Materials Science & Engineering A 820 (2021) 141504

Contents lists available at ScienceDirect

Materials Science & Engineering A

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

Grain size effect on tensile deformation behaviors of pure aluminum

B.B. Wang a, c, G.M. Xie b, L.H. Wu c, P. Xue c, *, D.R. Ni c, B.L. Xiao c, Y.D. Liu a, Z.Y. Ma c
a
School of Materials Science and Engineering, Northeastern University, 3-11 Wenhua Road, Shenyang, 110819, China
b
State Key Laboratory of Rolling and Automation, Northeastern University, 3-11 Wenhua Road, Shenyang, 110819, China
c
Shi-changxu Innovation Center for Advanced Materials, Institute of Metal Research, Chinese Academy of Sciences, 72 Wenhua Road, Shenyang, 110016, China

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

Keywords: Grain refinement is a very effective method to improve the mechanical properties of materials and attracts
Grain size effect widespread interests among researchers. However, the grain size effect on the mechanical properties is still
Tensile properties unclear due to the undesirable microstructure in ultrafine grained (UFG) materials. In the present work, series of
Hall-petch relationship
ideal materials with average grain sizes range from 0.7 μm to 30.0 μm containing high fraction of high angle
Strengthening mechanism
grain boundaries (HAGBs), equiaxed grains and low density of dislocations were produced by friction stir pro­
cessing (FSP). It was found that the Hall-Petch relationships could be classified into three stages as the grain size
reduced from coarse grain to UFG regimes, which were decided by the strengthening mechanisms during tensile
deformation. The strengthening effect of HAGBs (71 MPa⋅μm1/2) was almost three times of low angle grain
boundaries (25 MPa⋅μm1/2), resulting in the positive deviation of Hall-Petch slope in fine grain regime by the
increased specific surface area of HAGBs. The further positive deviation of the Hall-Petch slope in UFG regime
was affected by the occurrence of an extra dislocation source limited strengthening mechanism, which was up to
29 MPa and reached to about 20% of the yield strength. The increased recovery rate of dislocations at HAGBs
contributed to the decrease of mobile dislocation density, leading to the losing of work hardening in UFG regime
during tensile deformation.

1. Introduction such as equal channel angle pressing (ECAP), high-pressure torsion

(HPT), accumulative roll bonding (ARB), have become the main
As well-known, grain refinement is a very effective method to methods to prepare bulk UFG materials [13,14,16–18]. However, the
improve the mechanical properties. Therefore, nanostructured and materials usually exhibit elongated grains, high density of dislocations
ultrafine-grained (UFG) materials have attracted widespread interests and high fraction of low-angle boundaries (LAGBs) [19–30]. Besides the
among researchers due to the presence of large amount of grain grain size, grain aspect ratio (GAR), dislocation density, and grain
boundaries in the materials contributing to the excellent mechanical boundary characteristics will significantly affect the tensile properties
properties [1–6]. The contribution of grain refinement to the improved [31–33]. In order to investigate the grain size effect on the tensile
yield strength (YS) can be elucidated by the well-known Hall-Petch deformation behaviors, a series of model materials with uniform equi­
relationship in conventional coarse grain (CG) and fine grain (FG) re­ axed grains, low dislocation density, high fraction of high angle grain
gimes [7,8]. However, as the grain size decreased to UFG regime, boundaries (HAGBs, misorientation angle ≥ 15◦ ) and weak texture are
traditional work hardening mechanism would not work, as the compa­ required. Although researchers kept trying to avoid or minimize the
rable size between the ultrafine grains with the elementary unit of the microstructural variation in different grain size regimes, the results were
dislocation structures formed during tensile deformation process. not so satisfactory [19]. Therefore, preparing nearly ideal model mate­
Therefore, the relationship between grain size and tensile strength of rials with different grain sizes is still a difficult task.
UFG material should be different from that of CG and FG materials. Friction stir processing (FSP) is a new thermo-mechanical processing
Although, the deviation phenomenon of Hall-Petch slope in UFG range technology, which is invented, based on friction stir welding [34]. It has
has been reported in various studies, no consensus conclusion has been been proved that FSP is able to prepare CG, FG and UFG materials with
reached [9–12]. different grain sizes by adjusting the processing parameters [35–37].
Nowadays, various severe plastic deformation (SPD) technologies, More importantly, the grain refinement mechanism during FSP is

* Corresponding author.
E-mail address: [email protected] (P. Xue).

https://doi.org/10.1016/j.msea.2021.141504
Received 1 January 2021; Received in revised form 27 March 2021; Accepted 5 May 2021
Available online 4 June 2021
0921-5093/© 2021 Elsevier B.V. All rights reserved.
B.B. Wang et al. Materials Science & Engineering A 820 (2021) 141504

