invariably poor. Although it is possible that this is a result of experimental limitations or poorly
defined modelling boundary conditions, this lack of agreement suggests that these models are unable to accurately describe
deformation at the microstructural scale. The reliance on lattice curvature data makes it difficult to probe this lack of
agreement between models and experiments further, since more than one strain distribution can be conceived to explain
the development of lattice curvature (Acharya and Bassani, 2000) (Cermelli and Gurtin, 2001) (Di Gioacchino and Clegg,
2014). Therefore, even if the predicted evolution of lattice orientation matches that observed, this does not guarantee
that the predicted evolution of strain is a faithful representation of deformation at this scale. To fully test numerical
predictions, combined strain and lattice rotation measurements are required. This is particularly important for higher order
plasticity theories, where the details of the development of strain gradients and the associated lattice rotations are
fundamental.
Whereas diffraction techniques can only measure lattice rotation, full-field displacement mapping techniques, such as
digital image correlation (DIC) can measure strain. These measurements can be used in combination with orientation
maps to relate plastic deformation to the microstructure and starting crystal orientation (Lineau et al., 1997), (Hoc and
Rey, 2000), (Hoc et al., 2003), (Jin et al., 2008), (Schroeter and McDowell, 2003) (He ripre
 et al., 2007) (Marteau et al.,
2012) as well as local Schmid and Taylor factors (Kammers and Daly, 2013), (Tschopp et al., 2009) (Tasan et al., 2014),
(Han et al., 2013) (Walley et al., 2012) and even texture and lattice curvatures after deformation (Lineau et al., 1997)
(Tatschl and Kolednik, 2003) (Zhao et al., 2008) (Zhang and Tong, 2004). However, until recently, the spatial resolution of
DIC measurements has been inferior to that of EBSD measurements and therefore insufficient to relate deformation to the
lattice rotation and curvature at the required sub-micron scale. This is primarily because suitable pattern application
methods are needed to generate sufficiently fine speckles on the surface of the tested sample to enable high resolution
DIC (HRDIC).
Direct deposition of nanometer sized solid particles of selected materials is emerging as the class of most versatile
pattern application methods for HRDIC. A comprehensive review of such methods can be found in (Kammers and Daly,
2011). In the recent work of (Lim et al., 2014) (Tasan et al., 2014) (McMurtrey et al., 2014) HRDIC maps obtained using
deposition of nanometre copper, colloidal silica and gold particles were shown to resolve localization of strain at the
microstructural scale in tantalum oligocrystals, duplex steel and irradiated austenitic steel, respectively. In the present
study, we use the alternative HRDIC method introduced in (Di Gioacchino and Quinta da Fonseca, 2013), which makes use
of remodelling of a deposited metallic film. The speckles produced in this way, as also described in (Joo et al., 2013) are
homogeneously distributed and attached to sample surface so that these cannot displace with respect to the latter during
deformation. Correlating SEM images of the speckled areas at sufficiently low scan speeds, which helps reducing the noise
associated with the raster scan (Sutton et al., 2007), gives accurate and reliable deformation maps that can be confidently
compared to EBSD measurements (Di Gioacchino and da Fonseca, 2013). In the study presented here, the latter HRDIC
method is used to study the polycrystalline deformation of a stainless steel and obtain dense deformation gradient maps
with sub-micron resolution and measure the shear deformation associated with slip, material rotation and compare it to
lattice rotation from EBSD, which is also used to link the deformation gradients to the microstructure. The objectives are to
provide details of the spatial distributions of both plastic strain and lattice rotation (and curvature) with sub-micron
resolution and interpret these results in the context of the different approaches to model deformation at this scale. The
aim is to provide new insights that can guide the development of future models and provide data that can be used in their
validation.
94 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109

Table 1
Chemical composition of the 304L stainless steel in weight %.

Fe C Si Mn Cr Ni P S
Bal. 0.021 0.34 1.96 18.15 9.17 0.031 0.027

2. Material and experimental methodology

2.1. Material

The material chosen for this study was 304L stainless steel. This is an important structural engineering material, widely
used in power plants due to its corrosion resistance, good strength and toughness. As can be seen in Table 1, 304L contains
high amounts of Cr and Ni, which give the steel a low stacking fault energy (SFE) of about ~18 mJ m
2 (Schramm and Reed,
1975). The chemical composition is given in Table 1. The material was tested in the solution-annealed condition (1 h at 1050  C
and water quenched).

2.2. Micro-tensile testing and image acquisition

A flat dog-bone shaped specimen with a 5 mm gauge length and 2 mm thick was strained to 1.6% and then 6% elongation
using a ZeisseKammrath micro-tester at constant strain rate of 4  10
3 s
1. At each strain level the test was interrupted, the
specimen was unloaded and removed from the microtester and mounted on the SEM stage for image acquisition.
Before testing, the surface had been covered with nano-scale speckle pattern, using the gold remodelling procedure
described by Di Gioacchino and Quinta da Fonseca (2013) to provide sub-micron spatial resolution. This comprises of pol-
ishing the sample using a diluted colloidal silica (OPS) solution for 10 min, applying a 50 nm thick gold film, and then
remodelling it to give a nanoscale speckle pattern. The region of interest had been marked with a series of micro-indents, to
act as fiducial marks and allow quick identification of the area of interest and alignment of the sample following deformation.
The images were acquired using a FEG SEM (Sirion, FEI). Image acquisition was performed at resolution of 2756  1936
pixels and at a magnification of 2000, which corresponds to a field of view of about 60  45 mm, and a spatial resolution of
22 nm per pixel. Optimal imaging is achieved with backscattered electrons, a relatively high beam voltage and small working
distance. The values used are reported in Table 2. In order to cover a larger area and include several grains in the mea-
surement, images were acquired for adjacent regions in a 3  3 mosaic. The images were slightly overlapped to enable
seamless stitching into a larger map covering in total an area of about 170  130 mm2.

2.3. HRDIC measurements

The commercial software DaVis (LaVision) was used to carry out DIC analysis of the images in the deformation sequence
using adaptive, FFT based cross-correlation with an interrogation window of 8  8 pixels and no overlap, which corresponds
to 0.17  0.17 mm2 and therefore a distance between measurement points of only 0.17 mm.
The DIC analysis provides dense maps of in-plane displacement vectors, which can be used to calculate the components of
the deformation gradient associated with the in-plane deformation. Taking X1X2 as the plane of investigation and X3 the
normal to such plane, and ui,j with i ¼ 1, 2, 3 the displacement field u(X1, X2, 0), the components of the deformation gradient
tensor F describing in-plane deformation are calculated as:

Fij ¼ ui;j þ dij (1)

Where dij is the Kronecker delta representing the identity tensor (i.e. dij ¼ 0 for isj and dij ¼ 1 for i ¼ j) and the term ui,j ¼ vui/
vXj is the displacement gradient tensor. The latter describes the change of the i-component of the displacement, ui, on an
increment of distance dXj along the direction j, which is equivalent to the pitch of the grid in DIC measurements. Eq. 1 shows
that the components of F are related to displacement gradients and are not themselves measures of strain.

2.3.1. Plastic strain

Since deformation is assumed to occur by slip induced shear we are interested in measuring shear strain, g ¼ F12 þ F21.
However, out of plane shear can occur, which would not be visible in an in-plane shear strain map. To mitigate this effect, we
use an effective maximum shear strain, gmax to represent the distribution of plastic strain in the area of interest. This scalar is
essentially the maximum shear strain in plane strain and is given by (e.g. Dieter, 1986):

Table 2
SEM scan parameters used for image acquisition.

