Materials Science and Engineering A 528 (2011) 3722–3729

Contents lists available at ScienceDirect

Materials Science and Engineering A

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

Analysis of texture evolution in pure magnesium and the magnesium alloy AM30
during rod and tube extrusion
Somjeet Biswas a , Satyam Suwas a,∗ , R. Sikand b , Anil K. Gupta c
a
Department of Materials Engineering, Indian Institute of Science, Bangalore 560012, India
b
National Physical Laboratory, New Delhi 110012, India
c
Advance Materials and Processes Research Institute, Bhopal 462026, India

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

Article history: The evolution of microstructure and texture during extrusion of pure magnesium and its single phase
Received 3 October 2010 alloy AM30 has been studied experimentally as well as by crystal plasticity simulation. Microstructure and
Received in revised form 10 January 2011 micro-texture were characterized by electron back scattered diffraction (EBSD), bulk-texture was mea-
Accepted 10 January 2011
sured using X-ray diffraction and deformation texture simulations were carried out using visco-plastic
Available online 16 January 2011
self consistent (VPSC) model. In spite of clear indications of the occurrence of dynamic recrystallization
(DRX), simulations were able to reproduce the experimental textures successfully. This was attributed
Keywords:
to the fact that the textures were c-type fibers with their axis of rotation parallel to the c-axis and DRX
Magnesium
Texture
leads to simply rotate the texture around the c-axis.
Extrusion © 2011 Elsevier B.V. All rights reserved.
Dynamic recrystallization

1. Introduction reported for the extrusion process, as it provides the simplest

method to convert a cast billet to a finished wrought product. Many
Magnesium and its alloys, due to their lowest density and high variants to the normal extrusion process have been explored. Lee
specific strength amongst various structural materials, are poten- et al. [5] have shown that two-phase magnesium alloys formed
tial candidates for many automotive and aerospace applications by reciprocating extrusion may have superior strength as well
[1]. However, the use of magnesium as a wrought product is lim- as superplasticity. Chen et al. [6] have shown that an increase
ited by its poor workability at ambient temperatures [2]. This is in extrusion ratio of AZ31 Mg alloy leads to grain refinement
attributed to the inadequate number of slip systems in magnesium and improvement of mechanical properties. Murai et al. [7] have
owing to its hexagonal crystal structure. This limitation can only be observed that significant grain refinement is achieved by decreas-
overcome at higher forming temperatures. ing extrusion temperature and speed, resulting in an increase in
Since the minimum recrystallization temperature of magne- ductility.
sium is less than ∼100 ◦ C (0.4TM ), higher deformation temperatures The mechanical properties of magnesium alloys strongly
lead to dynamic (DRX), which, does not allow the grain size to be depend on texture due to the inherent anisotropy of the hexagonal
sufficiently reduced to enable strengths much higher than that can crystal structure. Processes like rolling and extrusion specifically
be achieved with castings. Therefore, in spite of the inherent lim- lead to the development of a strong texture with the basal plane
itation of cast materials, a majority of magnesium products for perpendicular to the deformation direction [8–13]. Although a
industrial applications are produced by die castings and not by great deal of work has been dedicated to establishing a relation-
wrought processing [3]. ship between the initial orientation and deformation mechanisms
Much research is underway to understand the deformation [14–18], the relationship between deformation mechanism and
behaviour of Mg and its alloys in order to increase formability and texture is limited to experimental observations only. It is expected
strength mostly by the methods to reduce grain size and also by tai- that a clear understanding of the underlying mechanism during
loring a favourable texture. Processing techniques such as rolling, extrusion could be developed through crystal plasticity based sim-
forging and extrusion are the conventional methods exploited to ulations. This study was, therefore, undertaken to investigate the
obtain the refined grain size [4]. Amongst these, most work is evolution of texture in pure magnesium and a newly developed
magnesium alloy AM30 in order to develop a detailed insight
into the deformation mechanisms that occur during extrusion.
∗ Corresponding author. Tel.: +91 80 22933245; fax: +91 80 23600472. Crystal plasticity simulations were coupled with the microstruc-
E-mail address: [email protected] (S. Suwas). tural investigation by electron backscatter diffraction (EBSD)

