Mechanics of Materials 160 (2021) 103934

Contents lists available at ScienceDirect

Mechanics of Materials
journal homepage: www.elsevier.com/locate/mecmat

Research paper

Nucleation of twinning dislocation loops in fcc metals

Sweta Kumari, Amlan Dutta ∗
Department of Metallurgical and Materials Engineering, Indian Institute of Technology Kharagpur, West Bengal 721302, India

ARTICLE INFO ABSTRACT

Keywords: Deformation twinning, which occurs in fcc metals only under particular conditions of intrinsic material
Deformation twinning properties, microstructure, and loading conditions, occupies an indispensable place in their deformation
Atomistic simulation mechanism maps. Nonetheless, dedicated studies have seldom been carried out to explore the fundamental
Dislocation loop
properties of twinning dislocations mediating this process. Here we employ a combination of atomistic
Nucleation
computations and continuum modeling to investigate the nucleation of twinning dislocation loops in metals
Nudged-elastic-band
with fcc crystal structure. Besides the conventional layer-by-layer model of twin nucleation, the newly proposed
alternate-shear mechanism has been investigated, and the energy barrier and shear stress for twinning loop
nucleation have been determined. We find the nucleation stress in the latter mechanism to be smaller than
that in the former. The study also highlights the non-uniform variations of the critical loop size and fractional
Burgers vector with the applied shear load.

1. Introduction very high SFE metal like platinum. Regardless of the exact mechanism
in action, the formation of a twin embryo involves plastic slips on three
Except for low stacking fault energy (SFE) systems, deformation successive {111} slip planes, and its growth occurs through subsequent
twinning is typically absent in pure metals and alloys with fcc crystal slip over the existing twin boundary. Slip of atomic planes, both dur-
structure. However, under the non-trivial conditions of loading, the ing the embryonic and growth stages of the twin-fault, is mediated
metals with moderate and high SFEs can also exhibit this mechanism. by the twinning partial dislocations. It underlines the significance of
Conditions favorable to deformation twinning include high strain rates twinning dislocations in the deformation mechanism maps of the fcc
and stresses, along with the reduced temperature (Gray, 2012; Zhao metals (Jo et al., 2014). A key contribution of the famous Orowan’s
et al., 2016; Kauffmann et al., 2011). In equal channel angular pressing relationship is establishing a trade-off between the density and velocity
of copper, twinning has been observed in the shear bands representing of line defects for a given strain rate. Consequently, a comprehensive
the regions of severe plastic deformation (Huang et al., 2006). In understanding and quantification of the deformation process remain
addition to the coarse-grained polycrystalline samples under special incomplete unless the mechanism of dislocation nucleation has been
loading conditions, nanocrystalline fcc metals like aluminum and cop- explored in detail. In the case of fcc metals, atomistic and line dynam-
per exhibit deformation twinning even at room temperature and low
ics simulations involving saddle point search have shown remarkable
strain rates (Zhu et al., 2004; Rottmann and Hemker, 2017; Sun et al.,
success in computing the activation energies for the nucleation of
2018). Similarly, nanosized fcc systems like nanoparticles (Lange et al.,
dislocation loops (Aubry et al., 2011; Jennings et al., 2013; Geslin et al.,
2016), nanowires (Kim et al., 2018), and nanopillars (Shu et al., 2017)
2017). These computational strategies have been duly complemented
also exhibit load-induced twinning.
with the development of analytical nucleation models. In this direction,
Given the elusive nature of deformation twinning, multiple mech-
the first breakthrough is probably attributable to Gutkin and Ovid’ko
anisms have been proposed to explain the formation and growth of
(2008), who put forward the idea of treating the Burgers vector as
twins in fcc metals. While the twin formation in bulk systems is
often described in terms of the pole-mechanism, fault-pair model, a variable non-crystallographic parameter. A significant refinement of
or layer-by-layer shear model, the shear model is the most feasible this approach was introduced by Aubry et al. (2011), who modified the
mechanism in their nanosized counterparts on account of the confined energetics equation of homogeneous loop nucleation by incorporating
space available (Chowdhury and Sehitoglu, 2018). Furthermore, Wang the generalized stacking fault energy (GSFE) as an atomistically ex-
et al. (2017) have recently demonstrated a variation of the conventional tracted parameter. Jennings et al. (2013) further extended this scheme
layer-by-layer mechanism, which can induce deformation twinning in a to study the surface nucleation of dislocations in the nanowires of

