R. [TÌPÁNEK et al.: MICROSTRUCTURAL CHANGES OF ECAP-PROCESSED MAGNESIUM ALLOY AZ91 ...
9–13
MICROSTRUCTURAL CHANGES OF ECAP-PROCESSED
MAGNESIUM ALLOY AZ91 DURING CYCLIC LOADING AT
DIFFERENT STRESS-AMPLITUDE LEVELS
MIKROSTRUKTURNE SPREMEMBE Z ECAP POSTOPKOM
PROCESIRANE MAGNEZIJEVE ZLITINE MED CIKLI^NIM
OBREMENJEVANJEM PRI RAZLI^NIH NIVOJIH NAPETOSTNE
AMPLITUDE
Roman [tìpánek1, Libor Pantìlejev1, Ondøej Man1, Mario Guagliano2, Maurizio Vedani2,
Ehsan Mostaed2
1Institute of Materials Science and Engineering, Faculty of Mechanical Engineering, Brno University of Technology,
Technická 2896/2, 616 69, Brno, Czech Republic
2Politecnico di Milano, Department of Mechanichal Engineering, Via Giuseppe La Masa 1, 20156, Milan, Italy
stepanek.r@fme.vutbr.cz
Prejem rokopisa – received: 2017-06-28; sprejem za objavo – accepted for publication: 2017-11-03
doi:10.17222/mit.2017.087
Microstructural changes of magnesium alloy AZ91 after fatigue loading in the EX-ECAP state were evaluated using EBSD. It
was found that both the number fraction of low-angle boundaries and parameter KAM decreased after the testing at a stress
amplitude of 160 MPa but started to increase with the increasing stress amplitude. This behaviour can be explained with a
mutual influence of dislocation accumulation (which is stronger with a higher stress amplitude) and dynamic softening (which is
weaker with a decreasing number of cycles/cycles to failure). The average grain size remained almost unchanged except at a
stress amplitude of 180 MPa, which could have been caused by certain conditions allowing an ideal development of both
mentioned phenomena.
Keywords: magnesium alloy, ECAP, EBSD, fatigue loading
V pri~ujo~em ~lanku avtorji opisujejo ovrednotenje mikrostrukturnih sprememb magnezijeve zlitine AZ91 z EBSD metodo
zaradi utrujanja po procesiranju z ECAP-postopkom. Ugotovili so, da se tako dele` malo kotnih mej kot tudi parameter razlik v
kristalografski orientaciji kristalnih zrn (KAM; angl.: Kernel Average Misorientation) zni`ujeta po testiranju pri napetostni
amplitudi toda pri~neta nara{~ati z nara{~ajo~o napetostno amplitudo. To obna{anje si je mo~ razlo`iti z isto~asnim vplivom
kopi~enja dislokacij, (ki je intenzivnej{e pri vi{jih napetostnih amplitudah) in dinami~nim meh~anjem, (ki je manj{e pri
manj{em razmerju med dejanskim {tevilom ciklov in {tevilom ciklov do poru{itve). Povpre~na velikost kristalnih zrn je ostala
skoraj nespremenjena pri napetostni amplitudi 180 MPa, kar je lahko posledica specifi~nih pogojev, ki omogo~ajo idealen potek
obeh omenjeni pojavov (mehanizmov).
Klju~ne besede: zlitina na osnovi magnezija, ekstremno mo~na plasti~na deformacija (SPD) dose`ena z enokanalnim
pravokotnim procesiranjem (ECAP), preiskovalna metoda na osnovi uklona povratno sipanih elektronov (EBSD), utrujanje oz.