dynamic recrystallization (DRX), which can avoid the high density of Huang et al. [39] studied the strengthening mechanisms in nano­
dislocations and strong shear textures, and uniform equiaxed grains with structured and UFG materials, and indicated that the grain boundary
high fraction of HAGBs were usually obtained in the processed zone (PZ) misorientation angle significantly affected the yield stress. Apparently,
[38]. Therefore, compared to the SPD methods, FSP provides an effec­ FSP samples show similar distribution of misorientation angle as
tive solution to prepare materials with different grain sizes and to decreasing the grain size, which can exclude the influence of grain
investigate the grain size effect on the tensile properties. boundary characteristic on the mechanical properties.
In the present work, pure Al samples with various grain sizes were Typical distribution of GARs for UFG-0.7 sample was shown in
prepared by FSP under different parameters. We will focus on the Fig. 4a. Only 5% of the GARs were larger than 2, and 68% of the GARs
transition of tensile deformation behavior and Hall-Petch relationship as were less than 1.5. The average GAR value was equal to 1.5 for UFG-0.7
the grain size reduced from the CG to UFG regimes, and the strength­ sample, and the distribution of the GAR was consistent with the normal
ening mechanisms will be discussed in detail. distribution. The relationship between GAR distributions and grain sizes
is shown in Fig. 4b. As decreasing the grain size, all samples exhibited
2. Experiment methods similar distribution characteristics. It was clear that 60%–70% of the
GARs were less than 1.5, and only 8% of the GARs were larger than 2 for
Commercially pure Al (1060) was used as the base material (BM). all FSP samples. Tsuji et al. [15] reported that the average grain size of
The plates were processed by FSP along the rolling direction, using a tool UFG Al fabricated by ARB was 270 nm, but the grains were elongated
with a concave shoulder 12 mm in diameter and a conical threaded pin with the length as large as 1.2 μm at the long axis direction. The average
4 mm in root diameter and the processing parameters are shown in GAR value was as high as 4.4 for the UFG Al prepared by this monotonic
Table 1. By adjusting the FSP parameters, samples with various average deformation method. Meanwhile, GAR significantly affected the accu­
grain sizes (d) of 30 μm, 4.6 μm, 3.2 μm, 1.4 μm, 1.2 μm, 0.9 μm, 0.8 μm, racy of grain size measurement and the mechanical anisotropy, and
and 0.7 μm were successfully obtained, which were measured from the further influenced the relationship between grain size and tensile
electron back-scattered diffraction (EBSD) data including more than 300 properties [40–42]. The average value of GAR was equal to about 1.5 for
grains and defined as CG-30, FG-4.6, FG-3.2, FG-1.4, FG-1.2, UFG-0.9, all FSP samples with different grain sizes, implying that nearly equiaxed
UFG-0.8, and UFG-0.7 samples, respectively. grains were achieved for all processing conditions.
The specimens for microstructural examinations were machined The TEM images in Fig. 5 exhibited the typical microstructural fea­
perpendicular to the FSP direction, as shown in Fig. 1. Microstructural tures of the specimens with various grain sizes. It is clear that equiaxed
characterization and analysis were carried out using optical microscopy grains were observed in all samples with different grain sizes. Most grain
(OM), scanning electron microscopy (SEM) and transmission electron boundaries were sharp, clear, and relatively straight, which were con­
microscopy (TEM). The specimens for OM observation were ground, sisted with the characteristics of the HAGBs [34,43], as illustrated by
polished, and then etched in a solution of 10 g NaOH and 100 ml H2O for arrows in Fig. 5. The dislocation density in grain interior was very low
about 2 min. TEM foils and EBSD samples were prepared by double-jet for all samples, owning to the occurrence of dynamic recovery and DRX
electrolytic polishing using a solution of 30 ml HNO3 and 70 ml CH3OH during FSP [34,44].
at 248 K under a potential of 12 V. Tensile specimens, with a gauge Based on the analysis of EBSD and TEM microstructures, pure Al
length of 10 mm, width of 3 mm, and thickness of 2 mm, were machined samples with grain sizes vary from 30 μm to 0.7 μm, containing high
in the center of the PZ parallel to the FSP direction (Fig. 1). Uniaxial fraction of HAGBs, similar GAR value and low dislocation density, were
tensile tests were carried out at room temperature with an initial strain prepared by FSP under different parameters. Compared to SPD methods,
rate of 1 × 10− 3 s− 1. FSP provides an effective solution to prepare nearly ideal model mate­
rials to investigate the real relationship between the grain size and the
3. Results tensile properties from CG to UFG regimes.

3.1. Microstructure 3.2. Tensile properties

It is observed from Fig. 1 that a basin-shaped PZ without defect and Fig. 6 shows the tensile stress-strain curves of FSP samples with
onion ring structure was obtained in the cross section perpendicular to various grain sizes, and the mechanical properties are listed in Table 1.
the processing direction. The microstructural features were character­ As decreasing the grain size, the YS and ultimate tensile strength (UTS)
ized by EBSD, and the images of some typical samples were shown in gradually increased, and accompanied with the decrease of the uniform
Fig. 2. The black lines and the white lines represent the HAGBs and elongation and total elongation. CG-30 sample showed a typical
LAGBs, respectively. All of the FSP samples exhibited uniform equiaxed continuous work hardening behavior with low YS of 43.6 MPa and high
grains with high fraction of HAGBs, indicating the occurrence of DRX in elongation of 59.5%. Compared with CG-30 sample, the curve of FG-1.2
the PZs during FSP processing. sample exhibited a distinct yielding peak, and followed by a short stress
As exhibited in Fig. 3, the distributions of grain boundary misori­ platform region with greatly increased YS of 92.2 MPa and decreased
entation angles in all FSP samples, irrespective of grain sizes, were elongation of 30.1%. As the grain size decreased to UFG regime, the
similar to that of the random distribution for a cubic polycrystalline, curve of UFG-0.7 sample showed a distinct yielding peak followed by
while all FSP samples contained a large fraction of HAGBs (>70%). rapid strain softening with the highest YS of 142.5 MPa and the lowest

Table 1
Tensile properties and grain sizes of FSP samples under different processing parameters.
Sample Rotation rate (rpm) Processing speed (mm/min) Grain size (μm) w value (μm) YS (MPa) UTS (MPa) σ(MPa) Elongation (%)