0921-5093/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.msea.2011.01.021
 S. Biswas et al. / Materials Science and Engineering A 528 (2011) 3722–3729 3723

technique to understand the micro-mechanisms involved in the is the uncorrelated misorientation distribution for the randomly
process. textured sample. From the correlated misorientation distribution
curves, it can be observed that for all the cases, a high fraction of
2. Experimental procedure (2–15◦ ) low angle grain boundaries (LAGBs) and a rather diffuse
population of (15–180◦ ) high angle grain boundaries (HAGBs) were
2.1. Materials and processing present in the extruded samples. For all the cases, peaks at ∼30◦
and ∼86◦ could be observed in the correlated and the uncorrelated
Ingots of unalloyed Mg and the alloy AM30 (Mg, 3% Al, 0.3% Mn) misorientation distribution profiles. The peak at 30◦ might corre-
were prepared by melting the starting materials in a steel crucible spond to 30◦ rotation of the deformation texture along the c-axis,
with SF6/CO2 protective cover gas. Billets 75 mm in diameter and as reported during recrystallization [19,20]. Alternately, it may also
230 mm long were cast at about 700 ◦ C in a permanent mould. The correspond to {1 0 1̄ 1} − {1 0 1̄ 2} double twinning that forms with
billets were preheated to 400 ◦ C for 2 h and extruded in a Well- 38◦ 1 1 2̄ 0 rotation between the twin variants and the untwined
man EnefcoTM 500 ton multipurpose vertical hydraulic press fitted matrix [21]. In that case, the peak at 86◦ might correspond to the
with a circular die. Solid rods, approximately 15 mm in diameter, {1 0 1̄ 2} tensile twin boundaries with 86◦ 1120 rotation orien-
corresponding to an extrusion ratio of 25:1 and tubes of 25 mm tation relationship between the twin variants and the untwined
diameter and 2.5 mm thick, corresponding to an extrusion ratio of matrix (hereafter referred to as ‘tt’). Another twin reported in mag-
19:1 were extruded using boron nitride lubricant at a billet speed of nesium is the {1 0 1̄ 1} compression twinning (ct) with orientation
10 mm s−1 and air cooled. An equivalent von-Mises strain imparted relationship 56◦ 112̄0. The fractions of boundaries corresponding
to the extruded rod was ∼6.4, while the same for the tube was ∼5.9. to different twins were calculated. It was revealed that a very low
fraction of boundaries correspond to tensile twins (<4%). No sig-
nificant boundary fraction corresponding to the compression and
2.2. Characterization double twins could be detected.