dinami~no (cikli~no) obremenjevanje
1 INTRODUCTION
Magnesium and its alloys, as one of the lightest
structural material, is used in applications where weight
reduction is required. However, mechanical properties of
the magnesium alloys in the as-cast state are usually
unsatisfactory for any advanced application due to its
coarse grain structure.1,2 Although conventional extru-
sion is often sufficient to achieve suitable mechanical
properties, further grain-refinement methods are sought
to meet higher demands.3,4
Severe-plastic-deformation (SPD) methods seem to
be suitable for achieving ultrafine-grain (UFG) or nearly
UFG microstructures, which often exhibit superior
mechanical properties. Equal-channel angular pressing
(ECAP) is one of the most used SPD methods, especially
due to its good controllability of the final microstruc-
ture.4,5 The materials processed via ECAP (or generally
with any SPD method) exhibit a certain microstructural
instability caused by high inner energy and unstable
(grain) boundaries. The microstructural stability loss can
occur during a thermal exposition6,7 or mechanical load-
ing (predominantly fatigue loading).8,9 The instability
can be manifested as grain coarsening connected with a
decrease in the mechanical properties, which can be
lower than the properties of an unprocessed, virgin
material.
The response of fine and ultrafine-grained micro-
structures to cyclic loading is very ambiguous and
depends on various parameters. Even nearly the same
alloys can exhibit different responses depending predo-
minantly on the details of the treatment processes
although the chemical compositions and microstructures
of the materials are, in principle, identical.10 Never-
theless, although UFG materials do not exhibit a uniform
behaviour during cyclic loading, their responses are
Materiali in tehnologije / Materials and technology 52 (2018) 1, 9–13 9
UDK 66.017:669.721.5:629.7.018.4 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 52(1)9(2018)
usually connected with a certain degree of microstruc-
tural changes such as grain coarsening or a creation and
annihilation of low-angle boundaries connected with
softening/hardening processes.11–13
The response of the microstructure of magnesium
alloy AZ91 processed with ECAP to cyclic loading is
analysed in this paper.
2 EXPERIMENTAL PART
The experimental material was an extruded magne-
sium alloy AZ91 of commercial quality with a chemical
composition given in Table 1. The alloy was further
processed at Politecnico di Milano with ECAP at 200 °C
using route BC, with a die having an angle between the
intersecting channels of 110° (the EX-ECAP state). The
dimensions of the billets obtained with the ECAP pro-
cess were a 10-mm diameter and a length of approxi-
mately 90 mm. The initial microstructural analysis was
performed using a light microscope Zeiss Axio Observer
Z1m and a scanning electron microscope Zeiss Ultra
Plus with an Oxford Instruments NordlysNano EBSD
detector.
Fatigue testing was performed on cylindrical samples
with a diameter of 6 mm and gauge length of 20 mm
fabricated from the EX-ECAP billets. Rotating bending-
fatigue tests were conducted at room temperature in the
load-controlled regime with a symmetrical loading cycle
with a frequency of 2000 min–1 using an Italsigma
2TM831 testing machine. The changes in the microstruc-
ture were assessed by means of EBSD using a scanning
electron microscope Zeiss Ultra Plus. The EBSD anal-
ysis was performed using a square grid with a step length
of 0.1 μm. The accelerating voltage was 20 kV and the
beam current was 3.3–8.6 nA.
Three specific parameters – grain size (GS),
low-angle boundaries (LAB), number fraction and kernel
average misorientation (KAM) were selected to quantify
the microstructural changes induced with fatigue testing.
A grain was defined as an area separated from its
surrounding by a boundary with a misorientation of at
least 5° and consisting of at least 5 pixels. The grain
diameter was defined as the diameter of an equivalent
circle with the same area as that of a detected grain (an
equivalent grain diameter). The misorientation angle was
defined in a range of 1–100° (the estimated measurement
accuracy <1°); the threshold between low- and high-
angle boundary (HAB) was set as 15°. The KAM distri-
bution was calculated within an angular range of 0–3°
over the second neighbour perimeter including interlaid
pixels; the modal values were then taken for comparison.
For the evaluation of the structural changes due to
fatigue loading, three areas of approximately 10800 μm2
each were considered for each analysed sample.
3 RESULTS
The microstructure of the analysed alloy contains
large areas consisting of fine grains filled with very fine
Mg17Al12 phase particles, whose morphology is partly
related to the grain layout after the first pass of ECAP
(Figure 1a – the dark areas). According to the EBSD
results, the matrix of the material is nearly ultrafine
grained, but with a few residual large grains (Figure 1b).