CG-30 1500 100 30 1.21 43.6 ± 2.3 89.3 ± 2.4 124.0 ± 2.4 59.5 ± 1.1
FG-4.6 800 100 4.6 1.30 59.8 ± 2.4 90.1 ± 2.0 115.2 ± 2.0 41.3 ± 0.7
FG-3.2 800 400 3.2 1.35 67.4 ± 1.4 94.3 ± 1.8 110.7 ± 1.8 25.6 ± 0.4
FG-1.4 800 800 1.4 1.23 85.0 ± 1.9 103.7 ± 1.2 121.7 ± 1.2 38.6 ± 1.3
FG-1.2 600 800 1.2 1.25 92.2 ± 2.7 109.2 ± 2.6 119.5 ± 2.6 30.1 ± 0.5
UFG-0.9 200 50 0.9 1.14 115.6 ± 2.1 125.2 ± 2.9 131.7 ± 2.9 28.5 ± 0.6
UFG-0.8 400 800 0.8 0.96 130.0 ± 3.4 155.2 ± 3.7 155.8 ± 3.7 16.1 ± 0.4
UFG-0.7 100 20 0.7 0.86 142.5 ± 2.3 172.8 ± 3.3 173.4 ± 3.3 11.3 ± 0.2

2
B.B. Wang et al. Materials Science & Engineering A 820 (2021) 141504

Fig. 1. Schematic illustration of FSP and the sample locations for tensile tests and microstructure observations.

Fig. 2. EBSD images of samples with different grain sizes: (a) FG-4.6, (b) FG-3.2, (c) FG-1.2, (d) UFG-0.7.

elongation of 11.3%. change of grain size:

It is well known that the YS (σ y ) is related to the average grain size (d) Type I (d > 1.2 μm): the curve shows continuous work hardening
by the classical Hall-Petch relationship: behavior similar to that of the CG material.
Type II (0.9≤ d≤1.2 μm): the curve exhibits a distinct yielding peak,
σ y = σ 0 + k⋅d− 1/2
(1) and follows by a short stress platform region (dynamic equilibrium be­
tween strengthening and softening processes) before a continuous strain
where σ0 and k represent the friction stress and the slope of Hall-Petch
softening.
relationship, respectively. Fig. 7 shows the Hall-Petch relationship of
Type III (d<0.9 μm): the curve shows a distinct yielding peak
pure Al specimens prepared by FSP in this study. The values of σ0 and k
following by a rapid strain softening.
were 28 MPa and 67 MPa⋅μm1/2 by fitting the data to equation (2) at CG
Previous studies indicated that rapid strain softening and attendant
and FG regimes, respectively. However, the fitting result of k value at
loss of ductility were the main tensile deformation characteristics as
UFG regime was 396 MPa⋅μm1/2, which was much higher than that of FG
grain size decreased to UFG [45–47]. Strain hardening and softening
material.
phenomena were usually associated with the difference value between
the steady state subgrain size and the actual grain size. A semi-empirical
4. Discussion
relationship between the shear stress, and the steady state subgrain size
(w) has been well established, which can be expressed by the following
4.1. Tensile deformation behavior
equation [36,48–50]:
According to the curve characteristics in Fig. 6, the tensile stress- G
w = kw b (2)
strain curves can be classified into three different categories with the σ

3
B.B. Wang et al. Materials Science & Engineering A 820 (2021) 141504

Fig. 3. Distributions of grain boundary misorientation angles of FSP samples: (a) FG-4.6, (b) FG-3.2, (c) FG-1.2, (d) UFG-0.7.

Fig. 4. (a) Typical distribution of GAR for UFG-0.7 sample, (b) the fraction of GAR of samples with various grain sizes.

where G is shear modulus (26 GPa), b is burgers vector (0.286 nm) and σ w value, the stress field of grain boundary would hinder the formation of
is the maximum value of true stress listed in Table 1 kw is a constant dislocation cell in grain interior.
value between 10 and 30, and the value of 20 was used in previous study The typical TEM microstructures of FG-4.6, FG-1.2, and UFG-0.7
[50]. W value represents the minimum size of the dislocation sub­ samples after 2% tensile strain and final fracture are shown in Fig. 8.
structure evolved during the tensile deformation process, and the For FG-4.6 samples, some dislocation tangles could be observed in grain
calculated results are listed in Table 1. When the grain size was less than interior even with a small deformation of 2% tensile strain, as shown in

4
B.B. Wang et al. Materials Science & Engineering A 820 (2021) 141504

Fig. 5. Typical TEM microstructures of FSP pure Al samples: (a) FG-4.6, (b) FG-3.2, (c) FG-1.2, (d) UFG-0.7.