∗ Corresponding author.
E-mail address: [email protected] (A. Dutta).

https://doi.org/10.1016/j.mechmat.2021.103934
Received 13 December 2020; Received in revised form 8 May 2021; Accepted 30 May 2021
Available online 2 June 2021
0167-6636/© 2021 Elsevier Ltd. All rights reserved.
S. Kumari and A. Dutta Mechanics of Materials 160 (2021) 103934

Fig. 1. Schematic illustration of the (a) layer-by-layer (ABC), and (b) alternate-layer
(ACB) slip configurations. The green layers indicate atomic planes with local fcc
packing, whereas the orange ones have local hcp structure. Two adjacent hcp layers
constitute an intrinsic stacking fault, while a single layer represents a twin boundary.
Subsequent slips over the twin boundaries of the embryo would cause it to grow. (For Fig. 2. Representative structures of the initial and final states for NEB calculations in
interpretation of the references to color in this figure legend, the reader is referred to the (a) layer-by-layer (ABC), and (b) alternate-layer (ACB) twinning routes. Only the
the web version of this article.) atoms in the twin fault are shown, where the green particles indicate atoms with fcc
structure, while the red ones have local hcp environments. The slab of hcp atoms in
the ACB configuration is shown as translucent so that the twinning loop at the center
of the slab becomes visible. (For interpretation of the references to color in this figure
several fcc metals. More recently, Faisal and Weinberger (2019) have legend, the reader is referred to the web version of this article.)
employed these nucleation models to analyze the power-law expo-
nent that appears in a phenomenological expression of the activation
free-energy. the conventional layer-by-layer scheme of deformation twinning, which
Despite the improvements in simulation and analytical strategies, begins with the ABC kind of embryo formation, as depicted in the
the studies on fcc metals have primarily confined to the nucleation of figure. This scheme has been referred to as Path-A by Wang et al. (c.f.
full or Shockley partial dislocation loops (Aubry et al., 2011; Jennings Fig. 3(a) in Wang et al., 2017). Another possible route is displayed in
et al., 2013; Geslin et al., 2017; Faisal and Weinberger, 2019; Burbery Fig. 1(b). Here the slips occur on two alternate {111} planes, separated
et al., 2016). In contrast, very little information is available regarding by a single plane in between, thereby creating two adjacent intrinsic
the nucleation of twinning dislocations in these systems. It entails stacking faults. A third slip on the intermediary plane creates the twin
a glaring gap in comprehending the deformation mechanism map, embryo. We can refer to this scheme as the ACB shear, which has been
which is expected to incorporate the deformation twinning mechanism designated as Path-B by Wang et al. (c.f. Fig. 3(b) in Wang et al., 2017)
depending upon the intrinsic characteristics of a material and external and is very similar to the mechanism reported by Wang and Huang
loading conditions. In this context, we must note that the slip of atomic (2004). On the one hand, the formation of intrinsic stacking faults in
planes leading to the formation of a stacking fault is crystallographi- both the mechanisms is associated with the glide of Shockley partials
cally distinct from a slip leading to the formation or migration of a twin ̄
with 𝑎⟨112⟩∕6 as their Burgers vector. On the other hand, slips leading
boundary. It results in a qualitative difference between the GSFE and to the formation and growth of the meta-twin or twin embryo are
the generalized planar fault energy (GPFE) of the {111} slip plane. In mediated by the glide of twinning dislocations with the same Burgers
the present study, we explore the homogeneous nucleation of twinning vector as that of a Shockley partial.
dislocation loops in fcc metals by atomistic simulations. In particular, The first set of atomistic computations estimates the minimum-
the nudged elastic band (NEB) method has been employed to measure energy paths of the nucleation of twinning dislocation loops by employ-
the activation barrier against the nucleation of twinning loops for two ing the climbing-replica version (Uberuaga and Jonsson, 2000) of the
different slip configurations in Ag, Au, and Cu. The choice of these three NEB method. Here the first configuration involves the loop nucleation
metals covers a fairly broad range of the intrinsic stacking fault energy, over an existing twin boundary and corresponds to the formation and
while the two types of slips correspond to the conventional layer- growth of a meta-twin or twin embryo associated with the conventional
by-layer shear and the newly discovered alternate-shear mechanism. ABC shear mechanism (Fig. 1(a)). The initial structure used in the NEB
Following the estimation of nucleation paths at different stresses, we calculations consists of a crystal of dimensions 16𝑎⟨112⟩ ̄ × 24𝑎⟨110⟩
̄ ×
compute the generalized twin fault energies for the two configurations 20𝑎⟨111⟩, with the 𝑥 and 𝑧 directions along the Burgers vector and
considered here. The GPFE acts as a critical material parameter fed to the normal to the slip plane, respectively. The crystal also contains an
the continuum nucleation model, whose predictive capability has been embryonic twin slab with two twin boundaries. This structure is relaxed
assessed vis-à-vis the outcome of the NEB calculations. to the desired shear stress (𝜏𝑥𝑧 ) so that a shear strain is developed. The
other end of the replica chain is constructed by creating a twinning
2. Modeling and simulation shear loop of radius ∼10 Å and Burgers vector 𝑎⟨112⟩∕6 ̄ on the back of
the upper twin boundary. Fig. 2(a) shows the two ends of the chain of
Fig. 1 schematically demonstrates the two varieties of slip con- states. The second configuration deals with the nucleation between two
sidered in the present work. In the first (Fig. 1(a)), the formation adjacent intrinsic stacking faults and corresponds to the twin-forming
of the twin embryo commences with the formation of an intrinsic slip of the ACB mechanism shown in Fig. 1(b). Here the initial state of
stacking fault (SF) when one part of the crystal slips over a {111} the NEB is similar to that in the first configuration, with the exception
̄
plane by a relative displacement of 𝑎⟨112⟩∕6. A subsequent slip over of having two adjacent intrinsic stacking faults instead of the twin
this stacking fault creates a meta-twin (MT) with two boundaries, and embryo. The end state consists of a small dislocation loop placed on
a third subsequent slip over the meta-twin boundary creates the twin the {111} slip plane between these two stacking faults, as exhibited in
embryo (TE). Further slips on the consecutive slip planes would cause Fig. 2(b). Periodic boundary conditions have been imposed in all the
the twin boundary to migrate, thereby widening the twin slab. This is three directions.

