Low cycle fatigue of an extruded Mg–3Nd–0.2Zn–0.5Zr magnesium alloy

Materials and Design 64 (2014) 63–73

Contents lists available at ScienceDirect

Materials and Design journal homepage: www.elsevier.com/locate/matdes

Low cycle fatigue of an extruded Mg–3Nd–0.2Zn–0.5Zr magnesium alloy F.A. Mirza a, D.L. Chen a,⇑, D.J. Li b, X.Q. Zeng b,⇑ a b

Department of Mechanical and Industrial Engineering, Ryerson University, 350 Victoria Street, Toronto, Ontario M5B 2K3, Canada The State Key Laboratory of Metal Matrix Composites, School of Materials Science and Engineering, Shanghai Jiao Tong University, 800 Dongchuan Road, Shanghai 200240, PR China

a r t i c l e

i n f o

Article history: Received 31 March 2014 Accepted 9 July 2014 Available online 23 July 2014 Keywords: Magnesium alloy Low cycle fatigue Twinning–detwinning Precipitate free zone Tension–compression yield asymmetry

a b s t r a c t The purpose of this study was to evaluate strain-controlled cyclic deformation behavior of an extruded Mg–3Nd–0.2Zn–0.5Zr (NZ30K) magnesium alloy. The microstructure of this alloy consisted of a bimodal microstructure with equiaxed recrystallized grains and unrecrystallized coarse grains along with a large number of smaller second-phase particles present inside the grains and larger particles along the grain boundaries alongside a characteristic precipitate free zone (PFZ). The average grain size was about approximately 5–7 lm. It was observed that unlike the higher RE-containing Mg–10Gd–3Y–0.5Zr (GW103K) magnesium alloy, the NZ30K alloy exhibited asymmetrical hysteresis loops in tension and compression in the fully reversed strain-control tests at a strain ratio of Re = 1. This was mainly due to the presence of relatively stronger crystallographic texture, PFZ, and the resultant twinning–detwinning activities during cyclic deformation. While this alloy exhibited cyclic softening at lower strain amplitudes and cyclic hardening at higher strain amplitudes, it had an equivalent fatigue life to that of other extruded Mg alloys. Fatigue crack was observed to initiate from the specimen surface with some isolated facets of the cleavage-like planes near the initiation site. Crack propagation was basically characterized by serrated fatigue striations. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Lightweighting in ground vehicles is deemed as one of the most effective strategies to improve fuel economy and reduce anthropogenic climate-changing, environment-damaging, costly and human death-causing1 emissions [1–7] due to the tremendous environmental concerns and escalating global energy demand combated by the transportation industry in recent years. Indeed, it is now depicted as the storm of lightweighting – a revolution in materials, processes, and business models – which is brewing on the horizon of the auto industry. To manufacture lighterweight vehicles, advanced highstrength steels, aluminum alloys, magnesium (Mg) alloys, and polymers are being used in the aerospace and automotive sectors, ⇑ Corresponding authors. Tel.: +416 979 5000x6487; fax: +416 979 5265 (D.L. Chen), tel.: +021 5474 2911; fax: +021 3420 2794 (X.Q. Zeng). E-mail addresses: [email protected] (D.L. Chen), [email protected] (X.Q. Zeng). 1 According to Science News entitled ‘‘Air pollution kills 7 million people a year’’ on March 25, 2014 at http://news.sciencemag.org/signal-noise/2014/03/air-pollutionkills-7-million-people-year: ‘‘Air pollution is not just harming Earth; it is hurting us, too. Startling new numbers released by the World Health Organization today reveal that one in eight deaths are a result of exposure to air pollution. The data reveal a strong link between the tiny particles that we breathe into our lungs and the illnesses they can lead to, including stroke, heart attack, lung cancer, and chronic obstructive pulmonary disease.’’ http://dx.doi.org/10.1016/j.matdes.2014.07.016 0261-3069/Ó 2014 Elsevier Ltd. All rights reserved.

but substantial weight reductions could be further achieved by employing ultra-lightweight Mg alloys due to their low density, high strength-to-weight ratio, and superior damping capacity [1,4,8,9]. There are currently intensive studies in the development of wrought Mg alloys with high strength, high corrosion resistance, and superior formability for structural applications [10,11]. Despite the potential of substantial reductions in weight, most wrought Mg alloys exhibited unusual mechanical properties, e.g., tension–compression yield asymmetry, limited ductility, and pronounced directional anisotropy arising from the presence of strong basal texture related to their hexagonal close-packed (HCP) crystal structure with a limited number of slip systems activated during extrusion or rolling processes [12–22]. Indeed, for the vehicle components subjected to dynamic cyclic loading, such mechanical anisotropy and tension–compression yield asymmetry could lead to irreversibility of cyclic deformation which may have an unfavorable influence on the material performance [20,23–26]. These problems could be conquered by means of texture modification via an addition of rare earth (RE) elements [20,23–25,27]. The addition of RE elements in Mg alloys can result in a decrease of the overall texture sharpness or intensity, allowing the activation of easy basal slip to a higher extent, compared to the RE-free wrought Mg alloys [23–25,28]. For example, Stanford and Barnett [28] reported that the anisotropy of yield strength of