Imaging ACC. Beam Working Dwell Magnification Image Spatial

mode Voltage (kV) Current (nA) distance (mm) (msec) size (pixels) resolution (nm)
BEI 22 1.1 5 40 2000 2756  1936 y22
 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109 95

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 
F11
F22 2 g2
gmax ¼ þ (2)
2 2

Clearly, this is not the actual maximum shear strain since deformation at the surface is not in plane strain, however, it
does have the advantage of including the apparent in-plane stretch associated with out of plane shear, as long as the shear
plane is not exactly perpendicular to the surface. That is, even if g is zero, F11 and or F22 will usually be non-zero and so will
gmax, i.e. all strain will be detected although the magnitude of shear will depend on the maximum shear direction, which is
unknown.

2.3.2. Material and lattice rotation

In the framework of small deformations, the antisymmetric part of the displacement gradient tensor gives the angle of
(rigid body) rotation about X3, u3, of the material line elements (Dieter, 1986):

1
u3 ¼ ðF
F21 Þ (3)
2 12

Unless pure shear occurs, for which F12 ¼ F21, shear generally gives rise to a material rotation. The latter differs from lattice
rotation in the kinematic model of crystal deformation (Asaro, 1983). For example, in the case of pure shear (where either F12
or F21 ¼ 0) due to slip on an active plane of trace X1 (or X2), there is material rotation because u3s0 but there is no lattice
rotation, that is w3 ¼ 0.
On the other hand, in a crystal undergoing non-homogenous deformation, lattice rotation can occur locally without
underlying shear, by rigid body rotation. In the case of single slip, this can be exploited to convert maps of material rotation
into maps of lattice rotation as described in detail in (Di Gioacchino and Clegg, 2014). In a region of the material deforming by
single slip with s being the slip trace on X1X2, if the direction X1 is made parallel to s then the lattice rotation is given by:

w3 ¼ F21=F ðwith X1 ≡sÞ (4)

11

where F11 is close to unity for small deformations. The variation of w3 was used here to describe the in-plane components of
the lattice curvature in regions of single slip and related to the deformation heterogeneity.

2.4. Acquisition and analysis of EBSD data

After image acquisition at the last loading step, the gold speckles were removed with few seconds of polishing with
0.25 mm diamond paste in preparation for lattice orientation measurements using EBSD. The sample surface was polished
further using 0.05 mm colloidal silica (OPS) for 5 min to improve the indexing of lattice orientation. In total, this removed
about 4 mm of material from the surface as deduced from the reduction in the size of the fiducial indents. A lattice orientation
map was then generated for the region of interest using a step size of 0.13 mm to give a map with a resolution comparable to
the HRDIC map. The acquisition was performed using (Channel 5 HKL, UK) software and the data analysed using Numerical
Python (Van der Walt et al., 2011) and plotted using Matplotlib (Hunter, 2007). The analysis included the correction for the
spatial distortion associated with the drifting of the electron beam due to charging of the sample during the EBSD mea-
surement, as shown in Fig. A1a of Appendix A.
EBSD measurements were used to define grain boundaries, identify phases, plot {111} plane traces and calculate values of
Schmid factor obtained for each slip system. EBSD was also used to validate the DIC measurements of w3 in regions where
there is no shear. The method used and the experimental limitations associated with the comparison of DIC and EBSD
measurements are also described in Appendix A.

3. Results

3.1. Plastic strain

Maps of gmax obtained for the region of interest after elongations of 1.6% and 6% are shown in Fig. 1a and b, respectively.
These show that the distribution of shear strain is heterogeneous from the outset. Deformation takes place by the formation
of thin shear bands at high angles with the loading direction. The bands form in groups with a characteristic band orientation
and a regular spacing of a few microns. These groups are seen to cover convex regions of material, which we refer to as
deformation domains.
As deformation progresses, the intensity of bands increases but the domains remain mostly unchanged and consequently
strain gradients become more intense at domain boundaries. The density of bands also increases and the amount of shear in
the new bands approaches that in the older, neighbouring bands. In addition, new deformation domains appear near the
edges of existing ones. These complementary deformation domains contain bands at high angles to those in the primary
96 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109

Fig. 1. Maximum shear strain maps at (a) 1.6% and (b) 6% strain. (a) White lines: location of the unit distance used to calculate the spacing between the bands. (b)
Dotted line: complementary deformation domain inside a parent domain (solid line). Region (A): undulated bands. Dashed lines (B1-4): bands change direction
abruptly. Regions (C1-3): elongated regions where bands fade.

deformation domain and can be seen at 6% macroscopic strain, Fig. 1b. In contrast, some regions, which appear dark in Fig. 1b,
have not deformed at all even though the applied macroscopic strain is 6%.
Careful observation reveals that some bands are not straight. In some cases bands contain undulations, e.g. region A. Since
the values in Fig. 1b are mapped in the initial configuration, these deformation bands must be curved from the outset. In other
cases the bands appear to change orientation abruptly along narrow regions (B1-4), either seamlessly (B1 and B2) or with
some degree of discontinuity (B3 and B4). There are also three elongated regions (C1-3) in which the bands appear to fade.
Profiles of maximum shear strain obtained at 6% elongation for three selected deformation domains are shown in Fig. 2.
The calculated maximum shear strain peaks at just above 0.3, whereas between the bands it drops down to below 0.02, which
is the order of magnitude of the noise in this kind of DIC measurements (Di Gioacchino and da Fonseca, 2013). Since for a
uniform extension of 6% the maximum ineplane shear strain is 0.045, this implies that the strain rates within the bands are
nearly an order of magnitude larger than the applied strain rate.

3.1.1. Plastic strain and the microstructure

The EBSD data in Fig. 3a was used to generate a grain boundary map that is shown in Fig. 3b. The region studied contains 26
grains, and about 1% of the area indexed as BCC iron (ferrite). There are two types of ferrite regions: elongated ferrite stringers,
for example the one crossing grain 2, the top of grain 23 and grain 4 and twin like features, for example in grains 6 and 9,
which presumably are martensite plates formed during deformation. Over half the grains also contain microtwins or
microtwin-like features, which, unlike the annealing twins, are known to form during deformation (El-Danaf et al., 1999).
 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109 97

Fig. 2. Maximum shear strain profiles for selected domains in the 6% strain map calculated using a three points moving average.

