Materials Science & Engineering A 576 (2013) 156–159

Contents lists available at SciVerse ScienceDirect

Materials Science & Engineering A

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

Rapid communication

Shear punch superplasticity in equal-channel angularly pressed

Mg–12Li–1Zn alloy
M. Karami, R. Mahmudi n
School of Metallurgical and Materials Engineering, College of Engineering, University of Tehran, Tehran, Iran

art ic l e i nf o a b s t r a c t

Article history: The superplasticity of a fine-grained Mg–12Li–1Zn alloy, processed by equal-channel angular pressing
Received 21 January 2013 (ECAP), was studied by shear punch testing. The strain rate sensitivity index of 0.45 and activation energy
Received in revised form of 71 kJ mol–1 is indicative of a superplastic shear behavior dominated by grain boundary sliding.
27 March 2013
& 2013 Elsevier B.V. All rights reserved.
Accepted 28 March 2013
Available online 9 April 2013

Keywords:
Superplasticity
ECAP
Shear punch test

1. Introduction by plotting shear stress against normalized displacement. SPT has

The superplasticity of materials relies primarily on the small
size of the grains, preferably of high-angle type [1]. Among many
possible means of attaining such fine structures, severe plastic
deformation techniques including processing by equal-channel
angular pressing (ECAP) have attracted special attention [2,3].
The ECAP method is a promising technique capable of producing
fine-grained microstructures suitable for superplastic deformation
of magnesium alloys [4,5]. In this process, a billet is pressed
repetitively through an angular die, to undergo a very high shear
strain without any change in the cross-sectional area, the princi-
ples of which have been reviewed by Valiev and Langdon [6].
Despite the great attention paid to the superplastic behavior of the
ECAPed Mg alloys [7], only a few studies have been performed on
the Mg–Li alloys [8]. Most of these studies have mainly concen-
trated on the determination of superplasticity by the conventional
tensile methods. The Mg–Li alloy with Li content greater than
11 wt% exhibits a single bcc β phase structure. This structure is
expected to provide more slip systems and higher ductility values,
as compared to the hcp-structured Mg–Li alloys with lower Li
contents. Therefore, it would be of interest to evaluate the super-
plasticity of bcc-structured Mg–Li alloys.
Shear punch test (SPT) is a miniature testing technique, which
has been recently used for evaluating strength properties of both
cast [9] and wrought [10,11] magnesium alloys. In this test,
mechanical properties such as shear yield stress (SYS), ultimate
shear strength (USS) and shear elongation values can be obtained
n
Corresponding author. Tel.: +98 21 8208 4137; fax: +98 21 8800 6076.
E-mail address: mahmudi@ut.ac.ir (R. Mahmudi)
not previously been used for evaluating superplastic behavior of
any Mg alloy through determination of the parameters commonly
considered as characteristics of superplasticity. One of these
parameters is the strain-rate sensitivity (SRS) index (m), which
can be used to assess the superplastic behavior. It is therefore the
aim of the present study to characterize the superplastic shear
behavior of a fine-grained Mg–12%Li–1%Zn alloy in the tempera-
ture range 473–548 K (T4 0.5 Tm), by measuring the SRS indices
and activation energies. It is also intended to evaluate deformation
mechanisms of the alloys having various grain sizes, obtained by
conventional extrusion and by ECAP.
2. Experimental procedure
The material used was an Mg–Li–Zn (LZ121) alloy containing
11.8 wt% Li and 0.8 wt% Zn. Extrusion was carried out with an
extrusion ratio of 11:1 at 573 K. Pressing of the extruded billets
was conducted at 473 K using a solid die with channel angles of
φ ¼901 and ψ¼201. Samples were pressed at a speed of 1 mm s–1
for two passes through the die using route BC.
One-millimeter thick slices were cut from the extruded and
ECAPed bars perpendicular to the pressing direction and ground to
a thickness of about 0.7 mm. Shear punch tests were conducted in
the temperature range 473–548 K and shear strain rate range
0.001–0.060 s–1, using a SANTAM universal testing machine.
A shear punch fixture with a 3.175 mm diameter flat cylindrical
punch and a 3.225 mm diameter receiving-hole was used. The
assembly of the specimen and fixture were accommodated by the
split furnace, and held for 15 min to establish thermal equilibrium