Nevertheless, the calculated average grain size is 1.6 μm.
Based on the fatigue-test results (the fatigue limit
determined for 107 cycles is 176 MPa), samples loaded at
stress amplitudes of (160, 180, 200 and 220) MPa were
chosen for a microstructural analysis after the tests.
The initial values of the analysed parameters (average
grain size – 1.61 μm, LAB number fraction – 63.1 %,
R. [TÌPÁNEK et al.: MICROSTRUCTURAL CHANGES OF ECAP-PROCESSED MAGNESIUM ALLOY AZ91 ...
10 Materiali in tehnologije / Materials and technology 52 (2018) 1, 9–13
Table 1: Chemical composition of the used AZ91 alloy, in mass fractions, (w/%)
Al Cu Fe Mn Si Zn other (total) Mg
AZ91 extruded 8.700 0.001 0.003 0.200 0.040 0.670 <0.030 balance
Table 2: Changes in the observed micro- and substructural parameters after fatigue tests
initial state
GS (ìm) LAB (%) KAM mode (°)
1.61 63.1 0.25
absolute value ÄGS absolute value ÄLAB absolute value ÄKAM
sample 10
a = 160 MPa
N = 16651000
site 1 1.46 -0.15 45.70 -17.40 0.15 -0.10
site 2 1.70 0.09 53.40 -9.70 0.15 -0.10
site 3 1.67 0.06 53.10 -10.00 0.25 0.00
sample 11
a = 180 MPa
N = 7063000
site 1 2.28 0.67 68.80 5.70 0.35 0.10
site 2 2.03 0.42 69.70 6.60 0.35 0.10
site 3 1.80 0.19 70.30 7.20 0.25 0.00
sample 5
a = 200 MPa
Nf = 34000
site 1 1.64 0.03 70.60 7.50 0.35 0.10
site 2 1.34 -0.27 74.90 11.80 0.35 0.10
site 3 1.69 0.08 58.70 -4.40 0.25 0.00
sample 4
a = 220 MPa
Nf = 24000
site 1 1.78 0.17 83.20 20.10 0.45 0.20
site 2 1.79 0.18 76.40 13.30 0.35 0.10
site 3 1.95 0.34 81.30 18.20 0.35 0.10
KAM mode – 0.25°, Table 2) were calculated as the
average values of more than 25 areas on the same alloy
in the virgin state (EX-ECAP). According to the EBSD
analysis, the average grain size remained almost
unchanged after the loading at 160 MPa and 200 MPa,
while it increased at 220 MPa and was even higher at
180 MPa. Gained inverse-pole figure maps (Figure 2)
clearly show that at the stress amplitude of 220 MPa, the
grain coarsening is rather homogeneous in the whole
investigated area, while at the stress amplitude of 180
MPa, the coarsening is more anisotropic with the grains
elongated in a certain direction. Even at 180 MPa, when
the grain coarsening is the most pronounced, the
absolute change in the grain diameter is lower than 1 μm
(Table 2).
The number fraction of LAB decreased at 160 MPa
but started to increase with the increasing stress ampli-
tude (Table 2). The KAM mode (which can be
connected with the local lattice distortion of the material)
exhibited a trend similar to the changes of the LAB
number fraction – it decreases at lower stress amplitudes
and increases at higher stress amplitudes (Table 2).