The map is overlaid onto the strain map in Fig. 4, allowing the deformation to be related to the microstructure. Since the
EBSD map was obtained in the deformed state, the strain map is now shown in the deformed configuration (note the stretch
along the loading direction compared to Fig. 1b). The first observation is that parent deformation domains correspond to grains
and that complementary deformation domains are usually found near grain boundaries. Regions such as (B1-4) where bands
were seen to abruptly change direction yet also correspond to grain boundaries. Microtwins and martensite bands (e.g. grains 6
and 9) lay on top of straight segments of shear bands. Furthermore, the ferrite stringer in the left-top corner corresponds to the
elongated region of fading bands (C1) in Fig. 1b. This suggests that the other elongated regions of locally reduced strain (C2 and
C3) are associated with ferrite stringers that were either removed by polishing or exist just below the surface.
EBSD data was used to calculate the orientation of slip traces, which are drawn using a blue-green-yellow-red colour
scheme, indicating the relative value of the corresponding Schmid factor with blue being lowest and red highest. Analysis for
all the 26 grains studied shows that the bands are aligned with a slip trace corresponding to a highly stressed {111} <110> slip
system, with the exception of grains 6, 7 and 26. In particular, the bands belonging to parent deformation domains are aligned

Fig. 3. EBSD map of the region of interest acquired after 6% deformation. (a) IPF map (b) Grain boundary map with BCC iron (ferrite) coloured in red. (For
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
98 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109

Fig. 4. Approximate positions of grain boundaries in Fig. 3. Deformation twin boundaries and transgranular ferrite stringers are removed for improved visibility.
{111} plane traces are also superimposed. Traces are coloured (blue, green, yellow, red) according to increasing values of Schmid factor, see Table 3. (For
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

with the traces of highest Schmid factor (red traces) with exception of those in grains 3, 5, 11 and 21 that are aligned with the
traces of second highest value (yellow traces) instead. The values of Schmid factor reported in Table 3 nevertheless show that
the first and second highest values are within only about 5% difference in the latter grains, that is, these are oriented for
double slip. This is also true for all the other grains showing complementary deformation domains, which are underscored in
Table 3, with the exception of grains 17 and 22. Peculiar among these is grain 26, which had slip bands aligned with traces of
{111} planes of smallest Schmid factor and which shows similar values for all the four {111} planes.
Measuring the angle between the active (111) planes and the plane of investigation, a, made it possible to calculate the
effective spacing, d, between the active planes as d ¼ ds sina, with ds being the spacing measured at the surface of the sample.
Values of ds were obtained by counting the number of bands within a reference distance of about 17 mm and only for bands
regularly spaced within single deformation domains, Fig. 1a. Results show that band spacing varies by a greater extent at 1.6%
than 6% strain at which an average of 16 bands per 17 microns are counted, therefore giving 1.06 mm spacing. This suggests
that the number of bands increases until a characteristic band spacing is reached, in a way consistent with early observations
of single crystal deformation (Orowan, 1941). A marked variation in band spacing is nevertheless possible within single grains,
as measured for grain 23 (regions 23a and 23b in Table 3), where the band spacing decreases in the region where they curve.

3.2. Material rotation

Values of u3 at 6% macroscopic strain are plotted in Fig. 5. The shear bands observed in the maximum shear maps are now even
more pronounced and are associated with alternating values of u3 from þ0.15 to
0.15. These are indicative of the material rotation
associated with shear and provide information on its direction. In domains where the bands have a positive slope, u3 is negative in
the bands showing that the sense of shear here is counter clockwise (negative) whereas it is positive in the material between the
bands, which therefore rotates clockwise. For bands with a negative slope, the opposite is true. Moreover, it appears that the higher
the intensity and density of bands, i.e. the higher the amount of shear in a region, Fig. 1b, the higher u3 between the bands. This is
evident, in particular, for grains 21, 22 and 23. this variation in shear intensity is more or less constant across these three grains,
forming a transgranular deformation zone perpendicular to slip band direction like a kink band, This transgranular heterogeneity is
an important feature which becomes even clearer when larger areas are considered (Di Gioacchino and da Fonseca, 2013).
 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109 99

Table 3
List of Schmid factor (SCM) values for {111} <110> slip systems and spacing of bands for each grain.

GRAIN SCM 1 (max) SCM 2 SCM 3 SCM 4 (min) Avg. Spacing (mm) 1.6%/6% Angle (radians) Actual spacing
1 0.49 0.44 0.22 0.11 1.36
2 0.47* 0.35 0.29 0.01 1.4/1.0 1.00 1.2/0.8
3 0.47 0.46* 0.40 0.34 3.4/1.2 1.18 3.4/1.1
4 0.47 0.37 0.21 0.02 1.18
5 0.43 0.40* 0.31 0.03 1.4/1.0 1.57 1.4/1.0
6 0.45 0.31 0.24 0.07 1.14
7 0.48 0.39 0.27 0.04 0.97
8 0.49* 0.42 0.31 0.11 2.4/1.1 1.54 2.4/1.1
9 0.43* 0.30 0.26 0.08 1.9/1.4 1.47 1.9/1.4
10 0.47* 0.37 0.36 0.07 3.4/1.4 1.01 1.7/1.2
11 0.45 0.43 0.09 0.05 0.92
12 0.47 0.45 0.42 0.33 1.56
13 0.32 0.26 0.24 0.08 1.14
14 0.49 0.46 0.39 0.26 0.93
15 0.48 0.39 0.35 0.09 0.90
16 0.48 0.39 0.36 0.10 1.28
17 0.49 0.41 0.33 0.11 1.08
18 0.46* 0.33 0.24 0.05 4.2/1.5 0.93 3.4/1.3
19 0.48 0.39 0.24 0.02 1.50
20 0.42 0.39 0.27 0.02 0.87
21 0.49 0.48* 0.35 0.32 2.8/1.0 1.39 2.8/1.0
22 0.39 0.38 0.22 0.07 0.54
23a** 0.44* 0.32 0.11 0.08 4.2/1.0 1.07 3.4/1.0
23b 0.45* 0.35 0.16 0.06 2.1/0.9 1.10 1.9/0.8
24 0.48 0.45 0.18 0.12 1.20
25 0.49 0.45 0.23 0.13 1.23
26 0.44 0.44 0.42 0.39 0.71
Mean ± STD (2.3 ± 0.8)/(1.0 ± 0.2)

Bold values of Schmid factor indicate active plane(s).

__Grains showing a second (complementary) active slip plane.
*Planes considered for measuring the actual band spacing.
**Values used to include grain 23 in the measurement of the mean spacing.

3.3. Lattice rotation and curvature

The calculation of lattice rotation using Eq. (4) was performed on a domain-to-domain basis by taking X1 as the direction of
the bands. The fact that the latter are near ±45 made it possible to generate two complementary seamless maps of w3: one for
bands near 45 , Fig. 6a, and one for those at
45 , Fig. 6b. Profiles 1e4 in Fig. 7 show that the variation of w3 is consistent with
that obtained using EBSD in Fig. 3a, providing confirmation that Eq. 4 succeeded in separating the rotation of the lattice from
the material rotation. The operation conserves the values of u3, that is w3 ¼ u3, in the material between the bands demon-
strating that its rotation coincides here with the rotation of the lattice. At the bands, values of w3 blend in with those obtained
for the surrounding material.
Deformation-induced lattice curvature is evidently greater at the interface between deformation domains showing bands
of opposite slope and that therefore rotate in opposite senses. For instance, at the boundary between grains 4 and 5, w3 goes
from about þ0.05 to
0.05 within a distance of about 3 mm, which gives a curvature of 0.033 radians (1.9  ) per micron. On the
other hand, the rotation of the lattice appears uniform across the grain boundaries separating deformation domains that
show bands with slopes of equal sign, such as grains 4, 23 and 24 or grains 8, 9 and 10. These grain boundaries include those in
regions (B1-4) in Fig. 1b where bands of adjacent domains meet to form continuous bands.
Within single deformation domains, the regions of greater lattice distortion correspond to the undulating bands in grain 23,
Fig. 6a, giving rise to a plume-shaped feature. This region is reproduced in Fig. 8 using a different range of values to further
accentuate the local feature. Similar features can be seen across the maps, such in grains 24 and 5 in Figs. 6a and 5b, respectively.
The plumes originate close to the triple points where single slip bands from the adjacent domains meet the boundaries. This
phenomenon has been documented elsewhere by combining EBSD observations and SEM images of slip bands (Britton and
Wilkinson, 2012). Here, the DIC data provide experimental evidence of the link between the (sense and intensity of) shear car-
ried by the band and the local rotation of the lattice induced inside the adjacent grain. It also shows that lattice curvature is
achieved not only by the activation of different slip systems but also by varying the intensity and spatial distribution of slip on only
one slip plane. This is a kind of kink band (Orowan, 1942) similar to the punch effect observed on single crystals by Jillson (1950).
Taking grain 21 as an example, additional parallels can be made between the strain heterogeneity and the lattice curvature.
In this grain, complementary domains sharing the same direction of slip bands surround the primary deformation domain,
appearing near the grain boundary on the left, at the top and at the top right, Fig. 9a. The parent domain rotates clockwise
whilst the complementary domains rotate anticlockwise, as shown in Fig. 6 and reproduced in Fig. 9. Combining these ob-
servations gives a map of w3 matching that calculated using EBSD, Fig. 9d. This shows that, in this case, lattice curvature
develops within the grain, separating the material in the centre from that near the boundary, like a core and mantle. Although
100 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109