Fig. 6. Tensile engineering stress-strain curves of FSP samples with different

grain sizes.
Fig. 7. Hall-Petch relationships of FSP pure Al samples.
Fig. 8a. Typical dislocation cell structures and a large amount of dislo­
cation tangles formed in the grain interior after final fracture, as shown tensile strain. Different from that of FG-4.6 sample, some dislocations
in Fig. 8d. The calculated w value was 1.30 μm for FG-4.6 sample, which began to accumulate near the grain boundaries, as shown by the arrows
was smaller than the grain size (4.6 μm). During the tensile test, the in Fig. 8b. Further, the final microstructure after fracture in FG-1.2
deformation promoted the nucleation and multiplication of dislocations sample was obviously different from that of FG-4.6 sample. The distri­
in the grain interior for the CG and FG materials with Type I tensile bution of dislocations was not uniform, resulting in tangled dislocations
stress-strain curves. With continued tensile deformation, dislocations in some large grains and free of dislocation in some small grains. The
multiplication and rearrangement resulted in the formation of disloca­ calculated w value of FG-1.2 sample (1.25 μm) was almost equal to the
tion cells in the grain interior, impeded the movement of dislocations, average grain size (1.2 μm). In this case, the balance of the work hard­
and enhanced the interactions between the dislocations. In this case, ening in the large grains (d > w) and strain softening in the small grains
work hardening phenomenon occurred during the tensile deformation (d < w) contributed to a short steady state deformation for the Type II
process for this type of materials. materials in tensile deformation.
Fig. 8b and e shows the TEM microstructures after 2% tensile strain Typical TEM microstructures of UFG-0.7 sample after 2% tensile
and final fracture of FG-1.2 sample, respectively. Only a few dislocations strain and final fracture are exhibited in Fig. 8c and f, respectively. As
could be observed in the grain interior after a small deformation of 2% the grain size decreased to UFG regime, the number of dislocations

5
B.B. Wang et al. Materials Science & Engineering A 820 (2021) 141504

Fig. 8. TEM microstructures of FG-4.6, FG-1.2, UFG-0.7 samples after 2% tensile strain (a)–(c) and after final fracture (d)–(e).

decreased dramatically, few dislocations could be found even after final

fracture. The calculated w value of UFG-0.7 sample (0.86 μm) was larger
than the grain size (0.7 μm), so the dislocation substructures were hard
to form in the grain interior due to the rapid recovery at the grain
boundary. Therefore, only a few dislocations piled up at the grain
boundary, and most of the grains were free of dislocation, which was
consisted with the strain softening performance and poor dislocation
storage ability [51].

4.2. Hall-Petch relationship

To make the comparison of Hall-Petch relationship more impartial,

materials prepared by FSP and SPD methods were compared in this
study [27,51–53]. The purity has an effect on the intercept of the
Hall-Petch relationship [8,54]. Therefore, all data were normalized by
making the intercept values consistent with the high purity Al to exclude
the influence of material purity, and the results were shown in Fig. 9.
Clearly, the Hall-Petch relationships can be divided into three stages. At Fig. 9. Hall-Petch relationships of pure Al samples [27,51–53].
the stage I (d ≥ 25 μm), the primary barrier for dislocation glide was
dominated by dislocation substructures during the tensile deformation can be written as the following equation:
process. The data points for CG materials followed the Hall-Petch rela­
σ II,1μm≤d<25 μm = σ FG + kII ⋅d− 1/2
= 28 ​ + ​ (60 ∼ 70) ​ ⋅d− 1/2
(4)
tionship with a lower kI value between 25 and 41 MPa⋅ μm1/2 [27]. kI
value can be affected by texture, dislocation density, and the fraction of In this stage, the specific surface area of HAGBs was increased as the
HAGBs, so the range was relatively wide due to the various micro­ decrease of grain size. Therefore, dislocation tangled or assembled at the
structures prepared by different methods. The YS in stage I can be HAGBs brought stronger strengthening effect than the LAGBs and
written as the following equation: dislocation substructures, resulting in an increase of kII compared to kI .
Previous studies indicated that the dislocation strengthening was even
σ I, d≥25 μm = σ 0 + kI ⋅d− 1/2
= 9.7 ​ + ​ (25 ∼ 41) ​ ⋅d− 1/2
(3)
larger than grain boundary strengthening due to high density of dislo­
At the stage II (1 μm ≤ d < 25 μm), the kII value was between 60 and cations in the materials prepared by cold rolling and ARB methods,
70 MPa⋅μm1/2, which was larger than that of stage I. The YS in stage II resulting in a higher Hall-Petch slope of 131 MPa⋅μm1/2 [52] in stage II,