2
S. Kumari and A. Dutta Mechanics of Materials 160 (2021) 103934

The replica-chain is initialized by creating fourteen intermediate

images between the end states through linear interpolation. In order
to prevent the loop from completely expanding and sweeping through
the slip plane, feedback forces are applied on the end state, which
confine it to an isopotential surface. In the NEB computation, the chain
of replicas is made to evolve through the recently improved FIRE algo-
rithm (Guénolé et al., 2020), which offers a significant enhancement in
performance over its older implementation. To extract the interatomic
forces, we select the embedded-atom-model potentials for Au (Grochola
et al., 2005), Ag (Zhou et al., 2004), and Cu (Mishin et al., 2001),
which produce the cohesive energies, SFEs, and stiffness constants in
reasonable agreement with those obtained from the DFT calculations.
The second set of atomistic computations involves quantifying the
generalized planar fault energy, which is an input parameter of the
nucleation model described further below. Here the set-up is similar to
the one created for the initial structure of the NEB calculations, but with
free surfaces normal to the ⟨111⟩ direction. In the first configuration
with an embryonic twin, the upper twin boundary is the slip plane,
whereas the plane between the two stacking faults is chosen as the slip
plane in the second configuration. The part of the crystal above the
Fig. 3. Generalized planar fault energies for the (a) stress-free and (b) stressed crystals
selected slip plane is displaced laterally relative to the lower part, and
in the ABC shear mechanism. GPFE profiles for the (c) stress-free and (d) stressed
the energy of the system is recorded as a function of disregistry. At each crystals in the ACB shear model. The indicated shear stress values are the maximum
step of the lateral displacement, the system is structurally relaxed, such stresses at which the nucleation barriers have been estimated for each of the metals.
that the atoms are allowed to move only normal to the slip plane. All
the atomistic computations reported here have been performed using
the LAMMPS molecular dynamics code (Plimpton, 1995), whereas present only a single GPFE profile for each metal in Fig. 3(a), which
the Atomsk (Hirel, 2015) and OVITO (Stukowski, 2009) codes have alone represents the planar fault energies for all the slips in the layer-
been employed for various pre-processing and post-processing tasks, ̄
by-layer model. In a perfect fcc crystal, the slip of 𝐛 = 𝑎⟨112⟩∕6 on the
respectively. {111} plane would produce a stable stacking fault energy, as seen in a
The atomistically informed nucleation model employed in this study typical generalized stacking fault energy profile. In contrast, the GPFE
is based upon the approach demonstrated by Aubry et al. (2011). It plots in Fig. 3(a) show translational invariance of structural energy at
expresses the enthalpy of formation of a dislocation loop of radius 𝑟 at the same disregistry. This is because unlike a perfect fcc crystal, where
a given stress 𝜏 as, ̄
a slip of 𝑎⟨112⟩∕6 creates an intrinsic stacking fault, the slips in the
( ) 𝐺𝑏2𝑓 [ 2 − 𝜈 ( 8𝑟 )
1
] ABC shear model merely cause the pre-existing twin boundary to shift
𝐸 𝑏𝑓 , 𝑟; 𝜏 = 2𝜋𝑟 ln −2 + by one atomic plane, while its structure and energy remain unaltered.
8𝜋 1 − 𝜈 𝑟0 2 (1)
[ ( ) ( )] In the presence of an applied shear load, the relative displacements
+𝜋 𝛾𝑔𝑝𝑓 𝑏𝑓 + 𝑢0 ; 𝜏 − 𝛾𝑔𝑝𝑓 𝑢0 ; 𝜏 𝑟2 − 𝜋𝑏𝑓 𝜏𝑟2 , corresponding to the translational invariance in the GPFE profiles get
where 𝑏𝑓 is a variable fractional Burgers vector and 𝛾(𝛿) is the GPFE reduced below the original value of 𝑎⟨112⟩∕6,̄ as evident from compar-