Fig. 5. HRDIC map of the material rotation, u3, following 6% macroscopic tensile strain along the horizontal direction. Grain boundaries in Fig. 3 are overlapped.

this looks similar to the mantle-core model proposed by Meyers and Ashworth (1982), the origin of this mantle and core is
very different from that in their model. Whereas Meyers and Ashworth propose that the mantle forms first and is harder due
to elastic anisotropy and incompatibility of neighbouring grains, our results show that yielding seems to start in the core, with
the mantle forming later when complementary slip is activated.

4. Discussion

4.1. Deformation heterogeneity

The high-resolution strain maps show that plastic deformation in 304 stainless steel is very heterogeneous right from
the start of uniaxial tensile deformation. It occurs as a discontinuous periodic variation in shear strain, with a wavelength of
a few microns, creating micro shear bands of regular spacing separated by regions with little or no strain. This is a
significantly different picture from that provided by previous DIC work such as the early work by Lineau et al. (1997) but
also the more recent work of Tatschl and Kolednik (2003), He ripre et al. (2007) and Tschopp et al. (2009). The strain
heterogeneity maps obtained in aluminium by Kammers and Daly (2013) or in duplex steel by (Tasan et al., 2014) are more
similar to ours, but the slip bands here are sharper and straighter. There is also some similarity with the maps in an Fe-Cr
alloy by Patriarca et al. (2013) although the slip bands in their work are much wider, presumably due to the lower reso-
lution of the measurements.
The observed distribution of shear strain resembles the typical deformation structures observed in low SFE materials
after deformation. This is made especially clear when the strain map is reduced to a skeleton of the slip bands as shown in
Fig. 10. The arrangement of slip bands into domains, separated by intergranular domain boundaries is consistent with the
Taylor lattice dislocation structures described by (Hughes, 1993) in Ale5.5 Mg, which also has low SFE. However, whereas
these dislocation structures were thought to develop gradually with homogeneous plastic deformation, our observations
imply that they are actually a consequence of a periodic variation in shear strain, which occurs naturally from the early
stages of plasticity. This suggests that plasticity in these materials is non-convex at this scale (Ortiz and Repetto, 1999). In
non-convex plasticity, non-homogeneous lamellar deformation is a lower bound solution to the variational problem of
elasto-plastic deformation by slip (Conti and Theil, 2003). In these formulations, rank 1 laminates are the lowest energy
configuration, which is in agreement with our observations. Non-convex plasticity has been used successfully to predict the
 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109 101

Fig. 6. HRDIC maps of lattice rotation, w3 for regions crossed by bands with (a) positive slope (<90 ) and (b) negative slope (<
90 ). Grain boundaries obtained
using EBSD in Fig. 3 are overlapped.

spacing between dislocation walls in heavily deformed microstructures (Aubry and Ortiz, 2003). Our results imply that
these deformation patterns start developing from very low applied strains instead of being created by dislocation entan-
glement and non-conservative rearrangement of uniformly distributed dislocations consistent with homogenous plastic
deformation.
Comparing the strain maps at different applied macroscopic strains shows that the deformation patterns evolve with
deformation but that their main characteristics are established very early on. Of course the appearance of slip lines on the
surface of polished metals after deformation is well known and was extensively studied in the 1940s (Brown, 1952), although
this early work relied on the appearance of slip steps on the surface of mostly single crystals at higher strain levels. One of the
questions at the time was whether slip localization did occur at strains of a few %, which we have clearly demonstrated here.
Our observations also provide direct evidence that, even in polycrystalline deformation, local strain rates are much higher
than the macroscopic strain rates. This is consistent with well-known experimental results on small single crystals that
showed that plasticity occurs via bursts of dislocation avalanches separated by time and space (Uchic et al., 2004), (Dimiduk
et al., 2006) and (Zaiser, 2006). Our results show that the constraint imposed by neighbouring grains does not cause
deformation to become homogeneous within grains. Instead, a quasi-regular pattern of lamellar deformation develops with a
well-defined spacing, arranged in packets, making up deformation domains.

4.2. Slip system activation and hardening

The slip bands measured are invariably well aligned with a highly stressed {111} slip plane trace. This implies that they are
the surface intersection of thin lamellas aligned parallel to the active slip plane. The high degree of crystallographic alignment
is probably due to the low SFE of the steel, which makes cross slip difficult and therefore promotes planar slip. It is important
102 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109

Fig. 7. Plots showing the variation of HRDIC and EBSD measured values of w3 for the profiles 1ein Figs. 3a and 6.

to note that it is not possible, from these observations alone, to rule out the activation of slip systems that share the same slip
trace, co-planar or not. The activity of non-coplanar systems can be ruled out in this case since most of the grains studied do
not contain slip planes with identical traces. In those that do, like grain 21 and grain 11, the Schmid factor is much lower in one
of them, making its activation unlikely. Ruling out co-planar slip is not possible without further measurements. However, in
most cases, the Schmid factor for one of the possible co-planar slip systems is much higher than that for the other two and
therefore it should dominate. This is consistent with the dislocation structures found in deformed material (Hughes, 1993).
The maps show that deformation in each grain is dominated by slip on one slip plane, which defines the primary
deformation domain. Therefore, although the applied strain is uniaxial tension, the deformation in most of each grain is best
described by in-plane shear, with the shear plane parallel to the slip plane. This is true from very low levels of strain, i.e. there
is no visible cross-slip before single plane slip (and on occasion double plane slip) occurs. This is at odds with the usual pattern
of slip activity predicted by continuum crystal plasticity FE models. A good example is the work on copper by (Musienko et al.,

Fig. 8. Detail of grain 23 in Fig. 6a highlighting a plume-like feature of local lattice distortion.
 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109 103

Fig. 9. Detail of grain 21 having complementary deformation domains at its boundary. (a) Material rotation. (b) and (c) HRDIC derived lattice rotation using the
direction of slip parallel to the primary bands (left) and secondary bands (right). (d) EBSD measured lattice rotation.