0921-5093/$ - see front matter & 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.msea.2013.03.080
 M. Karami, R. Mahmudi / Materials Science & Engineering A 576 (2013) 156–159 157

in the testing arrangement. The applied load P was measured a power-law relationship
automatically as a function of punch displacement; the data were
ε_ ¼ As1=m expð−Q =RTÞ ð3Þ
acquired by a computer so as to determine the shear stress using
the relationship [12] where A is a material parameter, m denotes SRS index, Q is the
deformation activation energy, R is the universal gas constant and
P
τ¼ ð1Þ T is the temperature. s can be either yield stress or peak stress.
πdt This equation can be simply modified for evaluating superplastic
where P is the punch load, t is the specimen thickness and d is the behavior in the SPT method by replacing ε_ with γ_ and s with τ. This
average of the punch and die diameters. The microstructure of the can be conducted using the Von Mises criterion for a state of pure
materials was revealed by optical microscopy, and average grain shear
pffiffiffi of kinematically
pffiffiffi hardening materials, which presents
sizes of 8 and 48 mm were measured for the ECAPed and extruded s¼ 3τ and ε ¼ (1/ 3)γ. Therefore, the modified form of Eq. (3)
conditions, respectively. can be rewritten in the form of
γ_ ¼ A1 τ1=m expð−Q =RTÞ ð4Þ
where γ_ is shear strain rate and A1 is a material parameter. Due to
3. Results and discussion the constancy of Q at a given temperature, it is possible to
determine the SRS index (m) from
Typical SPT curves, plotted as shear stress (τ) against normal-  
∂ln τ
ized punch displacement (δ¼h/t, where h is the displacement), are m¼ ð5Þ
∂ln_γ T
shown for both ECAPed and extruded materials at different shear
strain rates in Fig. 1. The shear strain rates were calculated using The above equation has been successfully used for Cu [13] and
the following equation [12]: Sn–Sb [14] alloys. The SRS index of 0.57 obtained by SPT of fine-
grained ECAPed Sn–5Sb alloy is in a good agreement with m ¼0.60
1 Z_ measured by the impression technique of the same material [1].
γ_ ¼ ð2Þ
2W The calculation of activation energy can be done by rearranging
where Z_ is the punch-displacement rate and W is the die-punch Eq. (4) and differentiating with respect to 1/T to obtain a relation-
clearance. As can be seen, similar to conventional superplastic ship of the form
 
tensile curves, after elastic region and a rather short strain ∂ln τ
Q ¼ nR ð6Þ
hardening stage, almost all of the SPT curves show a relatively ∂ð1=TÞ γ_
long strain softening region, which moves toward a near steady-
state region in the ECAPed condition. It is also evident that the where n is the stress exponent, which is equivalent to 1/m.
flow stress of the alloy significantly increases with increasing To determine SRS index, the shear yield stress was plotted on a
strain rate in both conditions, indicating the positive strain rate double-logarithmic scale against the shear strain rate at test
sensitivity of the material at high temperatures [1]. temperatures of 473, 498, 523 and, 548 K, as depicted in Fig. 2a
It is generally accepted that the high-temperature flow stress
(s) of superplastic materials can be related to the strain rate (_ε) by
100

1: γ = 6.0 × 10-2s-1 І ІІ ІІІ

Shear stress (MPa)

ECAPed
T = 548K 2: γ = 3.0 × 10-2s-1
3: γ = 1.2 × 10-2s-1
4: γ = 6.0 × 10-3s-1
τ (MPa)

10
5: γ = 1.2 × 10-3s-1
473 K, m = 0.21
498 K, m = 0.25
523 K, m = 0.30
ECAPed 548 K, m = 0.44
1
-4 10-3 10-2 10-1 100
10
Shear strain rate (s-1)
δ (mm/mm)
100
Extruded 1: γ = 6.0 × 10-2s-1
Shear stress (MPa)

T = 548 K 2: γ = 3.0 × 10-2s-1

3: γ = 1.2 × 10-2s-1
4: γ = 6.0 × 10-3s-1
τ (MPa)

5: γ = 1.2 × 10-3s-1 10

473 K, m = 0.18
498 K, m = 0.23
523 K, m = 0.25
Extruded 548 K, m = 0.27
1
10-4 10-3 10-2 10-1 100
Shear strain rate (s-1)
δ (mm/mm)
Fig. 2. Shear stress as a function of shear strain rate for (a) ECAPed, and
Fig. 1. SPT curves of the material obtained at different strain rates and T ¼ 548 K. (b) extruded conditions at various temperatures.
158 M. Karami, R. Mahmudi / Materials Science & Engineering A 576 (2013) 156–159