Microstructure and micro-texture of the extruded rods and

the extruded tube were obtained using the electron back-scatter
diffraction (EBSD) technique in a Field Emission Gun Scanning Elec-
tron Microscope (FEG-SEM) by FEI Company. In case of extruded 3.2. Texture characterization
rods, the EBSD scans were recorded on the axial plane perpendic-
ular to the extrusion direction (ED), while in the tubes, the same The inverse pole figures for extruded Mg and AM30 rods (Fig. 4)
was recorded at the centre of the thickness in the circumferential obtained from X-ray texture measurements shows a typical extru-
plane perpendicular to the radial direction (RD). Global textures sion texture, as generally observed in ideal hexagonal materials
were measured using a Bruker D8-Discover texture goniome- (with c/a = 1.634) with the basal planes along the periphery of the
ter using CuK␣ radiation. During X-ray texture measurements all pole figure. The texture in both the extruded rods is almost the
the samples were irradiated on the radial plane. Six pole figures same and a similar texture was obtained in the extruded tubes as
namely, (0 0 0 2), (1 0 1̄ 0), (1 0 1̄ 1), (1 0 1̄ 2), (1 1 2̄ 0) and (1 1 2̄ 2) well (after rotating the pole figures to a similar frame of reference).
were measured and orientation deformation functions (ODFs) were Fig. 5a and b shows the (0 0 0 2) pole figures for the extruded
calculated using ADC algorithm using a Labotex Software. Inverse rod and Fig. 5d and e for the tube of pure Mg and AM30 alloy, as
pole figures were calculated from the ODFs. Microhardness tests calculated from EBSD measurements. As mentioned earlier, for the
were carried out on the circumferential plane perpendicular to the extruded rods the EBSD scans were performed on the plane normal
radial direction with a CSM Microhardness tester. to the extrusion direction, while for the tubes, the measurements
were carried out on the plane normal to the radial direction. A typ-
3. Experimental results ical extrusion texture with c-axis along the radial direction could
be seen for all the cases. However, the textures obtained from EBSD
3.1. Microstructural characterization measurements exhibited slight difference between the rod and the
tubes for both materials. While the measurements could be done
The initial grain size of the as cast billets for pure magnesium at the centre of the rod, which provides texture intensity that is
was ∼1000 ␮m, while for AM30 Mg alloy the grain size was mea- almost the same in all radial directions, the measurements could
sured to be ∼200 ␮m. Fig. 1 shows the microstructure after rod and only be conducted on the mid-wall-thickness of the tubes, which is
tube extrusion for both the alloys. It can be seen that after extrusion, almost the periphery of the tube radius. The latter situation led to
the microstructure of both pure magnesium and AM30 alloys con- the intensity of the (0 0 0 2) orientation being closer to the centre of
sisted of equiaxed grains along the axial direction of the extruded the pole figure, the ND (normal direction) of the pole figure, which
rod (Fig. 1a and b) and along the radial direction of the extruded comprises the circumferential plane perpendicular to the radial
tube (Fig. 1c and d). The average grain size for all the samples is pre- direction. The intensity was less in the direction perpendicular to
sented in Table 1. It can be seen that for the same amount of strain, both ED, hence ND (ND refers to the normal to the irradiated plane).
the grain size reduced more for the AM30 alloy than pure Mg. The However, on a qualitative scale all the materials exhibited the typ-
grain size distribution was almost unimodal in all the cases (Fig. 2). ical extrusion texture of magnesium, similar to the macro-texture
Misorientation angle distribution for all the extruded material obtained from the X-ray goniometer analysis.
was calculated from the EBSD data and is shown in Fig. 3. The Fig. 6a–d shows the ϕ2 = 0 ◦ and ϕ2 = 30 ◦ sections of the ori-
red curve shows the correlated misorientation angle distribution. entation distribution function (ODF) for the pure Mg and AM30
(For interpretation of the references to color in this sentence, the extruded rod and tubes. It can be seen that the fiber texture extends
reader is referred to the web version of the article.) In this case, from ϕ1 = 0–180 ◦ , ϕ = 90 ◦ , ϕ2 = 0–30 ◦ . In the case of tube, the tex-
the misorientation distribution was calculated using the misorien- ture was rotated so that the extrusion direction turns into along
tations between neighbouring points. The blue curve shows the ND and becomes similar to the extruded rod. This operation would
uncorrelated or the texture derived misorientation distribution, render the texture of both the rod and the tube in the same ref-
where the misorientations were calculated from every point pair erence system in the Euler orientation space. It was also observed
in the scan and not just the neighbouring points. The green curve that the maximum texture intensity for all cases is at ϕ2 = 30 ◦ .
3724 S. Biswas et al. / Materials Science and Engineering A 528 (2011) 3722–3729

Fig. 1. Inverse pole figure maps for (a and c) pure magnesium and (b and d) the alloy AM30: (a) extruded rod of pure magnesium, (b) extruded rod of AM30 alloy, (c) extruded
tube of pure magnesium and (d) extruded tube AM30 alloy.

Fig. 2. Grain size distribution for (a) extruded rod of pure magnesium, (b) extruded rod of AM30 alloy, (c) extruded tube of pure magnesium and (d) extruded tube AM30
alloy.
 S. Biswas et al. / Materials Science and Engineering A 528 (2011) 3722–3729 3725

Fig. 3. Misorientation angle distributions for pure magnesium and AM30 alloy after rod and tube extrusions.

Fig. 4. Global texture, measured by X-ray diffraction, represented as inverse pole figures for (a) extruded rod of pure magnesium, (b) extruded rod of AM30 alloy, (c) extruded
tube of pure magnesium and (d) extruded tube AM30 alloy. The projected direction is the extrusion direction (ED).