2007) that assumes the initial slip resistance to be the same in all slip systems. It predicts that at the start of deformation
many non co-planar slip systems are activated, especially near the grain boundaries where the incompatibility between
grains needs to be accommodated, and as the strain increases the deformation within a grain quickly evolves to be dominated
by a single slip system, through latent hardening. This is in contrast to classical single crystal plasticity where a reduction in

Fig. 10. Skeleton of Fig. 1b showing that the slip bands form networks that resemble a Taylor lattice.
104 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109

the number of slip systems does not occur until strains higher than 0.3 (Wu et al., 1991). In contrast, our results show that the
number of active slip planes increases with strain rather than decrease, as secondary deformation domains develop. The
grains appear to behave as if they were single crystals although, unlike a single crystal, the activation of secondary slip planes
occurs only in certain regions of the grain and at small strains, before significant crystal rotation occurs. This is probably a
consequence of local stress fluctuations caused by interactions with neighbouring grains. The observation that the material
rotation between the slip bands within each deformation domain coincides with the lattice rotation is further evidence of
single crystal like behaviour.
This difference in predicted vs observed slip system activity helps explain why the texture evolution in materials with low
SFE metals is less well predicted using the Taylor model, in which five slip systems are assumed to be active, and better
predicted by models that assume single crystal-like behaviour, like the Sachs models (Leffers and Ray, 2009) especially when
modified to account for local random stress fluctuations (Leffers, 1979). When CPFEM is used to model texture evolution,
strong latent hardening is required to obtain the low SFE textures, but this approach fails at larger strains (Miraglia et al.,
2007) possibly because the latent hardening formulation prevents the occurrence of complementary slip required to form
secondary deformation domains. To capture these effects, more detailed hardening formulations will be required, like those
developed for single crystal deformation and which are capable of giving a high initial latent hardening, which gradually
reduces (Wu et al., 1991) (Bassani and Wu, 1991).
The difficulty with polycrystalline deformation with only one slip plane per grain is that it should lead to incompatibilities
at grain boundaries. In fact, this is the principal objection to using the Sachs model to predict texture, despite the fact that it
makes better texture predictions for low SFE metals (Kocks, 2000) (Leffers and Ray, 2009). Our results show how these in-
compatibilities are accommodated in practice and the effect they have on the development of lattice curvature.

4.3. The origins of lattice curvature in polycrystals

Since in low SFE materials slip bands are made up of arrays of edge dislocation dipoles, they produce almost no lattice
curvature. As a consequence the bands are not visible in the EBSD lattice rotation maps. However, since the lattice rotations of
neighbouring deformation domains are different, pronounced lattice curvatures develop between domains, increasing the
mosaicity within grains.
The lattice rotation within domains depends on the spacing between the shear bands, the intensity of the shear within the
bands and finally by the orientation of the bands, as shown schematically in Fig. 11. Within a grain, these changes in the local
slip character must be caused by the interactions with neighbouring grains. Therefore, the development of deformation
domains is a non-local effect and is not determined by grain orientation alone. If the deformation is compatible across grain
boundaries, then deformation domains can extend across one or more grains. If not, then grain boundaries coincide with
deformation domain boundaries and since slip occurs in discrete volumes and only on 1e2 slip systems, incompatibilities
must develop between grains that must somehow be accommodated.
Before focussing on the incompatibilities, it is interesting to note that, in many cases, there is limited or no incompatibility at
grain boundaries because the orientation of bands is similar either side of it. Such compatibility requires that neighbouring
grains are suitably oriented, which happens to be the case in the region studied here. However, in an f.c.c. metal, a grain will often
contain 2 highly stressed complementary slip systems for which the Schmid factor is similar. It is therefore possible to activate
either of these slip systems, depending on the stresses imposed by neighbouring grains. Once deformation starts, slip in one
grain can influence slip selection in a neighbouring grain, leading to the activation of a compatible slip system. This reduces the
incompatibility at the grain boundaries and hence minimises the amount of lattice curvature generated during deformation.
At the boundaries where incompatibilities do arise, they are accommodated in two main ways. In the first, the intensity of
slip decreases as the bands approach the grain boundaries, giving rise to lattice curvature. This is equivalent to the creation of

Fig. 11. Deformation domains and lattice rotation in a grain. In domain 1 all slip bands have the same orientation but the spacing and slip intensity varies, give
rise to a gradient in lattice orientation. The slip bands in domain 2 are perpendicular to those in domain 1 and therefore lattice rotation occurs in the opposite
direction. Domains 3 and 4 do no rotate within the grain but for different reasons: in 3 there is no slip, whereas in 4 two balanced slip systems are active giving
rise to no lattice rotation.
 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109 105

geometric necessary dislocations as proposed by (Ashby, 1970) and is shown schematically in Fig. 12aec. In some cases, the
magnitude of slip on the impinging bands does not change and instead cause a change in the distribution of slip in the
neighbouring grain, causing slip bands to bend, increase in intensity and decreasing their spacing, therefore promoting cur-
vature of the lattice. A decrease in the band spacing also signifies that stress concentrates here during deformation (Orowan,
1941). In both cases lattice curvature is generated near the grain boundary by varying the local strain but preserving single slip.
The second way in which incompatibilities are accommodated is by the activation of a complementary slip system as
shown schematically in Fig. 12def. This corresponds to the formation of a complementary deformation domain. In this case,
the change in lattice rotation occurs not at the grain boundary but at the boundary between deformation domains. This leads
to a breakup of the grain and because deformation in each domain is dominated by only one or two slip systems, large
misorientations are generated within grains, with associated high densities of GNDs.
In all cases, although slip is clearly discontinuous, the changes in lattice rotation are not. This implies that GNDs making up
domain boundaries are generated by the differences in lattice rotation of neighbouring domains and are not debris from slip
made up of unbalanced statistically stored dislocations. This is important because polycrystalline crystal plasticity models
usually only minimize for the work dissipated by slip without considering the development of deformation domains. This
suggests that models with extra degrees of freedom like Cosserat or strain gradient plasticity models (Forest, 2008) would be
required to simulate these non-local effects. However, since in most implementations strain gradients only modify the hard-
ening terms, it is not clear to what extent they can simulate the predominance of single slip and the associated development of
deformation domains. Predictions using first order crystal plasticity (Barbe et al., 2001a), (Pokharel et al., 2014) are that lattice
curvature is always highest at the grain boundaries, suggesting that deformation domains do not emerge in these models.

5. Conclusions and implications for modelling plasticity at the microstructural scale

By combining high-resolution DIC (HRDIC) and EBSD we were able to study in detail the development of plastic strain
at the microstructural scale and link it to the development of lattice curvature. Deformation occurs by the formation of slip
bands well aligned with highly stressed slip systems. Deformation is lamellar in nature at low strains (1.6%), which
supports the idea that plasticity at this scale is non-convex, with near instantaneous infinitely strong latent hardening. The
high resolution of the strain measurements and the strong crystallographic alignment of the slip bands suggest that
deformation is dominated by single slip. This leads to the formation of deformation domains separated by boundaries of
high lattice curvature. Incompatibilities between grains are accommodated by variation of slip intensity or the formation
of complementary deformation domains but without ever activating more than 2 slip systems. This single crystal like

Fig. 12. Mechanisms of deformation-induced lattice curvature observed. (a)e(c) Lattice curvature associated with gradient of slip on a slip band. (d)e(f)
Equivalent curvature associated with the activation of a second slip system.
106 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109

behaviour is consistent with the ability of single slip models like the Sachs model to predict the texture evolution in these
materials.
Current full field crystal plasticity models do not predict the lamellar nature of deformation and since this strain locali-
zation is certain to be accompanied by high local stresses (Ahmed et al., 2014), the ability to predict the local stress using these
models is limited, which explains why grain boundary failure trends determined experimentally cannot be related to local
predicted stresses (Diard et al., 2005), (Gonzalez et al., 2014). There still remains the question to what extent the smooth,
continuous deformation fields predicted by these models are representative of the average behaviour of a volume containing
several slip bands. Our results suggest that only partial agreement should be possible and offer a possible explanation for why
crystal plasticity models are able to predict mean grain behaviour, which is dominated by grain orientation and slip activity
but fail to predict the behaviour of individual grains, where the development of deformation domains dominates.
In any case, it would appear that there is much to be gained by studying the development of deformation patterns at the
microstructural scale and that strain measurements with sub-micron spatial resolution provide unique insights. This work
should be extended to different microstructures, deformation modes and materials with different stacking fault energies.
These results should prove invaluable to the development of higher order crystal plasticity models and to better predictions of
stresses and strains at the microstructural scale.