obtained as a function of the disregistry, 𝛿. In the above equation, ing Fig. 3(b) to (a). Moreover, the applied shear load also reduces the
𝑢0 is the offset-displacement calculated by equilibrating the applied unstable twin fault energies, given by the intermediate energy maxima
stress against the gradient of the GPFE. Apart from the GPFE, the seen in the figures. Both of these stress-effects are attributable to the
other intrinsic material parameters are the effective shear modulus (𝐺) alteration in atomistic configuration across the slip-plane due to the
and Poisson’s ratio (𝜈), which in general, depend upon the relevant elastic shear strain.
crystallographic orientation (Fitzgerald and Aubry, 2010). 𝑟0 is the As compared to the ABC shear, a contrasting feature is visible
effective cut-off radius, which is the only fitting parameter in this in the GPFE plots corresponding to slip in the alternate-layer (ACB)
model. Although a somewhat simpler expression for the loop’s enthalpy mechanism. Here we find that the disregistry of 𝑎⟨112⟩∕6 ̄ leads to
has been employed by Jennings et al. (2013), Eq. (1) is found to yield a structure with energy, which is lower than the initial energy, and
slightly better results for the systems studied here. effectively represents a negative stacking fault energy (Fig. 3(c)). This
A noteworthy feature of this nucleation model is recognition of the peculiar feature can be understood by realizing that the initial structure
fact that at large enough stress, the GPFE itself becomes sensitive to with two stacking faults on alternate planes has substantially elevated
the applied stress, thereby justifying the incorporation of the stress, 𝜏, structural energy, for it is effectively four hcp layers sandwiched be-
as an extrinsic parameter in Eq. (1). Once the stress-dependent GPFE ̄
tween fcc phases. A relative slip of 𝑎⟨112⟩∕6 between the two adjacent
is obtained, the loop-energy is computed over a two-dimensional grid stacking faults, as illustrated in the third step of Fig. 1(b), causes the
consisting of 𝑏𝑓 and 𝑟, and the activation barrier against nucleation of middle two layers to return back to their fcc configurations. It explains
the twinning loop is computed at the saddle point of this energy surface. the drop in the energy of the system with respect to the initial structure.
In the thermodynamic viewpoint perceiving an intrinsic stacking fault
3. Result and discussions as a thin hcp phase embedded within the fcc system (Dash and Dutta,
2020), the slip studied in Fig. 3(c) can be perceived as a process that
Fig. 3(a) displays the generalized planar fault energies correspond- is opposite to that of the formation of a stacking fault in fcc. Similar to
ing to the ABC shear model for a stress-free crystal, where the slip what is observed in Fig. 3(b), an applied shear stress affects the GPFE
occurs over a pre-existing twin fault. As recently shown by Pulagam profile in the ACB shear model as well (Fig. 3(d)).
and Dutta (2020), both the slips illustrated in Fig. 1(a) for the layer- Having explored the generalized planar fault energies corresponding
by-layer model and the subsequent slips entailing further thickening to the two shear models, we now examine the transition pathways
of the twin slab, exhibit identical GPFE profiles. This observation is revealed by the NEB computations. Fig. 4(a) and (b) display the rep-
attributed to the equivalent stacking sequence of atomic planes around resentative plots showing the variations in structural energies along
the fault boundary in each of these slip configurations. Therefore, we the minimum energy paths of twinning loop nucleation in the ABC