Table 1
Grain size and hardness for pure Mg and AM30 alloy.

Pure magnesium AM30

Grain size Hardness (MPa) Grain size Hardness (MPa)

Initial material >1000 ␮m 296 ± 18 >200 ␮m 457 ± 37

Extruded rod 18.6 ␮m 453 ± 35 10.4 ␮m 653 ± 63
Extruded tube 22.1 ␮m 363 ± 15 15.8 ␮m 603 ± 24
3726 S. Biswas et al. / Materials Science and Engineering A 528 (2011) 3722–3729

Fig. 5. (a) (0 0 0 2) pole figure of extruded pure magnesium, (b) (0 0 0 2) pole figure of extruded AM30 alloy, (c) simulated (0 0 0 2) pole figure for the rod extrusion of material
having (c/a = 1.624), (d) (0 0 0 2) pole figure of tube extruded pure magnesium, (e) (0 0 0 2) pole figure of tube extruded AM30 alloy, (f) simulated (0 0 0 2) pole figure tube
extrusion of material having (c/a = 1.624). All the experimental pole figures are in the same reference system as the corresponding simulated ones, where ‘1’ is extrusion
direction (ED) and 2, 3 are the radial directions.

Fig. 6. ϕ2 = 0 ◦ and ϕ2 = 30 ◦ sections of the ODFs for (a) extruded rod of pure magnesium, (b) extruded rod of AM30 alloy, (c) extruded tube of pure magnesium, (d) extruded
tube AM30 alloy and (e) simulated extrusion texture for the rod.
 S. Biswas et al. / Materials Science and Engineering A 528 (2011) 3722–3729 3727

pyr.c+a−I pyr.c+a−II
3.3. Hardness 0 /0basal , 0 /0basal ] was found out to be [1,10,12].
Deformation twinning was not considered for both materials in
The hardness values obtained from the micro-hardness tests are the simulations as limited twinning activity was observed in the
shown in Table 1 for the as received and extruded materials. The microstructures. The strain rate sensitivity, m was taken as 0.2,
finer grain size of AM30 in both the rods and tubes correlated with which is an approximate average value for pure Mg and Mg alloys
the significantly higher hardness compared with pure Mg. In all the based on published experimental data (Agnew and Duygulu [16],
cases, measurements were done on the RD plane. Kim et al. [24]). The m-value is considered to be the same as the
strain rate sensitivity exponent of slip in the constitutive law [23]:
4. Simulation    m−1
 ˙ s,f m f ˙
s,f  ˙ s,f 
 s,f
=
f
0 
s,f
sgn(˙ )  = 0   (4)
4.1. Simulation philosophy ˙ 0  ˙ 0  ˙ 0 

In this simulation, the deformation during both rod and tube Here  s,f is the resolved shear stress in the slip system indexed by
f
extrusion were assumed to be discontinuous at the entry point. s of the family indexed by f, ˙ s,f is the slip rate, the 0 value is the
In the case of rod extrusion, the following velocity gradient in the reference stress level (at which the slip rate is ˙ 0 ). The reference
plane of symmetry (1, 2, 3 reference system) was followed: shear rate ˙ 0 is supposed to be the same for all slip systems. Eq. (1)
  has been widely used in crystal plasticity simulations [for example,
1 0 0  23,25–27]. The slip systems are grouped into “families” for which

L =  0 −0.5

0  (1) purpose the index f is used. It is assumed that the reference shear
 0 0 −0.5  f
1,2,3 stress 0 is the same within a given slip system family, but can be
different from one family to another. Slip system hardening during
where 2 and 3 represent the radial directions and 1 corresponds to
deformation was not modelled.
the extrusion direction.
In the case of tube extrusion, the velocity gradient was expressed
4.3. Simulated texture
in terms of the tube reference system (which is incidentally similar
to rolling reference system). The velocity gradient in this case is
Simulations were conducted to parallel the experimental steps
represented as:
  used. The simulation results obtained are presented through pole
1 0 0  figures (Fig. 5) and ϕ2 = 0 ◦ and ϕ2 = 30 ◦ section ODFs (Fig. 6).
 