R. [TÌPÁNEK et al.: MICROSTRUCTURAL CHANGES OF ECAP-PROCESSED MAGNESIUM ALLOY AZ91 ...
Materiali in tehnologije / Materials and technology 52 (2018) 1, 9–13 11
Figure 2: IPF maps of AZ91 alloy after the fatigue test: a) a = 160 MPa, b) a = 180 MPa, c) a = 200 MPa, d) a = 220 MPa
Figure 1: Microstructure of AZ91 alloy (EX-ECAP state): a) light microscope, b) IPF map (EBSD – SEM)
4 DISCUSSION
The clear trend in the LAB and KAM change beha-
viour can be explained with two significant processes
that occur in the materials (especially in the fine- and
ultrafine-grained ones) during the cyclic loading. The
first process is a dislocation accumulation with an in-
creasing overall dislocation density, which can lead up to
an increased number of LAB and also an increase in the
local lattice distortion and KAM mode.14 The second
process, occurring during the cyclic loading, is dynamic
softening, driven by the cyclic loading of the material.14
Due to their nature, both processes are mutually counter-
active. At lower levels of the stress amplitude, the effect
of the dislocation accumulation is weaker, while dyna-
mic-softening processes are more pronounced due to a
higher number of the loading cycles so that the overall
effect of the cyclic loading on the (sub) structure would
be manifested by a decrease in both the LAB number
fraction and KAM mode. With the increasing stress
amplitude, dislocation accumulation becomes stronger
and the number of cycles to failure decreases so that the
dynamic-softening processes cannot be developed as in
the previous case; therefore, the overall change in the
LAB number fraction and KAM mode start increasing
and the increase is more pronounced with the higher
stress amplitude (and lower number of cycles to failure).
This common assumption is supported by the results of
an image-quality (IQ) analysis (the value of IQ can be
connected to the overall dislocation density at the
qualitative level – a higher dislocation density results in a
lower diffraction and lower IQ), which indicate a de-
crease of IQ with the increasing stress amplitude (stress
amplitude/average IQ: 160 MPa/105, 180 MPa/68, 200
MPa/65, 220 MPa/57).
The change in the grain size did not exhibit any parti-
cular trend. This behaviour can also be explained with
dynamic-softening processes, which can be accompanied
by the grain coarsening under certain conditions.14,15 The
changes in the grain size could be caused by a proper
combination of a sufficiently high amplitude and an
adequate loading time (i.e., a higher number of cycles). It
is possible that a slight increase in the grain size at 220
MPa was caused simply by a high stress amplitude,
which allowed the grain coarsening to partially develop
even during 24.000 cycles. In contrast with this assump-
tion, the changes at 180 MPa could be attributed to more
than 7×106 cycles, which allowed the grain coarsening to
develop at the tested amplitude (which was very close to
the estimated fatigue limit). At 160 MPa and 200 MPa,
the conditions for the development of grain coarsening
were probably insufficient so that the grain size re-
mained almost unchanged.
5 CONCLUSIONS
Changes of the (sub)structural parameter of fine-
grained Mg alloy AZ91 processed with ECAP were
analysed:
• At lower stress amplitudes, the local lattice distortion
and the amount of LAB decreased because of the
influence of dynamic-softening processes that coun-
teracted the dislocation accumulation.
• With the increasing stress amplitude (and decreasing
number of cycles/cycles to failure), the influence of
the dislocation accumulation increased and the
softening processes did not fully develop so that both
the LAB number fraction and local lattice distortion
gradually increased.
• The change in the average grain size did not exhibit
any particular trend so that during the cyclic loading
below the fatigue limit, the microstructure could be
considered as stable.
Acknowledgement
This work was supported by the specific research
project no. FSI-S-14-2511 and the European Regional
Development Fund in the framework of research project
NETME Centre under the Operational Program Research
and Development for Innovation No. CZ.1.5/2.1.00/
01.002, and within project NETME plus (Lo1202), a
project of the Ministry of Education, Youth and Sports
under the "National sustainability programme".