6
B.B. Wang et al. Materials Science & Engineering A 820 (2021) 141504

as illustrated by arrows in Fig. 9. On the contrary, the dislocation density Hall-Petch relationship, but ignored the influence of grain boundary
of FSP material was very low attributed to the DRX process. characteristics [56,57]. However, the strengthening effect was different
At the stage III (d < 1 μm), the kIII value was between 200 and 400 between LAGBs and HAGBs, leading to a strong influence on the rela­
MPa⋅μm1/2 in UFG regime, which was significantly higher than CG and tionship of grain size and tensile properties. Therefore, grain boundary
FG materials. The microstructural differences were very large for strengthening divides into LAGBs strengthening and HAGBs strength­
different UFG materials due to the different deformation degree during ening, which can be written as:
SPD processes, and a high fraction of non-equilibrium grain boundaries √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ √̅̅̅̅̅̅̅̅̅̅̅
were usually obtained. Therefore, the deviation of the actual grain size σ GB = σ LAGB + σ HAGB = kLAGB
f2◦ ≤θ<15◦
+ kHAGB
fθ≥15◦
(8)
was very big, resulting in a big range of kIII value. The YS in stage III can d d
be written as the following equation:
where kLAGB is the Hall-Petch slope for LAGBs strengthening, kHAGB is the
σ III, d<1μm = σ UFG + kIII ⋅d − 1/2 − 1/2 ​
= − 307 ​ + ​ (200 ∼ 400) ​ ⋅d (5) Hall-Petch slope for HAGBs strengthening. The kLAGB can be obtained at
the situation where the LAGB dominated the grain boundary strength­
In equation (5), the intercept value (σ UFG ) didn’t represent the lattice ening for the CG materials at the stage I. kLAGB was equal to 25 MPa⋅μm1/
friction stress, which contained the lattice friction stress and other 2
, which was the minimize value of kI through fitting the Hall-Petch
strengthening mechanisms, and a negative intercept value (− 307 MPa) relationship of LAGBs, as shown in Fig. 9. σ HAGB and kHAGB can be
was suited for the Hall-Petch relationship in stage III. Yu et al. [51] had calculated by the following equations:
reported that the fitting result of k value for ECAP UFG material was
much higher than that of FG material, and the initial inhomogeneous σ HAGB = σYS − σ0 − σ dis − σLAGB (9)
flow resulted in the positive deviation of the Hall–Petch relationship in √̅̅̅̅̅̅̅̅̅̅̅
the UFG regime. Meanwhile, the dislocation densities of UFG materials kHAGB = σHAGB
/ fθ≥15◦
(10)
produced by ECAP and ARB were higher than the FG materials prepared d
by combination of SPD and following annealing processes, which would The values of σHAGB and kHAGB of the FG-4.6 and FG-3.2 samples were
contribute higher dislocation strengthening in SPD UFG material than calculated through equation (9) and (10), as shown in Table 2. The kHAGB
that in FG material. However, no obvious inhomogeneous flow was value, representing the HAGBs strengthening for pure Al with any grain
observed in UFG-0.7 sample. Furthermore, the dislocation density of FSP size, can be defined as the average value of those of the FG-4.6 and FG-
UFG material was similar to that of FG material and kept a relatively low 3.2 samples, which was equal to 71 MPa⋅μm1/2. In order to verify the
value. Therefore, the kIII value of FSP UFG materials was much higher accuracy of the above results, kLAGB and kHAGB were used to calculate the
than kII and kI , and even higher than that of ECAP UFG material. This YS of FG-1.2 sample. The calculated YS of FG-1.2 sample was 95.5 MPa,
was not be affected by the inhomogeneous flow and increased disloca­ which was nearly equal to the actual result (92.2 MPa), proving the
tion density, so an extra strengthening mechanism should play a more accuracy of klAGB and kHAGB values. Obviously, the kHAGB value (71
important role. MPa⋅μm1/2) is almost three times of klAGB (25 MPa⋅μm1/2), meaning that
the HAGBs are more efficient than LAGBs in material strengthening,
which resulted in a higher Hall-Petch slope in stage.
4.3. Strengthening mechanisms When the grain size decreases to UFG regime (stage III), the grain
refinement increased the probability of dislocation sinking at the
Some models have been proposed to explain the Hall-Petch rela­ HAGBs. The increased recovery rate of dislocations near the grain
tionship and give expression for k value [54]. In these models, the k boundaries led to the decrease of mobile dislocation density. Previous
value was decided by dislocation density, GAR, and grain boundary studies have proposed that significantly reduced dislocation source
characteristics, which were related to the strengthening mechanism of density resulted in an additional strengthening mechanism and a higher
materials. stress for UFG or nanostructured materials by activating alternative
The strengthening mechanism of CG and FG materials mainly con­ dislocation source [27]. Huang et al. [39,58] found that the YS of UFG
tained the grain boundary strengthening and dislocation strengthening material could be increased by annealing, because the generation and
(σ dis ). Most of the dislocations stored in grain interior and LAGBs. Pre­ interaction of dislocations reduced under heat treatment, leading to an
vious research indicated that the LAGBs with misorientation angle increase in YS in order to activate new dislocation sources during tensile
below 2◦ acted as dislocation strengthening and the rest of the grain deformation. Kamikawa et al. [27] called this strengthening mechanism
boundaries were regarded as grain boundary strengthening [52]. Based as dislocation source limited strengthening, which usually appeared in
on this assumption, the YS can be written in the following way: UFG and nanostructure materials where dislocations were significantly
σ YS = σ0 + σ dis + σGB (6) reduced. The dislocation source limited strengthening (σ source ) in
UFG-0.7 sample can be acquired by the following equation:
where σ dis represents the dislocation strengthening, and σ GB is grain
σ source = σ YS, UFG− − σ 0 − σdis − σLAGB − σ HAGB (11)
boundary strengthening. σdis can be written as:
0.7

√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ The σ source value of UFG-0.7 sample was calculated as 29 MPa ac­
σ dis = M αGb ρ0 +
3(1 − fθ≥2◦ )θave,<2◦
(7) cording to equation (11), which reached to about 20% of the YS of UFG-
bd 0.7 sample.
Fig. 10 shows the distribution of different strengthening mechanisms
where fθ≥2◦ represents the fraction of grain boundaries with misorien­
contributed to the YS of some typical samples. Strengthening mecha­
tation angles above 2◦ , θave,<2◦ is the average value of misorientation
nisms of FG materials mainly included dislocation strengthening and
angle below 2◦ , M is the Taylor factor (3.06), α is a constant (0.24), G is
HAGBs strengthening. The contribution of LAGBs strengthening to YS
the shear modulus (26 GPa), b is the burgers vector (0.286 nm), ρ0 is
was less than 10% for FG and UFG materials. Previous research indi­
dislocation density in grain interior [27,52]. The dislocations mainly
cated that the dislocation strengthening was larger than grain boundary
stored in the LAGBs with misorientation angle below 2◦ , and the dislo­
strengthening due to the high density of dislocations stored in LAGBs
cation density in grain interior (ρ0 ) was relatively small due to the DRX
with misorientation angle below 2◦ [52]. However, the dislocation
during FSP, which can be ignored in this study.
strengthening of FSP sample was relatively weak due to the low fraction
Previous research indicated that grain boundary characteristic had a
of LAGBs. The primary strengthening mechanisms of UFG materials
strong effect on the tensile deformation behavior [55]. Most of the
were HAGBs strengthening and dislocation source limited strengthening
previous literatures only focused on the grain size effect on the