L =  0 −x 0  (2) Although, the extrusion textures are normally depicted through
 0 0 −y  inverse pole figures, as shown in Fig. 4 for comparison, the exper-
1,2,3
imental textures have also been re-plotted on the pole figures. It
where 1 represents the extrusion direction, 2 the tangential direc- can be seen that for the rod as well as the tube of both materi-
tion and 3 corresponds to the radial direction. Fig. 7 displays the als, the experimental results are well reproduced by simulation.
variations of x and y from the circumference towards the centre For both the cases, simulated textures were symmetric, however,
of the extruded billet. This expression for the velocity gradient is it can be observed from the experimental ODFs (Fig. 6a–d) of the
valid for rod extrusion as well; however, the reference system 1, rod and tube extrusion for both materials, that the fiber texture for
2, 3 in the pole figures (Fig. 5) will change. The benefit of adapt- all the cases extend from ϕ1 = 0 − 180 ◦ , ϕ = 90 ◦ and ϕ2 = 0 − 30 ◦ .
ing this reference system is that the heterogeneity of the texture It is well known that texture rotation due to DRX by ∼30◦ along
from the centre to the circumference of the tube could be taken into ϕ2 could not be noticed in the ϕ2 = 0 ◦ and ϕ2 = 30 ◦ section ODFs.
account. At the centre, the value of x and y is 0.5 as in Eq. (1). As we Fig. 6e shows the ϕ2 = 0 ◦ and ϕ2 = 30 ◦ section ODFs for the simu-
go towards the y direction (Fig. 7) from the centre to the surface lated normal extrusion (rod) condition, where it can be observed
(circumference), y increases from 0.5 to 1 and x decreases from 0.5 that the maximum intensity is located at ϕ2 = 0 ◦ of the fiber instead
to 0. In the same way as we go towards the x direction, x = 1 and of ϕ2 = 30 ◦ . The difference in experimental and simulated texture
y = 0 at the circumference. Therefore, the velocity gradient is same could be visualized from the position of the maximum intensity of
as that of rolling at the circumference of the extruded rod or tube. the fiber. As the simulated texture could consider the deformation
For simulating the tube texture at the mid-thickness, the veloc- only, the intensity at ϕ2 = 30 ◦ for the experimental texture could
ity gradient which matches almost exactly with the experimental be considered due to DRX.
texture was found to be as follows: As demonstrated in the simulation, the velocity gradient (Eq.
 
1 0 0  (2)) had a variable x and y components with respect to the position

L =  0 −0.4

0  (3) of the texture measurement from the centre to the circumference
 0 0 −0.6  along RD. In short, the texture changes from the centre to the cir-
1,2,3
cumference along the radial direction for extrusion. In the case
L was incorporated in polycrystalline self consistent model [22,23] of the tubes x = 0.4 and y = 0.6 was used to obtain the same tex-
up to a strain of 6.4 in the case of rod extrusion and 5.9 in the case ture. As the texture measurements were done at the mid-thickness
of tube extrusion. of the tube, ∼1.25 mm from the circumference, it was understood
that the change in texture is more prominent near the surface of
4.2. Simulation procedure the extruded rod or tube. This heterogeneity of texture could be
accounted for the change in the imposed velocity gradient along the
As the initial cast material had a weak texture, the initial texture RD near the circumference of the extruded rod or tube. However,
was assumed to be composed of 2000 randomly distributed spher- the possibility of the presence of friction at the surface could not
ical grains. The following families of slip systems were assumed to be ruled out. For better visualization of the resulting texture, a pic-
operate: basal {0 0 0 1}1 2̄ 1 0, prismatic {1 1̄ 0 0}1 1 2̄ 0, pyrami- torial representation showing orientation of the hexagonal closed
dal a  {1 0 1̄ 1} 1̄ 2 1̄ 0, pyramidal c + a  type-I {1 0 1̄ 1} 2 1̄ 1̄ 3̄ and packed (HCP) unit cells around the circumference of the extruded
pyramidal c + a  type-II {2 1̄ 1̄ 2} 2̄ 1 1 3. Based on previous studies rod and tube is also shown in Fig. 7.
on Mg and its alloys [22,23], the reference stress values correspond- To illustrate the hardening or softening induced by texture, the
prism. pyr.a
ing to the slip systems:[0basal /0basal , 0 /0basal , 0 /0basal , activity of the slip systems are plotted (Fig. 8). As can be seen, there
3728 S. Biswas et al. / Materials Science and Engineering A 528 (2011) 3722–3729