3
S. Kumari and A. Dutta Mechanics of Materials 160 (2021) 103934

Fig. 5. Radii (circular symbols) of the critical twinning loops and associated fractional
Burgers vectors (square symbols) plotted against the applied shear loads in the (a)
layer-by-layer and (b) alternate-shear mechanisms. To facilitate comparison among the
three metals, the fractional Burgers vector (𝑏𝑓 ) is given as fraction of the corresponding
√
crystallographic Burgers vector, 𝑏 = 𝑎∕ 6.
Fig. 4. Structural energies plotted against the reaction coordinates for twinning loop
nucleation in Cu in the (a) layer-by-layer and (b) alternate-shear mechanisms. Energy
barriers displayed as the functions of applied shear stress in (c) ABC and (d) ACB shear
models. Circular symbols present the results of NEB calculations, whereas the dashed
lines indicate the continuum model’s fitted outcomes. loop in the ACB model have local fcc structure in contrast to the hcp
atoms outside it. Fig. 4(d) displays a reduction in the nucleation barrier
with an increase in the applied shear stress in the alternate-shear
mechanism. The fitted cut-off radii used to plot the energy barriers with
and ACB shear models, respectively. Here the critical loop nucleus the continuum model for Ag, Au, and Cu are obtained as 0.99𝑏, 0.65𝑏,
is easily identified as the replica with the largest potential energy, and 0.66𝑏, respectively. A direct comparison with Fig. 4(b) shows that
and the corresponding snapshots of the critical twinning loops are the nucleation stresses of twinning dislocation loops in all the three
also displayed as the translucent backgrounds. The energy barriers are metals are indeed smaller than those in the ABC model, as expected
computed at different stresses for the layer-by-layer shear mechanism from the GPFE profiles.
and are exhibited in Fig. 4(c). Here, the fitting parameter 𝑟0 , which We have already utilized Eq. (1) to estimate the energy barrier as
produces the energy barriers in remarkable agreement with the NEB the energy of the saddle point along the transition pathway. However,
results, has the values 0.85𝑏, 0.66𝑏, and 0.67𝑏 for Ag, Au, and Cu, the same saddle point also provides the values of the fractional Burgers
respectively. We find that the stress required for a given energy barrier vector (𝑏𝑓 ) and loop radius (𝑟) as the parameters of the critical twinning
is largest for Cu and smallest for Au. This trend can be intuitively loop. Fig. 5(a) displays these critical parameters calculated with the
understood by realizing that the unstable twin fault energy dominates fitted values of 𝑟0 for the ABC shear model, whereas Fig. 5(b) presents
the mechanism of loop nucleation modeled in Eq. (1), which follows the same for the ACB mechanism. It is pertinent to recall that in the
the order Cu > Ag > Au (Fig. 3(a) and (b)). conventional model of loop nucleation, the Burgers vector is assumed
In the case of a shear loop of Shockley partial, a positive stacking to be fixed and the same as the crystallographic Burgers vector, and
fault energy tends to oppose its expansion. Therefore, it is interesting predicts a uniform reduction in the critical radius with an increase
to note that the negative fault energy observed in Fig. 3(c) and (d) is in applied shear stress. However, the results in Fig. 5 underline the
expected to support the expansion of the loop in the ACB nucleation implications of replacing the crystallographic Burgers vector (𝑏) by its
model. Further, the unstable fault energies are smaller than their cor- fractional counterpart (𝑏𝑓 ). We note that in both the shear models, the
responding counterparts in the layer-by-layer model. This observation variation of loop radius with applied load is not essentially uniform
leads us to predict reduced stresses for loop nucleation. The critical loop and can even be non-monotonic. For instance, in the ACB shear, the
nucleus in the alternate-layer mechanism, as seen in Fig. 4(b), reveals critical loop radius in Ag at 1 GPa is slightly larger than that at 0.9 GPa
that unlike the loop in the ABC model, where the atoms inside the loop (Fig. 5(b)). Such a trend can be understood by observing that the frac-
have the same twin-boundary structure (Fig. 4(a)), the atoms inside the tional Burgers vector’s variation with increasing applied stress is highly