7
B.B. Wang et al. Materials Science & Engineering A 820 (2021) 141504

Table 2
Structural parameters for different strengthening mechanisms.
Sample kLAGB (MPa⋅μm1/2) LAGBs strengthening (MPa) Dislocation strengthening (MPa) kHAGB (MPa⋅μm1/2) HAGBs strengthening (MPa)

FG-4.6 25.00 4.75 22.13 70 23.22

FG-3.2 25.00 6.75 17.23 72 33.72
FG-1.2 25.00 9.62 24.51 71 55.28
UFG-0.7 25.00 9.48 20.44 71 73.81

Credit author statement

P. Xue and D.R. Ni led the project. B.B. Wang, G.M. Xie, L.H. Wu, D.R.
Ni and B.L. Xiao designed and performed all of the experiments and
analyzed the data including sample processing, property testing and
microstructure characterization. B.B. Wang and P. Xue wrote the paper
under the guidance of Y.D. Liu and Z.Y. Ma. All of the authors have
discussed the results and revised the paper together.

Data availability

The raw/processed data required to reproduce these findings cannot

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

Declaration of competing interest

The authors declare that they have no known competing financial

Fig. 10. The distributions of different strengthening mechanisms contribute to interests or personal relationships that could have appeared to influence
the YS. the work reported in this paper.

which contributed a higher Hall-Petch slope in stage. Acknowledgements

The calculation substantiates that the HAGBs strengthening in­
creases with the decrease of grain size and acts as a dominant role in This work was supported by the National Natural Science Foundation
mechanical properties. The dislocation source limited strengthening of China under grant No. 52071317, the Open Research Fund from the
reached to about 20% of the YS of UFG materials. Compared to FG State Key Laboratory of Rolling and Automation, Northeastern Univer­
material, the extra strengthening of UFG material, i.e. dislocation source sity (2020RALKFKT009), and the Youth Innovation Promotion Associ­
limited strengthening, required higher stress to nucleate more mobile ation of the Chinese Academy of Sciences (2017236).
dislocations to maintain the plastic flow during tensile tests. Therefore,
dislocation source limited strengthening contributed to higher YS,
References
leading to the loss of work hardening during tensile tests.
[1] R.Z. Valiev, R.K. Islamgaliev, I.V. Alexandrov, Bulk nanostructured materials from
5. Conclusions severe plastic deformation, Prog. Mater. Sci. 45 (2000) 103–189.
[2] Y. Estrin, A. Vinogradov, Extreme grain refinement by severe plastic deformation: a
wealth of challenging science, Acta Mater. 61 (2013) 782–817.
Commercially pure Al (1060) samples with grain sizes range from [3] X.H. An, S.D. Wu, Z.G. Wang, Z.F. Zhang, Significance of stacking fault energy in
0.7 μm to 30.0 μm were produced by FSP with different processing pa­ bulk nanostructured materials: insights from Cu and its binary alloys as model
rameters. Based on the experimental observations and discussions, the systems, Prog. Mater. Sci. 101 (2019) 1–45.
[4] T.G. Langdon, Twenty-five years of ultrafine-grained materials: achieving
following conclusions can be obtained: exceptional properties through grain refinement, Acta Mater. 61 (2013)
7035–7059.
1. Irrespective of grain size, the grain morphology was equiaxed, and [5] L.S. Toth, C. Gu, Ultrafine-grain metals by severe plastic deformation, Mater. Char.
92 (2014) 1–14.
almost no dislocations were observed in grain interior. The distri­ [6] A. Azushima, R. Kopp, A. Korhonen, D.Y. Yang, F. Micari, G.D. Lahoti, P. Groche,
butions of grain boundary misorientation angles were similar to the J. Yanagimoto, N. Tsuji, A. Rosochowski, A. Yanagida, Severe plastic deformation
random distribution and the fraction of the HAGBs was higher than (SPD) processes for metals, CIRP Ann 57 (2008) 716–735.
[7] H. Fujita, T. Tabata, Effect of grain-size and deformation substructure on
70%. mechanical properties of polycrystalline aluminum, Acta Metall. 21 (1973)
2. The Hall-Petch relationships were classified into three stages as the 355–365.
grain size reduced from CG to UFG regime. The strengthening effect [8] N. Hansen, The effect of grain size and strain on the tensile flow stress of
aluminium at room temperature, Acta Metall. 25 (1977) 863–869.
of HAGBs was almost three times of LAGBs, resulting in the positive
[9] H.J. Choi, S.W. Lee, J.S. Park, D.H. Bae, Positive deviation from a Hall-Petch
deviation of Hall-Petch slope in stage II. The further positive devia­ relation in nanocrystalline aluminum, Mater. Trans. 50 (2009) 640–643.
tion of Hall-Petch relationship at UFG region was due to the occur­ [10] W. Xu, L.P. Davila, Tensile nanomechanics and the Hall-Petch effect in
rence of dislocation source limited strengthening. nanocrystalline aluminium, Mater. Sci. Eng., A 710 (2018) 413–418.
[11] S.N. Naik, S.M. Walley, The Hall-Petch and inverse Hall-Petch relations and the
3. The dislocation source limited strengthening reached to about 20% hardness of nanocrystalline metals, J. Mater. Sci. 55 (2020) 2661–2681.
for the YS of UFG material, which resulted in the increased the YS [12] B. Mirshekari, A. Zarei-Hanzaki, A. Barabi, H.R. Abedi, S.J. Lee, H. Fujii, An
and the decreased uniform elongation caused by the losing of work anomalous effect of grain refinement on yield stress in friction stir processed
lightweight steel, Mater. Sci. Eng., A 799 (2021) 140057.
hardening during tensile deformation. [13] A. Zhilyaev, T. Langdon, Using high-pressure torsion for metal processing:
fundamentals and applications, Prog. Mater. Sci. 53 (2008) 893–979.
[14] H. Lanjewar, L.A.I. Kestens, P. Verleysen, Damage and strengthening mechanisms
in severely deformed commercially pure aluminum: experiments and modeling,
Mater. Sci. Eng., A 800 (2021) 140224.