Fig. 7. Variations in velocity field from the centre to the circumference of the extruded rod and the velocity gradient used during simulation of tube extrusion and the
visualization of orientation around the circumference of the extrusion.

are significant changes; basal slip decreases from ∼99% to ∼50% and axis in (ϕ = 90 ◦ ) ODF, which is the rotation of ∼30◦ along the basal
then increases to ∼85%, while the pyramidal slip activity increases axis from the orientation of the deformed grains [30,31]. ϕ2 = 0 ◦ and
slightly and then again decreases at a von-Mises strain of ∼2. Pris- ϕ2 = 30 ◦ section ODFs (Fig. 6) show the texture evolution after the
matic slip activity remains less than 10%. Activities of all the slip rod and tube extrusion After the extrusion, the maximum intensity
systems remained almost the same for rest of the deformation. of the fiber formed due to deformation should be at ϕ2 = 0 ◦ if there
While an increase in activity of the harder slip systems (e.g. pyrami- is no recrystallization. This is discussed in detail in the simulation
dal slip) will induce apparent initial hardening [28], this gave way part of this paper (Section 4). It can be seen that the maximum
to softening due to basal glide that became predominant later. intensity is at ϕ2 = 30 ◦ supporting recrystallization.
In order to observe the strain incorporated in the material after
extrusion, Kernel Average Misorientation (KAM) was plotted for
5. Discussion
each case (Fig. 9). Average misorientation of each EBSD spot with
all of its neighbouring spots was calculated with the provision that
The texture measured for the rod on the plane perpendicular to
misorientations exceeding 5◦ were excluded from the average cal-
the extrusion direction (ED), and of the tube parallel to ED, is very
culation. The misorientation between a grain at the centre of the
similar to the extrusion texture measured by different researchers
kernel and all points at the perimeter of the kernel were measured.
[11,13,28]. Though extrusion of single phase Mg alloys leads to the
The local misorientation value assigned to the centre point was the
formation of c-type fiber, the presence of second phase drastically
average of these misorientations, which could be obtained from
alter the extrusion texture [28]. As an example, Mackenzie et al. [29]
the EBSD scans. Maps constructed using this method are helpful in
observed the same texture for AZ31 and ZC71 alloy, but showed
visualizing the distribution of local misorientation within a grain.
that WE43 alloy possessed texture with basal planes at ∼45◦ to
As the peak of the KAM curve shifts towards the left (origin) with
ED. As for the tube texture, the intensity of the (0 0 0 2) poles is
increasing number fraction, the material is expected to be more
more along the normal direction (ND) and matches well with that
recrystallized. As the maxima always occurred for a misorientation
observed for cold extruded AZ31 magnesium alloy sheet [30,31].
angle lower than 1◦ , the microstructures can be considered recrys-
The rod and tube extrusions were clearly dynamically recrystallized
tallized. The small difference in the recrystallization response of the
(DRX) since the extrusion temperature of 400 ◦ C was much higher
two materials could be realized based on this criterion; indeed the
than the recrystallization temperature of magnesium and its alloys
AM30 extruded rod had the lower level of recrystallization, while
(∼0.4TMelting ≈ 100◦ C). This was confirmed by the equiaxed grains
the pure Mg tube the most. The presence of Al and Mn as alloying
for both extruded materials. Based on previous studies, a unimodal
grain size distribution (Fig. 2) also indicates the occurrence of pri-
mary recrystallization [32]. On texture evolution during DRX, it is
well known that recrystallized grains rotate by ∼30◦ along the ϕ2