4
S. Kumari and A. Dutta Mechanics of Materials 160 (2021) 103934

non-uniform and unpredictable. In the example mentioned above, we CRediT authorship contribution statement
find that a substantial drop in 𝑏𝑓 from the stress of 1 GPa to 0.9 GPa
causes the loop radius to increase instead of decrease, as opposed to our Sweta Kumari: Conceptualization, Methodology, Software, Investi-
intuitive expectation. Such results firmly imply that the conventional, gation, Visualization, Writing - original draft. Amlan Dutta: Conceptu-
text-book model of loop nucleation is not only quantitatively inaccurate alization, Writing - review & editing, Supervision.
but can be inadequate in predicting the qualitative trend as well.
At this juncture, it can be pointed out that the nucleation stresses Declaration of competing interest
shown in Figs. 4 and 5 are significantly large, for they represent the
mechanism of homogeneous nucleation. In principle, a similar study is The authors declare that they have no known competing finan-
possible for the mechanisms of heterogeneous nucleation of twinning cial interests or personal relationships that could have appeared to
loops at surfaces, grain boundaries, and other planar interfaces, where influence the work reported in this paper.
the stresses are expected to be much smaller. The mathematical models
for such processes would invariably involve treating the image stresses Acknowledgments
and ledge formation on the surface or interface. Jennings et al. (2013)
have tried to model the heterogeneous nucleation of Shockley loops We acknowledge the Department of Science and Technology, Govt.
at the surfaces of nanowires under tension and compression. However, of India, for partially funding the present work under the ECRA scheme.
their model did not exhibit the remarkable agreement with the atom- We also thank the Indian Institute of Technology Kharagpur, India, for
istic results, which is observed in the case of homogeneous nucleation. financial support through the ISIRD grant.
It suggests that there are factors, which need to be incorporated in the
models of heterogeneous nucleation to predict the nucleation stresses References
with better accuracy. The development of such refined models in the
future can also pave the way for reliable quantitative prediction of the Aubry, S., Kang, K., Ryu, S., Cai, W., 2011. Scr. Mater. 64, 1043.
Burbery, N., Das, R., Ferguson, W.G., Po, G., Ghoniem, N., 2016. Int. J. Comput.
heterogeneous nucleation of twinning loop.
Methods 13, 1641006.
Chowdhury, P., Sehitoglu, H., 2018. J. Eng. Mater. Technol. 140, 020801.
4. Conclusion Dash, S., Dutta, A., 2020. Mater. Manuf. Process. 35, 635.
Faisal, A.H.M., Weinberger, C.R., 2019. Phys. Rev. Mater. 3, 103601.
By way of summary, this study examines the nucleation of twinning Fitzgerald, S.P., Aubry, S., 2010. J. Phys. Condens. Matter. 22, 295403.
Geslin, P.-A., Gatti, R., Devincre, B., Rodney, D., 2017. J. Mech. Phys. Solids 108, 49.
dislocation loops in fcc metals through the technique of transition- Gray, G.T. (Rusty), 2012. Annu. Rev. Mater. Res. 42, 285.
pathway-search and atomistically informed continuum model. In par- Grochola, G., Russo, S.P., Snook, I.K., 2005. J. Chem. Phys. 123, 204719.
ticular, we consider the newly proposed alternate-shear model of twin Guénolé, J., Nöhring, W.G., Vaid, A., Houllé, F., Xie, Z., Prakash, A., Bitzek, E., 2020.
formation, in addition to the conventional layer-by-layer model. A Comput. Mater. Sci. 175, 109584.
Gutkin, M., Ovid’ko, I., 2008. Acta Mater. 56, 1642.