Acknowledgements

The authors would like to thank Prof. William J. Clegg for the recent discussions and comments. Dr. Xin Sun is also
acknowledged for her invite to discuss the paper at the symposium “Microstructure Based Property Prediction and Small
Scale Experimental Validation” held in October 2012 in Pittsburgh, US. This was occasion for early comments and useful
discussions. The authors thank Serco TCS (UK) for funding and support.

Appendix A

Because of the slow scanning during the EBSD measurement and unexpected charging of the sample, significant beam
drift occurred during the EBSD measurement. This was corrected by determining the shape of the region measured using SE
imaging and using it to correct the EBSD map as shown in Fig. A1.
Quaternion algebra is used for the ease offered in performing operations on rotations. The four components of the unit
quaternion q, jqj ¼ 1, representing the rotation R(w, r) about the unit vector r can be constructed as:
   w  
w
 
w
  
w
q ¼ q0 ; q ¼ cos ; sin r1 ; sin r2 ; sin r (A1)
2 2 2 2 3

Using the convention in (Pantleon, 2008) a rotation q1 followed by a second rotation q2 is equivalent to a rotation p ob-
tained as the quaternion product q2q1 defined as:
 
p ¼ q2 q1 ¼ q02 q01
q2  q1 ; q02 q1 þ q01 q2 þ q2  q1 (A2)

Where the symbols $ and  represents the dot the vector product, respectively.
The conjugate of q, q* ¼ (q0,
q), represents the inverse rotation. It is therefore possible to describe the rotation m that
would follow q1 to give q2, in sample coordinates, as:

m ¼ q*1 q2 (A3)

Eq. A3 can be used to calculate the rotation of the lattice at one point taking the initial orientation as q*1 and the current one
as q2 and considering lattice symmetry. As lattice rotations are generally small, these can be treated as infinitesimal rotations
and their components wi about the principal axes calculated as (Pantleon, 2008):
 
2$arccos m0
wi ¼ mi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 2 (A4)
1
m0

To reveal the curvature of the lattice within a grain, it is therefore possible to take the lattice orientation at an arbitrary
point as the reference orientation q*1 and compute wi for the remaining points. This was achieved here for w3; the results are
shown in Fig. A2. It is nevertheless to note that an exact agreement between the DIC and EBSD measurements is not expected
because of experimental limitations. In order to perform the EBSD after testing, polishing is in fact required to remove the gold
pattern and any (micro) topography occurred with deformation. Therefore, material removal is generally uneven and the two
techniques probe different layers of material. This is believed to be the cause of the vertical features that appear in grains 4
and 23 in Fig. A2.
 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109 107

Fig. A1. (a) Example of the marking caused by the EBSD measurement showing the actual shape of the region covered due to drifting of the electron beam for the
material and SEM settings used here. Deformation induced topography is also visible (b) EBSD map corrected for the spatial distortion.

Fig. A2. EBSD measurements of the in-plane component of lattice rotation, w3, calculated with respect to reference orientations on a grain-to-grain basis. Note
difference in the colour scheme of the scale bars if comparing to Fig. 6.

References

Acharya, A., Bassani, J.L., 2000. Lattice incompatibility and a gradient theory of crystal plasticity. J. Mech. Phys. Solids 48, 1565e1595. http://dx.doi.org/10.
1016/S0022-5096(99)00075-7.
Ahmed, I.I., Grant, B., Sherry, A.H., Quinta da Fonseca, J., 2014. Deformation path effects on the internal stress development in cold worked austenitic steel
deformed in tension. Mater. Sci. Eng. A 614, 326e337. http://dx.doi.org/10.1016/j.msea.2014.07.005.
Anahid, M., Samal, M.K., Ghosh, S., 2011. Dwell fatigue crack nucleation model based on crystal plasticity finite element simulations of polycrystalline
titanium alloys. J. Mech. Phys. Solids 59, 2157e2176. http://dx.doi.org/10.1016/j.jmps.2011.05.003.
Anand, L., Gurtin, M.E., Reddy, B.D., 2015. The stored energy of cold work, thermal annealing, and other thermodynamic issues in single crystal plasticity at
small length scales. Int. J. Plast. 64, 1e25. http://dx.doi.org/10.1016/j.ijplas.2014.07.009.
Asaro, R.J., 1983. Crystal plasticity. J. Appl. Mech. 50, 921e934. http://dx.doi.org/10.1115/1.3167205.
Ashby, M.F., 1970. The deformation of plastically non-homogeneous materials. Philos. Mag. 21, 399e424. http://dx.doi.org/10.1080/14786437008238426.
Aubry, S., Ortiz, M., 2003. The mechanics of deformationeinduced subgrainedislocation structures in metallic crystals at large strains. Proc. R. Soc. Lond. Ser.
Math. Phys. Eng. Sci. 459, 3131e3158. http://dx.doi.org/10.1098/rspa.2003.1179.
Bache, M.R., Dunne, F.P.E., Madrigal, C., 2010. Experimental and crystal plasticity studies of deformation and crack nucleation in a titanium alloy. J. Strain
Anal. Eng. Des. 45, 391e399. http://dx.doi.org/10.1243/03093247JSA594.
Barbe, F., Decker, L., Jeulin, D., Cailletaud, G., 2001a. Intergranular and intragranular behavior of polycrystalline aggregates. Part 1: F.E. model. Int. J. Plast. 17,
513e536. http://dx.doi.org/10.1016/S0749-6419(00)00061-9.
Barbe, F., Forest, S., Cailletaud, G., 2001b. Intergranular and intragranular behavior of polycrystalline aggregates. Part 2: results. Int. J. Plast. 17, 537e563.
http://dx.doi.org/10.1016/S0749-6419(00)00062-0.
Bassani, J.L., Wu, T.-Y., 1991. Latent hardening in single crystals II. Analytical characterization and predictions. Proc. Math. Phys. Sci. 435, 21e41.
Bay, B., Hansen, N., Hughes, D.A., Kuhlmann-Wilsdorf, D., 1992. Evolution of f.c.c. deformation structures in polyslip. Acta Metall. Mater. 40, 205e219. http://
dx.doi.org/10.1016/0956-7151(92)90296-Q.
108 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109