8
B.B. Wang et al. Materials Science & Engineering A 820 (2021) 141504

[15] N. Tsuji, Y. Ito, Y. Saito, Y. Minamino, Strength and ductility of ultrafine grained [36] S. Shukla, M. Komarasamy, R.S. Mishra, Grain size dependence of fatigue
aluminum and iron produced by ARB and annealing, Scripta Mater. 47 (2002) properties of friction stir processed ultrafine-grained Al-5024 alloy, Int. J. Fatig.
893–899. 109 (2018) 1–9.
[16] Q. Liu, X. Huang, D.J. Lloyd, N. Hansen, Microstructure and strength of [37] H. Zhao, Q. Pan, Q. Qin, Y. Wu, X. Su, Effect of the processing parameters of
commercial purity aluminium (AA 1200) cold-rolled to large strains, Acta Mater. friction stir processing on the microstructure and mechanical properties of 6063
50 (2002) 3789–3802. aluminum alloy, Mater. Sci. Eng., A 751 (2019) 70–79.
[17] O.V. Mishin, D. Juul Jensen, N. Hansen, Microstructures and boundary populations [38] Y. Hu, H. Liu, S. Du, Achievement of high-strength 2219 aluminum alloy joint in a
in materials produced by equal channel angular extrusion, Mater. Sci. Eng., A 342 broad process window by ultrasonic enhanced friction stir welding, Mater. Sci.
(2003) 320–328. Eng., A (2020) 140587.
[18] T. Huang, L. Shuai, A. Wakeel, G. Wu, N. Hansen, X. Huang, Strengthening [39] X. Huang, N. Kamikawa, N. Hansen, Strengthening mechanisms in nanostructured
mechanisms and Hall-Petch stress of ultrafine grained Al-0.3%Cu, Acta Mater. 156 aluminum, Mater. Sci. Eng., A 483–484 (2008) 102–104.
(2018) 369–378. [40] A.B. Lopes, F. Barlat, J.J. Gracio, J.F.F. Duarte, E.F. Rauch, Effect of texture and
[19] M. Kawasaki, Z. Horita, T.G. Langdon, Microstructural evolution in high purity microstructure on strain hardening anisotropy for aluminum deformed in uniaxial
aluminum processed by ECAP, Mater. Sci. Eng., A 524 (2009) 143–150. tension and simple shear, Int. J. Plast. 19 (2003) 1–22.
[20] M. Reihanian, R. Ebrahimi, N. Tsuji, M.M. Moshksar, Analysis of the mechanical [41] D.J. Jensen, N. Hansen, Flow-stress anisotropy in aluminum, Acta Metall. Mater.
properties and deformation behavior of nanostructured commercially pure Al 38 (1990) 1369–1380.
processed by equal channel angular pressing (ECAP), Mater. Sci. Eng., A 473 [42] O. Engler, Texture and anisotropy in cold rolled and recovery annealed AA 5182
(2008) 189–194. sheets, Mater. Sci. Technol. 31 (2014) 1058–1065.
[21] S. Qu, X.H. An, H.J. Yang, C.X. Huang, G. Yang, Q.S. Zang, Z.G. Wang, S.D. Wu, Z. [43] Z.Y. Ma, Friction stir processing technology: a review, Metall. Mater. Trans. 39
F. Zhang, Microstructural evolution and mechanical properties of Cu-Al alloys (2008) 642–658.
subjected to equal channel angular pressing, Acta Mater. 57 (2009) 1586–1601. [44] X. Meng, Y. Huang, J. Cao, J. Shen, J.F. dos Santos, Recent progress on control
[22] X.H. An, S.D. Wu, Z.G. Wang, Z.F. Zhang, Enhanced cyclic deformation responses strategies for inherent issues in friction stir welding, Prog. Mater. Sci. 115 (2021)
of ultrafine-grained Cu and nanocrystalline Cu-Al alloys, Acta Mater. 74 (2014) 100706.
200–214. [45] A. Mishra, B. Kad, F. Gregori, M. Meyers, Microstructural evolution in copper
[23] P. Bazarnik, Y. Huang, M. Lewandowska, T.G. Langdon, Structural impact on the subjected to severe plastic deformation: experiments and analysis, Acta Mater. 55
Hall-Petch relationship in an Al-5Mg alloy processed by high-pressure torsion, (2007) 13–28.
Mater. Sci. Eng., A 626 (2015) 9–15. [46] F. Tang, J.M. Schoenung, Strain softening in nanocrystalline or ultrafine-grained
[24] A. Loucif, R.B. Figueiredo, T. Baudin, F. Brisset, R. Chemam, T.G. Langdon, metals: a mechanistic explanation, Mater. Sci. Eng., A 493 (2008) 101–103.
Ultrafine grains and the Hall-Petch relationship in an Al-Mg-Si alloy processed by [47] Y.J. Li, X.H. Zeng, W. Blum, Transition from strengthening to softening by grain
high-pressure torsion, Mater. Sci. Eng., A 532 (2012) 139–145. boundaries in ultrafine-grained Cu, Acta Mater. 52 (2004) 5009–5018.