and b for the ECAPed and extruded conditions, respectively. It can increasing temperature, toward a typical sigmoidal shape having
be observed that the strain rate dependence of the shear stress for three distinct regions: І, ІІ and ІІІ (Fig. 2a). According to Eq. (5), the
the fine-grained ECAPed material deviates from linearity with slope of the curves in any of these regions gives the corresponding
SRS index (m). For the temperatures of 523 and 548 K, the SRS
index showed maximum values of 0.30 and 0.45 in the inter-
100 mediate strain-rate region ІІ. This is in agreement with those
obtained by conventional tensile test for superplastic Mg–Li alloys
Qave = 71 kJ/mol
[8,15]. In contrast to the ECAPed condition, the strain rate
Shear stress (MPa)

dependence of the extruded material is well expressed by linear

relationships at all testing temperatures with slopes in the range
10 0.18–0.27 (Fig. 2b). It is notable that m-value increases in both
conditions, as the temperature increases from 473 to 548 K.
γ = 3.0 × 10-2 s-1 The results of tensile tests on a superplastic Mg–8Li–2Zn alloy
γ = 1.2 × 10-2 s-1 are indicative of m-values increasing from 0.24 to 0.51 as the
γ = 6.0 × 10-3 s-1 temperature increased from 473 to 593 K [15]. Other similar
ECAPed γ = 1.2 × 10-3 s-1
1 results obtained for the compression testing of Mg–11.5Li–1.5Al
1.7 1.9 2.1 2.3 have yielded n-values of about 3.0 (m ¼0.33) in the temperature
1000/T (K-1) range 473–723 K [16]. The activation energy of the ECAPed condi-
tion in stage ІІ of the sigmoidal curves (71 kJ mol–1), as the
representative activation energy for the superplastic ECAPed
100 material, could be acquired using the slope of the fitted lines in
Fig. 3a. For the extruded condition, Fig. 3b, average activation
Shear stress (MPa)

energy of about 106 kJ mol–1 was attained over the whole range of
Qave = 106 kJ/mol
strain rates tested.
The different microstructures of the processed materials, shown in
10 Fig. 4, can explain the observed difference in the deformation behavior
γ = 6.0 × 10-2 s-1 of the two tested conditions. It is clear that in both conditions,
γ = 3.0 × 10-2 s-1 recrystallization has comprehensively occurred, leading to the forma-
γ = 1.2 × 10-2 s-1 tion of new equiaxed grains. The extruded single-phase microstructure
Extruded γ = 6.0 × 10-3 s-1 exhibits some large grains together with many finer grains, forming an
1 inhomogeneous β-Li grain structure with an average grain size of
1.7 1.9 2.1 2.3
48 mm (Fig. 4b). In the ECAPed condition, however, a more refined
1000/T (K-1) microstructure with an average grain size of 8 mm could be achieved,
Fig. 3. Temperature-dependence of the shear stress for (a) ECAPed, and as exhibited in Fig. 4a. The mechanism of grain refinement during the
(b) extruded conditions. ECAP process can be described as continuous dynamic recovery and
recrystallization (CDRR), in which a combination of mechanical shear-
ing and subsequent dynamic recovery, recrystallization and growth of
grains and subgrain cells produce refined and equiaxed grains [17].
This fine-grained microstructure with high-angle grain boundaries
may give rise to grain boundary sliding (GBS), and thereby high SRS
indices in the ECAPed material. The microstructure of the ECAPed
material after shear deformation at 548 K is depicted in Fig. 5. It is
evident that the material has relatively retained its fine grain structure
after deformation. Moreover, despite its high ductility provided by
superplastic behavior, the ECAPed sample shows no visible flow
pattern after extensive deformation. This may give rise to the
possibility of a grain boundary movement mechanism such as grain
boundary sliding, which might have been activated to cause the

Fig. 4. Optical micrographs of the material for (a) ECAPed, and (b) extruded Fig. 5. Optical micrograph of the ECAPed material after SPT at 548 K with a grain
conditions with grain sizes of 8 and 48 mm, respectively. size of about 11 mm.
 M. Karami, R. Mahmudi / Materials Science & Engineering A 576 (2013) 156–159 159

Table 1 particularly advantageous when the material is only available as

Values of m, n and activation energies of the ECAPed and extruded materials. small test pieces such as those usually obtained by severe plastic
deformation processes. The deformation behavior of the fine-
Condition Test temperature (K) m-value n-value Q (kJ mol−1)
grained ECAPed material was characterized by sigmoidal shear
ECAPed 473 0.21 4.76 71 stress–strain rate curves with three distinct regions, in which the
498 0.25 4.00 intermediate region corresponds to a superplastic behavior. The
523 0.30 3.33 SRS index of 0.45 found at 548 K, and the activation energy of
548 0.44 2.27
71 kJ mol–1 are indicative of a grain boundary sliding mechanism.
Extruded 473 0.18 5.55 106 The extruded material could not develop m-values higher
498 0.23 4.35
than 0.27.
523 0.25 4.00
548 0.27 3.70