Bieler, T.R., Eisenlohr, P., Roters, F., Kumar, D., Mason, D.E., Crimp, M.A., Raabe, D., 2009. The role of heterogeneous deformation on damage nucleation at
grain boundaries in single phase metals. Int. J. Plast. 25, 1655e1683. http://dx.doi.org/10.1016/j.ijplas.2008.09.002. Exploring New Horizons of Metal
Forming Research.
Britton, B.T., Wilkinson, A.J., 2012. Stress fields and geometrically necessary dislocation density distributions near the head of a blocked slip band. Acta
Mater. 60, 5773e5782. http://dx.doi.org/10.1016/j.actamat.2012.07.004.
Brown, A.F., 1952. Surface effects in plastic deformation of metals. Adv. Phys. 1, 427e479. http://dx.doi.org/10.1080/00018735200101241.
Cermelli, P., Gurtin, M.E., 2001. On the characterization of geometrically necessary dislocations in finite plasticity. J. Mech. Phys. Solids 49, 1539e1568.
http://dx.doi.org/10.1016/S0022-5096(00)00084-3.
Conti, S., Theil, F., 2003. Single-slip elastoplastic microstructures. Arch. Ration. Mech. Anal. 178, 125e148.
Diard, O., Leclercq, S., Rousselier, G., Cailletaud, G., 2005. Evaluation of finite element based analysis of 3D multicrystalline aggregates plasticity: application
to crystal plasticity model identification and the study of stress and strain fields near grain boundaries. Int. J. Plast. 21, 691e722. http://dx.doi.org/10.
1016/j.ijplas.2004.05.017.
Dieter, G.E., 1986. Mechanical Metallurgy. McGraw-Hill.
Di Gioacchino, F., Clegg, W.J., 2014. Mapping deformation in small-scale testing. Acta Mater. 78, 103e113. http://dx.doi.org/10.1016/j.actamat.2014.06.033.
Di Gioacchino, F., da Fonseca, J.Q., 2013. Plastic strain mapping with sub-micron resolution using digital image correlation. Exp. Mech. 53, 743e754.
Dimiduk, D.M., Woodward, C., LeSar, R., Uchic, M.D., 2006. Scale-free intermittent flow in crystal plasticity. Science 312, 1188e1190. http://dx.doi.org/10.
1126/science.1123889.
Dunne, F.P.E., 2014. Fatigue crack nucleation: mechanistic modelling across the length scales. Curr. Opin. Solid State Mater. Sci. 18, 170e179. http://dx.doi.
org/10.1016/j.cossms.2014.02.005. Slip Localization and Transfer in Deformation and Fatigue of Polycrystals.
El-Danaf, E., Kalidindi, S.R., Doherty, R.D., 1999. Influence of grain size and stacking-fault energy on deformation twinning in fcc metals. Metall. Mater. Trans.
A 30, 1223e1233. http://dx.doi.org/10.1007/s11661-999-0272-9.
Forest, S., 2008. Some links between Cosserat, strain gradient crystal plasticity and the statistical theory of dislocations. Philos. Mag. 88, 3549e3563. http://
dx.doi.org/10.1080/14786430802154815.
Gonzalez, D., Simonovski, I., Withers, P.J., Quinta da Fonseca, J., 2014. Modelling the effect of elastic and plastic anisotropies on stresses at grain boundaries.
Int. J. Plast. 61, 49e63. http://dx.doi.org/10.1016/j.ijplas.2014.03.012.
Han, Q., Kang, Y., Hodgson, P.D., Stanford, N., 2013. Quantitative measurement of strain partitioning and slip systems in a dual-phase steel. Scr. Mater. 69,
13e16. http://dx.doi.org/10.1016/j.scriptamat.2013.03.021.
Hansen, N., Mehl, R.F., Medalist, A., 2001. New discoveries in deformed metals. Metall. Mater. Trans. A 32, 2917e2935. http://dx.doi.org/10.1007/s11661-001-
0167-x.
He ripre
, E., Dexet, M., Cre pin, J., Ge le
bart, L., Roos, A., Bornert, M., Caldemaison, D., 2007. Coupling between experimental measurements and polycrystal
finite element calculations for micromechanical study of metallic materials. Int. J. Plast. 23, 1512e1539. http://dx.doi.org/10.1016/j.ijplas.2007.01.009.
Hoc, T., Cre pin, J., Ge
 le
bart, L., Zaoui, A., 2003. A procedure for identifying the plastic behavior of single crystals from the local response of polycrystals. Acta
Mater. 51, 5477e5488. http://dx.doi.org/10.1016/S1359-6454(03)00413-0.
Hoc, T., Rey, C., 2000. Effect of the free surface on strain localization in mild steel. Scr. Mater. 42, 1053e1058. http://dx.doi.org/10.1016/S1359-6462(00)00337-7.
Hughes, D.A., 1993. Microstructural evolution in a non-cell forming metal: AleMg. Acta Metall. Mater. 41, 1421e1430. http://dx.doi.org/10.1016/0956-
7151(93)90251-M.
Hunter, J.D., 2007. Matplotlib: a 2D graphics environment. Comput. Sci. Eng. 9, 90e95. http://dx.doi.org/10.1109/MCSE.2007.55.
Jillson, D.C., 1950. An experimental survey of deformation and annealing processes in zinc. Trans. AIME 188, 1009.
Jin, H., Lu, W.-Y., Korellis, J., 2008. Micro-scale deformation measurement using the digital image correlation technique and scanning electron microscope
imaging. J. Strain Anal. Eng. Des. 43, 719e728. http://dx.doi.org/10.1243/03093247JSA412.
Joo, S.-H., Lee, J.K., Koo, J.-M., Lee, S., Suh, D.-W., Kim, H.S., 2013. Method for measuring nanoscale local strain in a dual phase steel using digital image
correlation with nanodot patterns. Scr. Mater. 68, 245e248. http://dx.doi.org/10.1016/j.scriptamat.2012.10.025.
Kammers, A.D., Daly, S., 2013. Digital image correlation under scanning electron microscopy: methodology and validation. Exp. Mech. 53, 1743e1761. http://
dx.doi.org/10.1007/s11340-013-9782-x.
Kammers, A.D., Daly, S., 2011. Small-scale patterning methods for digital image correlation under scanning electron microscopy. Meas. Sci. Technol. 22,
125501. http://dx.doi.org/10.1088/0957-0233/22/12/125501.
Kocks, U.F., 2000. Texture and Anisotropy: Preferred Orientations in Polycrystals and Their Effect on Materials Properties. Cambridge University Press.
Kuhlmann-Wilsdorf, D., Hansen, N., 1991. Geometrically necessary, incidental and subgrain boundaries. Scr. Metall. Mater. 25, 1557e1562. http://dx.doi.org/
10.1016/0956-716X(91)90451-6.
Kuroda, M., 2014. On scale-dependent crystal plasticity models. In: Schro €der, J., Hackl, K. (Eds.), Plasticity and Beyond. CISM International Centre for
Mechanical Sciences. Springer, Vienna, pp. 305e353.
Leffers, T., 1979. A modified sachs approach to the plastic deformation of polycrystals as a realistic alternative to the taylor model. In: Gerold, P.H., Kostorz, G.
(Eds.), Strength of Metals and Alloys, pp. 769e774. Pergamon.
Leffers, T., Ray, R.K., 2009. The brass-type texture and its deviation from the copper-type texture. Prog. Mater. Sci. 54, 351e396. http://dx.doi.org/10.1016/j.
pmatsci.2008.09.002.
Lim, H., Carroll, J.D., Battaile, C.C., Buchheit, T.E., Boyce, B.L., Weinberger, C.R., 2014. Grain-scale experimental validation of crystal plasticity finite element
simulations of tantalum oligocrystals. Int. J. Plast. 60, 1e18. http://dx.doi.org/10.1016/j.ijplas.2014.05.004.
Lineau, C., Rey, C., de Lesegno, P.V., 1997. Experimental analysis of plastic deformation of steel grains comparison with polycrystal models predictions. Mater.
Sci. Eng. A 234e236, 853e856. http://dx.doi.org/10.1016/S0921-5093(97)00336-5.
Marteau, J., Haddadi, H., Bouvier, S., 2012. Investigation of strain heterogeneities between grains in ferritic and ferritic-martensitic steels. Exp. Mech. 53,
427e439. http://dx.doi.org/10.1007/s11340-012-9657-6.
McDowell, D.L., Dunne, F.P.E., 2010. Microstructure-sensitive computational modeling of fatigue crack formation. Int. J. Fatigue 32, 1521e1542. http://dx.doi.
org/10.1016/j.ijfatigue.2010.01.003. Emerging Frontiers in Fatigue.
McMurtrey, M.D., Was, G.S., Cui, B., Robertson, I., Smith, L., Farkas, D., 2014. Strain localization at dislocation channelegrain boundary intersections in
irradiated stainless steel. Int. J. Plast. 56, 219e231. http://dx.doi.org/10.1016/j.ijplas.2014.01.001.
Meyers, M.A., Ashworth, E., 1982. A model for the effect of grain size on the yield stress of metals. Philos. Mag. A 46, 737e759. http://dx.doi.org/10.1080/
01418618208236928.
Miraglia, M., Dawson, P., Leffers, T., 2007. On the influence of mechanical environment on the emergence of brass textures in FCC metals. Acta Mater. 55,
799e812. http://dx.doi.org/10.1016/j.actamat.2006.07.017.
Musienko, A., Tatschl, A., Schmidegg, K., Kolednik, O., Pippan, R., Cailletaud, G., 2007. Three-dimensional finite element simulation of a polycrystalline
copper specimen. Acta Mater. 55, 4121e4136. http://dx.doi.org/10.1016/j.actamat.2007.01.053.
Orowan, E., 1942. A new type of plastic deformation in metals. Nature 149, 643e644. http://dx.doi.org/10.1038/149643a0.
Orowan, E., 1941. Origin and Spacing of Slip Bands. Nature 147, 452e453. http://dx.doi.org/10.1038/147452a0.
Ortiz, M., Repetto, E. A, 1999. Nonconvex energy minimization and dislocation structures in ductile single crystals. J. Mech. Phys. Solids 47, 397e462. http://
dx.doi.org/10.1016/S0022-5096(97)00096-3.
Pantleon, W., 2008. Resolving the geometrically necessary dislocation content by conventional electron backscattering diffraction. Scr. Mate. 58, 994e997.
http://dx.doi.org/10.1016/j.scriptamat.2008.01.050.
Patriarca, L., Abuzaid, W., Sehitoglu, H., Maier, H.J., 2013. Slip transmission in bcc FeCr polycrystal. Mater. Sci. Eng. A 588, 308e317. http://dx.doi.org/10.1016/
j.msea.2013.08.050.
 F. Di Gioacchino, J. Quinta da Fonseca / International Journal of Plasticity 74 (2015) 92e109 109