Fig. 9. Kernel Average Misorientation (KAM) for the extruded rod and tube of mag-
Fig. 8. Activities of various slip system during extrusion. nesium and AM30 alloy.
 S. Biswas et al. / Materials Science and Engineering A 528 (2011) 3722–3729 3729

elements is believed to be responsible for the lower DRX fraction 5. The average grain size for AM30 alloy was lower compared
observed in the AM30 alloy. with pure magnesium. No difference has been found in the
The presence of significant fraction of LAGBs and a negligible grain sizes in the extruded rods and tubes of the same
fraction of twin boundaries indicates that the deformation was material.
dominated by slip for both extruded alloys. Based on this observa-
tion, VPSC simulations (discussed a priori) were carried out without Acknowledgement
accounting for twinning to validate the deformation mechanism.
The hardness measurements in all the cases were carried out on The authors are grateful to Prof. P. Rama Rao for his encour-
RD plane. The hardness of the rod was higher than the tube, as the agement at various levels of their collaborative research. The work
strain induced in the rod was higher than in tube and consequently involved the Institute X-ray Facility and Institute Nanoscience
a relatively lower grain size could be obtained in rod compare to Initiative (INI) facility sponsored by DST-FIST program at I.I.Sc.,
the tube for both unalloyed and alloyed magnesium. On the other Bangalore. The authors are also thankful to Dr. C.N. Tome and
hand, though the overall deformation mechanism and texture for Prof. L.S. Tóth for providing the VPSC code for texture simula-
both materials in both rod and tube form were similar, the higher tion. Special thanks are due to Dr. Anil K. Sachdev of GMR&D
hardness noted for the AM30 alloy can be explained as due to solid Center for providing the materials used for this study as well as
solution strengthening with Al and Mn, and a significant reduction for a very critical reading of the manuscript and many valuable
in its grain size and sub-grain size compared with pure Mg. suggestions.
In short, the crystal plasticity simulations using visco-plastic
self consistent model indicated that during extrusion, the critical References
resolved shear stress for the prismatic and pyramidal a slip could
be ∼12 times and pyramidal c + a I and II could be ∼10 times that [1] K.U. Kainer, F. von Buch, in: K.U. Kainer (Ed.), Magnesium Alloys and Technolo-
gies, DGM, Weinheim, 2003, p. 1.
of basal slip for both the materials, pure Mg and the alloy AM30. [2] A.A Luo, J. Mater. Manuf. (2005) 411.
This indicates that alloying additions does not affect the deforma- [3] I.J. Polmear, Mater. Sci. Technol. 10 (1994) 1.
tion mechanism at these temperatures. The relative activity of the [4] F.H. Froes, D. Eliezer, E. Aghion, JOM 50 (1998) 30.
[5] W. Lee, Y.-L. Chen, H.-Y. Wang, C.-F. Yang, J.-W. Yeh, Mater. Sci. Eng. A 464
slip systems during the extrusion process is presented in Fig. 8. The
(2007) 76.
basal slip activity reduces and non-basal slip activity increases up [6] Y. Chen, Q. Wang, J. Peng, C. Zhai, W. Ding, J. Mater. Proc. Technol. 182 (2007)
to a strain of ∼0.5. Beyond this strain level, the activity of basal slip 281.
again increases accompanied with the decrease in non-basal activ- [7] T. Murai, S.-I. Matsuoka, S. Miyamoto, Y. Oki, J. Mater. Proc. Technol. 141 (2003)
207.
ities. This transition occurs within a strain of ∼2%. During the rest [8] S.R. Agnew, M.H. Yoo, C.N. Tomé, Acta Mater. 49 (2001) 4277.
of the deformation, the basal slip activity remains more than 80%, [9] A.D. Rollett, S.I. Wright, in: U.-F. Kocks, C.N. Tome, H.-R. Wenk (Eds.), Texture