particularly interesting feature of the twinning dislocations is that while
Hirel, P., 2015. Comput. Phys. Comm. 197, 212.
their Burgers vectors resemble those of Shockley partials, the gener- Huang, C., Wang, K., Wu, S., Zhang, Z., Li, G., Li, S., 2006. Acta Mater. 54, 655.
alized fault energy in the layer-by-layer nucleation does not exhibit Jennings, A.T., Weinberger, C.R., Lee, S.-W., Aitken, Z.H., Meza, L., Greer, J.R., 2013.
the stable stacking fault energy, which is otherwise a quintessential Acta Mater. 61, 2244.
characteristic of the GSFE. Besides, the alternate-shear model reveals a Jo, M., Koo, Y.M., Lee, B.-J., Johansson, B., Vitos, L., Kwon, S.K., 2014. Proc. Natl.
Acad. Sci. 111, 6560.
negative stacking fault energy, which motivates a comparison between Kauffmann, A., Freudenberger, J., Geissler, D., Yin, S., Schillinger, W., Sarma, V.S., Bah-
the twinning loop nucleation in these two mechanisms. Our computa- manpour, H., Scattergood, R., Khoshkhoo, M., Wendrock, H., Koch, C., Eckert, J.,
tions indicate that the nucleation stress of twinning loops is larger for Schultz, L., 2011. Acta Mater. 59, 7816.
metal with higher unstable fault energy. Furthermore, the nucleation Kim, S.-H., Kim, H.-K., Seo, J.-H., Whang, D.-M., Ahn, J.-P., Lee, J.-C., 2018. Acta
Mater. 160, 14.
stress for a given metal is smaller in the alternate-shear mechanism, as
Lange, A., Samanta, A., Majidi, H., Mahajan, S., Ging, J., Olson, T., van Benthem, K.,
compared to its layer-by-layer counterpart. When tuned with a single Elhadj, S., 2016. Acta Mater. 120, 364.
free parameter, the continuum model produces excellent fits to the Mishin, Y., Mehl, M.J., Papaconstantopoulos, D.A., Voter, A.F., Kress, J.D., 2001. Phys.
NEB results. A closer look at the critical loop reveals that its fractional Rev. B 63, 224106.
Burgers vector plays a vital role in dictating the critical loop size and Plimpton, S., 1995. J. Comput. Phys. 117, 1.
Pulagam, S., Dutta, A., 2020. arXiv:2008.11937.
stress of nucleation.
Rottmann, P.F., Hemker, K.J., 2017. Acta Mater. 140, 46.
The computational results highlight the significance of perceiving Shu, X., Kong, D., Lu, Y., Long, H., Sun, S., Sha, X., Zhou, H., Chen, Y., Mao, S., Liu, Y.,
the twinning deformation in terms of nucleation and growth of twin- 2017. Sci. Rep. 7, 1.
ning partial loops in the fcc materials. An obvious extension of this Stukowski, A., 2009. Model. Simul. Mat. Sci. Eng. 18, 015012.
Sun, L., He, X., Lu, J., 2018. Npj Comput. Mater. 4, 1.
work is to study the process of heterogeneous nucleation of twinning
Uberuaga, B., Jonsson, H., 2000. J. Chem. Phys. 113, 9901.
loops. Moreover, as the computations provide us with the energy bar- Wang, L., Guan, P., Teng, J., Liu, P., Chen, D., Xie, W., Kong, D., Zhang, S., Zhu, T.,
rier to loop nucleation, it can be subsequently utilized to estimate the Zhang, Z., Ma, E., Chen, M., Han, X., 2017. Nature Commun. 8, 2142.
kinetic parameters associated with the process at different strain rates Wang, J., Huang, H., 2004. Appl. Phys. Lett. 85, 5983.
and temperatures. The results and insights rendered here can motivate Zhao, F., Wang, L., Fan, D., Bie, B.X., Zhou, X.M., Suo, T., Li, Y.L., Chen, M.W., Liu, C.L.,
Qi, M.L., Zhu, M.H., Luo, S.N., 2016. Phys. Rev. Lett. 116, 075501.
further investigations into the fundamental computational modeling of
Zhou, X.W., Johnson, R.A., Wadley, H.N.G., 2004. Phys. Rev. B 69, 144113.
twinning dislocations. Zhu, Y., Liao, X., Srinivasan, S., Zhao, Y., Baskes, M., Zhou, F., Lavernia, E., 2004.
Appl. Phys. Lett. 85, 5049.