Pokharel, R., Lind, J., Kanjarla, A.K., Lebensohn, R.A., Li, S.F., Kenesei, P., Suter, R.M., Rollett, A.D., 2014. Polycrystal plasticity: comparison between grain e
scale observations of deformation and simulations. Annu. Rev. Condens. Matter Phys. 5, 317e346. http://dx.doi.org/10.1146/annurev-conmatphys-
031113-133846.
Roters, F., Eisenlohr, P., Hantcherli, L., Tjahjanto, D.D., Bieler, T.R., Raabe, D., 2010. Overview of constitutive laws, kinematics, homogenization and multiscale
methods in crystal plasticity finite-element modeling: theory, experiments, applications. Acta Mater. 58, 1152e1211. http://dx.doi.org/10.1016/j.actamat.
2009.10.058.
Schramm, R.E., Reed, R.P., 1975. Stacking fault energies of seven commercial austenitic stainless steels. Metall. Trans. A 6, 1345e1351. http://dx.doi.org/10.
1007/BF02641927.
Schroeter, B.M., McDowell, D.L., 2003. Measurement of deformation fields in polycrystalline OFHC copper. Int. J. Plast. 19, 1355e1376. http://dx.doi.org/10.
1016/S0749-6419(02)00040-2.
Sutton, M.A., Li, N., Garcia, D., Cornille, N., Orteu, J.J., McNeill, S.R., Schreier, H.W., Li, X., Reynolds, A.P., 2007. Scanning electron microscopy for quantitative
small and large deformation measurements part I: SEM imaging at magnifications from 200 to 10,000. Exp. Mech. 47, 789e804. http://dx.doi.org/10.
1007/s11340-007-9042-z. Springer.
Tasan, C.C., Hoefnagels, J.P.M., Diehl, M., Yan, D., Roters, F., Raabe, D., 2014. Strain localization and damage in dual phase steels investigated by coupled in-
situ deformation experiments and crystal plasticity simulations. Int. J. Plast. 63, 198e210. http://dx.doi.org/10.1016/j.ijplas.2014.06.004. Deformation
Tensors in Material Modeling in Honor of Prof. Otto T. Bruhns.
Tatschl, A., Kolednik, O., 2003. On the experimental characterization of crystal plasticity in polycrystals. Mater. Sci. Eng. A 342, 152e168. http://dx.doi.org/10.
1016/S0921-5093(02)00278-2.
Tschopp, M.A., Bartha, B.B., Porter, W.J., Murray, P.T., Fairchild, S.B., 2009. Microstructure-dependent local strain behavior in polycrystals through in-situ
scanning electron microscope tensile experiments. Metall. Mater. Trans. A 40, 2363e2368. http://dx.doi.org/10.1007/s11661-009-9938-6.
Uchic, M.D., Dimiduk, D.M., Florando, J.N., Nix, W.D., 2004. Sample dimensions influence strength and crystal plasticity. Science 305, 986e989. http://dx.doi.
org/10.1126/science.1098993.
Van der Walt, S., Colbert, S.C., Varoquaux, G., 2011. The NumPy array: a structure for efficient numerical computation. Comput. Sci. Eng 13, 22e30. http://dx.
doi.org/10.1109/MCSE.2011.37.
Walley, J.L., Wheeler, R., Uchic, M.D., Mills, M.J., 2012. In-situ mechanical testing for characterizing strain localization during deformation at elevated
temperatures. Exp. Mech. 52, 405e416. http://dx.doi.org/10.1007/s11340-011-9499-7.
Wu, T.-Y., Bassani, J.L., Laird, C., 1991. Latent hardening in single crystals I. Theory and experiments. Proc. R. Soc. Lond. Math. Phys. Eng. Sci. 435, 1e19. http://
dx.doi.org/10.1098/rspa.1991.0127.
Zaiser, M., 2006. Scale invariance in plastic flow of crystalline solids. Adv. Phys. 55, 185e245. http://dx.doi.org/10.1080/00018730600583514.
Zhang, N., Tong, W., 2004. An experimental study on grain deformation and interactions in an Ale0.5%Mg multicrystal. Int. J. Plast. 20, 523e542. http://dx.
doi.org/10.1016/S0749-6419(03)00100-1. Owen Richmond Memorial Special Issue.
Zhao, Z., Ramesh, M., Raabe, D., Cuitin ~ o, A.M., Radovitzky, R., 2008. Investigation of three-dimensional aspects of grain-scale plastic surface deformation of
an aluminum oligocrystal. Int. J. Plast. 24, 2278e2297. http://dx.doi.org/10.1016/j.ijplas.2008.01.002.