whereas the non-basal slip activities are less than 10%. Very high and Anisotropy, Cambridge University Press, Cambridge, 1998, p. 179.
[10] S. Kleiner, P.J. Uggowitzer, Mater. Sci. Eng. A 379 (2004) 258.
of basal slip activity and comparatively low non-basal slip activity [11] S.J. Liang, Z.U. Liu, E.D. Wang, Mater. Lett. 62 (2008) 3051.
at the extrusion temperature (400 ◦ C) for both the alloys could be [12] F. Yang, S.M. Yin, S.X. Li, Z.F. Zhang, Mater. Sci. Eng. A (2008).
due to larger initial grain size. [13] T. Homma, N. Kunito, Kamado, Scr. Mater. 61 (2009) 644.
[14] J. Koike, T. Kobayashi, T. Mukai, H. Watanabe, M. Suzuki, K. Maruyama, K.
Higashi, Acta Mater. 51 (2003) 2055.
6. Conclusions [15] J.A. del Valle, F. Carren, O.A. Ruano, Acta Mater. 54 (2006) 4247.
[16] S.R. Agnew, Ö. Duygulu, Int. J. Plast. 21 (2005) 1161.
[17] S.-B. Yi, H.-G. Brokmeier, J. Bohle, D. Letzig, K.U. Kainer, Physica B 350 (2004)
Pure magnesium and a binary Mg–Al alloy AM30 were sub- 507.
jected to rod and tube extrusion in order to examine the evolution [18] S. Suwas, G. Gottstein, R. Kumar, Mater. Sci. Eng. A 471 (2007) 1.
[19] S. Biswas, S.S. Dhinwal, S. Suwas, in: A. Haldar, S. Suwas, D. Bhattacharjee
of microstructure and texture. The experimental results and the
(Eds.), Microstructure and Texture in Steels and Some Other Materials, Springer
crystal plasticity simulations led to the following conclusions: Verlag, London, 2009, p. 465.
[20] R.K. Nadella, I. Samajdar, G. Gottstein, in: K.U. Kainer (Ed.), Proceedings of
1. The texture of the extruded rod as well as tube consists of a fiber the Magnesium Alloys and their Applications, Wiley-VCH, Weinheim, 2003,
p. 1052.
texture that extends from ϕ1 = 0 − 180 ◦ , ϕ = 90 ◦ and ϕ2 = 0 − 30 ◦ [21] T.A. Samman, G. Gottstein, Mater. Sci. Eng. A 488 (2008) 406.
in the Euler space, forms in both the materials. The (0 0 0 2) poles [22] S.R. Agnew, P. Mehrotra, T.M. Lillo, G.M. Stoica, P.K. Liaw, Acta Mater. 53 (2005)
were perpendicular to the extrusion direction in all cases. 3135.
[23] B. Beausir, S. Suwas, L.S. Tóth, K.W. Neale, J.J. Fundenberger, Acta Mater. 56
2. Crystal plasticity simulations using polycrystalline visco-plastic (2007) 200.
self consistent model successfully has reproduced the experi- [24] W.J. Kim, S.W. Chung, C.S. Chung, D. Kum, Acta Mater. 49 (2001)
mental texture for both the rod and the tube. The heterogeneity 3337.
[25] R.J. Asaro, A. Needleman, Acta Metall. 33 (1985) 923.
of texture during extrusion is attributed to the change in the [26] B. Beausir, L.S. Tóth, K.W. Neale, Acta Mater. 55 (2007) 2695.
imposed velocity gradient with respect to the depth from the [27] B. Beausir, L.S. Tóth, F. Qods, K.W. Neale, J. Eng. Mater. Technol. Trans. ASME
circumference towards the radial direction. 131 (2009) 131.
[28] N. Stanford, M.R. Barnett, Scr. Mater. 58 (2008) 179.
3. The critical resolved shear stress for the prismatic and pyramidal [29] L.W.F. Mackenzie, B. Davis, F.J. Humphreys, G.W. Lorimer, Mater. Sci. Technol.
a slip is 12 times and pyramidal c + a I and II is 10 times that 23 (2007) 1173.
of basal slip, as estimated from the VPSC simulations. [30] S.J. Liang, Z.Y. Liu, E.D. Wang, Mater. Lett. 62 (2008) 4009.
[31] C.C. Huang, J.C. Huang, Y.K. Lin, Y.M. Hwang, Mater. Trans. 45 (2004)
4. Dynamic recrystallization is observed for the rod as well as the
3142.
tube. There is a shift along ϕ2 = 30 ◦ between the deformed and [32] F.J. Humphreys, M. Hatherly, Recrystallization and Related Phenomena, Perga-
the recrystallized grains. mon, Oxford, 1995.