[25] X.H. An, Q.Y. Lin, S.D. Wu, Z.F. Zhang, R.B. Figueiredo, N. Gao, T.G. Langdon, [48] R. Kapoor, J.B. Singh, J.K. Chakravartty, High strain rate behavior of ultrafine-
Significance of stacking fault energy on microstructural evolution in Cu and Cu-Al grained Al-1.5 Mg, Mater. Sci. Eng., A 496 (2008) 308–315.
alloys processed by high-pressure torsion, Philos. Mag. A 91 (2011) 3307–3326. [49] N. Kumar, R.S. Mishra, C.S. Huskamp, K.K. Sankaran, Critical grain size for change
[26] R.W. Armstrong, 60 Years of Hall-petch: past to present nano-scale connections, in deformation behavior in ultrafine grained Al-Mg-Sc alloy, Scripta Mater. 64
Mater. Trans. 55 (2014) 2–12. (2011) 576–579.
[27] N. Kamikawa, X. Huang, N. Tsuji, N. Hansen, Strengthening mechanisms in [50] H. Mughrabi, On the grain-size dependence of metal fatigue: outlook on the fatigue
nanostructured high-purity aluminium deformed to high strain and annealed, Acta of ultrafine-grained metals, Investigations and Applications of Severe Plastic
Mater. 57 (2009) 4198–4208. Deformation 80 (2000) 241–253.
[28] C.S. Pande, B.B. Rath, M.A. Imam, Effect of annealing twins on Hall-Petch relation [51] C.Y. Yu, P.W. Kao, C.P. Chang, Transition of tensile deformation behaviors in
in polycrystalline materials, Mater. Sci. Eng., A 367 (2004) 171–175. ultrafine-grained aluminum, Acta Mater. 53 (2005) 4019–4028.
[29] Y.Z. Tian, S. Gao, L.J. Zhao, S. Lu, R. Pippan, Z.F. Zhang, N. Tsuji, Remarkable [52] N. Kamikawa, T. Hirochi, T. Furuhara, Strengthening mechanisms in ultrafine-
transitions of yield behavior and Lüders deformation in pure Cu by changing grain grained and sub-grained high-purity aluminum, Metall. Mater. Trans. 50 (2019)
sizes, Scripta Mater. 142 (2018) 88–91. 234–248.
[30] X.H. An, S.D. Wu, Z.F. Zhang, R.B. Figueiredo, N. Gao, T.G. Langdon, Enhanced [53] Y.S. Sato, M. Urata, H. Kokawa, K. Ikeda, Hall-Petch relationship in friction stir
strength–ductility synergy in nanostructured Cu and Cu-Al alloys processed by welds of equal channel angular-pressed aluminium alloys, Mater. Sci. Eng., A 354
high-pressure torsion and subsequent annealing, Scripta Mater. 66 (2012) (2003) 298–305.
227–230. [54] V. Bata, E.V. Pereloma, An alternative physical explanation of the Hall-Petch
[31] K. Matsumoto, T. Shibayanagi, Y. Umakoshi, Effect of grain-size on grain relation, Acta Mater. 52 (2004) 657–665.
orientations and grain-boundary-character-distribution in recrystallized Al-0.3% [55] P.L. Sun, C.Y. Yu, P.W. Kao, C.P. Chang, Influence of boundary characters on the
Mg alloy, Scripta Metall. Mater. 33 (1995) 1321–1326. tensile behavior of sub-micron-grained aluminum, Scripta Mater. 52 (2005)
[32] P.L. Sun, E.K. Cerreta, G.T. Gray III, J.F. Bingert, The effect of grain size, strain rate, 265–269.
and temperature on the mechanical behavior of commercial purity aluminum, [56] Y.Z. Tian, Y.P. Ren, S. Gao, R.X. Zheng, J.H. Wang, H.C. Pan, Z.F. Zhang, N. Tsuji,
Metall. Mater. Trans. 37 (2006) 2983–2994. G.W. Qin, Two-stage Hall-Petch relationship in Cu with recrystallized structure,
[33] M. Furukawa, Z. Horita, M. Nemoto, R.Z. Valiev, T.G. Langdon, Microhardness J. Mater. Sci. Technol. 48 (2020) 31–35.
measurements and the Hall-Petch relationship in an Al-Mg alloy with [57] Y.S. Sato, M. Urata, H. Kokawa, K. Ikeda, Hall-Petch relationship in friction stir
submicrometer grain size, Acta Mater. 44 (1996) 4619–4629. welds of equal channel angular-pressed aluminium alloys, Mater. Sci. Eng., A 354
[34] R.S. Mishra, Z.Y. Ma, Friction stir welding and processing, Mater. Sci. Eng. R 50 (2003) 298–305.
(2005) 1–78. [58] X.X. Huang, N. Hansen, N. Tsuji, Hardening by annealing and softening by
[35] T. Hirata, T. Oguri, H. Hagino, T. Tanaka, S.W. Chung, Y. Takigawa, K. Higashi, deformation in nanostructured metals, Science 312 (2006) 249–251.
Influence of friction stir welding parameters on grain size and formability in 5083
aluminum alloy, Mater. Sci. Eng., A 456 (2007) 344–349.