Acknowledgments

observed superplasticity in the fine grained ECAPed structure. Due to

The authors thank the Iran National Science Foundation (INSF)
the fact that GBS plays a minor role in the deformation of the coarse-
for financial support of this work under Grant no. 91041956.
grained materials, the m-values in the extruded condition stand in the
lower levels, compared to those of ECAPed condition. The increase of
m-value with temperature in both conditions might be the result of
References
the easier dislocation movement and diffusional processes at higher
temperatures.
[1] R. Mahmudi, R. Alizadeh, A.R. Geranmayeh, Scr. Mater. 64 (2011) 521–524.
It is believed that the stress exponent values of about 2, which [2] S.M. Masoudpanah, R. Mahmudi, Mater. Sci. Eng. A526 (2009) 22–30.
correspond to m¼ 0.5, are associated with GBS controlled by lattice [3] S.A. Torbati-Sarraf, R. Mahmudi, Mater. Sci. Eng. A527 (2010) 3515–3520.
or grain boundary diffusion [18]. Therefore, stress exponent value [4] M. Furui, H. Kitamura, H. Anada, T.G. Langdon, Acta Mater. 55 (2007)
1083–1091.
of 2.3 and an activation energy of 71 kJ mol–1, which is close to [5] R.B. Figueiredo, T.G. Langdon, Scr. Mater. 61 (2009) 84–87.
67 kJ mol–1 for the grain boundary diffusion in β-Li [19], may [6] R.Z. Valiev, T.G. Langdon, Prog. Mater. Sci. 51 (2006) 881–981.
propose that grain boundary sliding could be the dominant [7] Y. Miyahara, Z. Horita, T.G. Langdon, Mater. Sci. Eng. A420 (2006) 240–244.
[8] M. Furui, C. Xu, T. Aida, M. Inoue, H. Anada, T.G. Langdon, Mater. Sci. Eng.
deformation mechanism in the ECAPed condition at 548 K. For
A410–411 (2005) 439–442.
the coarse-grained extruded material, with n-values of 3.7–5.5 and [9] R. Alizadeh, R. Mahmudi, Mater. Sci. Eng. A527 (2010) 3975–3983.
Q¼106 kJ mol–1, which is very close to the activation energy for [10] S.M. Masoudpanah, R. Mahmudi, Mater. Des. 31 (2010) 3512–3517.
lattice diffusion in β-Li (103 kJ mol–1) [20], it is suggested that [11] M. Karami, R. Mahmudi, Mater. Lett. 81 (2012) 235–238.
[12] R.K. Guduru, K.A. Darling, R. Kishore, R.O. Scattergood, C.C. Koch, K.L. Murty,
lattice diffusion-controlled dislocation climb is the creep mechan- Mater. Sci. Eng. A395 (2005) 307–314.
ism. Values of m, n and activation energies, for both conditions, are [13] R.K. Guduru, Ph.D. Thesis, NC State University, Raleigh, NC, USA, 2006.
summarized in Table 1. [14] R. Mahmudi, R. Alizadeh, Sh. Azhari, Mater. Lett. 97 (2013) 44–46.
[15] X. Liu, H. Zhan, S. Gu, Z. Qu, R. Wu, M. Zhang, Mater. Sci. Eng. A528 (2011)
6157–6162.
[16] O. Sivakesavam, Y.V.R.K. Prasad, Mater. Sci. Eng. A323 (2002) 270–277.
4. Conclusions [17] C.W. Su, L. Lu, M.O. Lai, Mater. Sci. Eng. A434 (2006) 227–236.
[18] T.G. Langdon, Acta Metall. Mater 42 (1994) 2437–2443.
[19] E.M. Taleff, O.A. Ruano, J. Wolfenstine, O.D. Sherby, J. Mater. Res. 7 (1992)
It is demonstrated that the SPT, as an easy-to-perform method,
2131–2135.
is capable of evaluating the shear superplastic behavior of a fine- [20] P. Metenier, G. Gonzalez-Doncel, O.A. Ruano, J. Wolfenstine, O.D. Sherby,
grained LZ121 alloy by measuring the SRS index. This can be Mater. Sci. Eng. A125 (1990) 195–202.