6 REFERENCES
1 B. L. Mordike, T. Ebert, Magnesium: Properties – applications –
potential, Materials Science and Engineering A, 302 (2001), 37–45,
doi:10.1016/S0921-5093(00)01351-4
2 P. Poddar, S. Bagui, K. Ashok, A. P. Murugesan, Experimental inves-
tigation on microstructure and mechanical properties of gravity-
die-cast magnesium alloys, Journal of alloys and compounds, 695
(2017), 895–908, doi:10.1016/j.jallcom.2016.10.193
3 H. M. Ruzi, M. Norhamidi, S. Abu Bakar, R. J. Khairur, M. N. Nor
Hafiez, A. Sufizar, I. I. M. Halim, A review of workability of
wrought magnesium alloys, Advanced Manufacturing Research
Group ’09 Seminar 3, 2009
4 K. Kubota, M. Mabuchi, K. Higashi, Review: Processing and mecha-
nical properties of fine-grained magnesium alloys, Journal of
Materials Science, 34 (1999), 2255–2262, doi:10.1023/
A:1004561205627
5 R. B. Figueiredo, T. G. Langdon, Grain refinement and mechanical
behavior of a magnesium alloy processed by ECAP, J. Mater. Sci., 45
(2010), 4827–4836, doi:10.1007/s10853-010-4589-y
6 J. Stráská, M. Jane~ek, J. ^í`ek, J. Stráský, B. Hadzima, Micro-
structure stability of ultra-fine grained magnesium alloy AZ31
processed by extrusion and equal-channel angular pressing
(EX–ECAP), Materials Characterization, 94 (2014), 69–79,
doi:10.1016/j.matchar.2014.05.013
7 H. K. Kim, W. J. Kim, Microstructural instability and strength of an
AZ31 Mg alloy after severe plastic deformation, Materials Science
and Engineering A, 385 (2004), 300–308, doi:10.1016/j.msea.2004.
06.055
R. [TÌPÁNEK et al.: MICROSTRUCTURAL CHANGES OF ECAP-PROCESSED MAGNESIUM ALLOY AZ91 ...
12 Materiali in tehnologije / Materials and technology 52 (2018) 1, 9–13
8 H. W. Höppel, M. Kautz, C. Xu, M. Murashkin, T. G. Langdon, R. Z.
Valiev, H. Mughrabi, An overview: Fatigue behaviour of ultrafine-
grained metals and alloys, International Journal of Fatigue, 28
(2006), 1001–1010, doi:10.1016/j.ijfatigue.2005.08.014
9 S. Fintová, L. Kunz, Fatigue properties of magnesium alloy AZ91
processed by severe plastic deformation, Journal of the Mechanical
Behavior of Biomedical Materials, 42 (2015), 219–222,
doi:10.1016/j.jmbbm.2014.11.019
10 L. Wu, G. M. Stoica, H. Liao, S. R. Agnew, E. A. Payzant, G. Wang,
D. E. Fielden, L. Chen, P. K. Liaw, Fatigue-property enhancement of
magnesium alloy AZ31B through equal-channel-angular pressing,
Metallurgical and Materials Transactions A, 38 (2007), 2283–2289,
doi:10.1007/s1161-007-9123-8
11 L. Kunz, P. Luká{, L. Pantìlejev, O. Man, Stability of microstructure
of ultrafine-grained copper under fatigue and thermal exposition,
Strain, 47 (2011), 476–482, doi:10.1111/j.1475-1305.2009.00710.x
12 R. Zhu, Y. J. Wu, W. Q. Ji, J. T. Wang, Cyclic softening of ultrafine-
grained AZ31 magnesium alloy processed by equal-channel angular
pressing, Materials Letters, 65 (2011), 3593–3596, doi:10.1016/
j.matlet.2011.07.111
13 Z. J. Zhang, P. Zhang, Z. F. Zhang, Cyclic softening behaviours of
ultra-fine grained Cu-Zn alloys, Acta Materialia, 121 (2016),
331–342, doi:10.1016/j.actamat.2016.09.020
14 H. W. Höppel, H. Mughrabi, Cyclic deformation and fatigue pro-
perties of very fine-grained metals and alloys, International Journal
of Fatigue, 32 (2010), 1413–1427, doi:10.1016/j.ifatigue.2009.
10.007
15 T. Sakai, A. Belyakov, R. Kaibyshev, H. Miura, J. J. Jonas, Dynamic
and post-dynamic recrystallization under hot, cold and severe plastic
deformation conditions, Progress in Materials Science, 60 (2014),
130–207, doi:10.1016/j.pmatsci.2013.09.002
R. [TÌPÁNEK et al.: MICROSTRUCTURAL CHANGES OF ECAP-PROCESSED MAGNESIUM ALLOY AZ91 ...
Materiali in tehnologije / Materials and technology 52 (2018) 1, 9–13 13