64

F.A. Mirza et al. / Materials and Design 64 (2014) 63–73

(a)

(b)

(c)

(d)

PFZ PFZ

PFZ PFZ

Fig. 1. Microstructures of extruded NZ30K alloy, (a) and (c) OM images, and (b) and (d) SEM back-scattered electron images taken at different magnifications.

a rolled Mg–Zn–RE alloy could be reversed as compared with the traditional Mg–Zn alloy; the tensile-compressive yield asymmetry of a Mg–10Gd–3Y–0.5Zr (GW103K) alloy in different conditions was observed to be nearly absent via the weakening of crystallographic texture due to the addition of RE elements [23–25]. Furthermore, it was reported that RE elements could develop more random textures during hot extrusion and/or generate the socalled ‘‘RE texture component’’ even at low alloying levels, which could result in a substantial gain of ductility and reduce the extent of tension–compression asymmetry in the conventional wrought Mg alloys [27]. Despite the fact that the addition of RE elements sheds some light on the alteration in the mechanical anisotropy, the potential advantage of such RE–Mg alloys as structural components under cyclic loading condition has not been well appreciated. The earlier studies were conducted mainly on the high cyclic fatigue properties of such alloys, e.g., [20,29–36]. Recently, there were several investigations which reported the strain-controlled low cycle fatigue behavior of RE-Mg alloys [23–25,37–42]. For instance, the cyclic deformation and low cycle fatigue behavior of an extruded Mg–8.0Gd–3.0Y–0.5Zr (GW83) alloy under fully reversed straincontrolled tension–compression loading along the extrusion

direction was investigated [39], where the GW83 alloy displayed nearly symmetrical stress–strain hysteresis loops (but a detailed examination indicated that they were not fully symmetrical when the strain amplitude was large) and marginal cyclic hardening with almost zero mean stress, and the strain–life curve was practically identical to that of RE-free AZ31 alloy [43]. In our earlier studies [23–25], low cycle fatigue behavior of a higher RE-containing extruded GW103K alloy was evaluated at varying strain amplitudes and strain ratios. It was observed that unlike the RE-free Mg alloys (e.g., extruded AZ31 and AM30 alloys [12–16], and extruded ZK60, AZ31B, and AZ61A alloys [44–48]), the RE-containing extruded GW103K alloy exhibited basically symmetrical hysteresis loops during cyclic deformation under fully reversed cycling, indicating the absence of unfavorable or unwanted tension–compression yield asymmetry due to the weaker texture, smaller grain size, and stronger resistance of RE particles or precip 2g extension twinning. While the high perforitates to the f1 0 1 mance of the GW103K alloy was achieved, it was somewhat expensive due to the addition of a fairly high amount (totally 13 wt.%) of RE elements. In the automotive applications cost is one of the major considerations. This has driven the development of relatively low RE-containing Mg alloys, e.g., recently developed

8

Intensity, Counts x100

(a)

Mg

(b)

6

4

2

Nd

Nd

Nd

0 0

5

10

15

20

25

30

Distance, μm Fig. 2. SEM back-scattered electron images indicating (a) EDS line scan positions and (b) the corresponding EDS line scan results of NZ30K alloy.

F.A. Mirza et al. / Materials and Design 64 (2014) 63–73

600

Intensity, counts

Mg–3Nd–0.2Zn–0.5Zr (NZ30K) alloy. This NZ30K alloy offers high strength and ductility as well as good creep resistance, it is thus very attractive to automotive powertrain and structural applications [29]. However, it is unclear to what extent the NZ30K Mg alloy would exhibit the tension–compression asymmetry, whether cyclic hardening would occur, and what are the cyclic stress response and fatigue life. The present study was, therefore, aimed at identifying cyclic deformation behavior of the low RE containing extruded NZ30K alloy, and evaluating the fatigue life under varying strain amplitudes.

Mg Mg12Nd Mg3Nd

500

65

400 300 200 100

2. Material and experimental procedure ▲

0 10

20

30

40

50

60

70

The material investigated was a recently developed extruded Mg alloy with the following nominal composition (wt.%): 3Nd, 0.2Zn, 0.5Zr, and Mg (balance). The received alloy bars with a diameter of 20 mm were extruded at 400 °C with an extrusion ratio of 9:1 at an extrusion ram speed of 3 mm/s and a die angle of 60°.

80

Diffraction angle 2θ, ° Fig. 3. X-ray diffraction pattern of the extruded NZ30K alloy.

(a)

Basal plane dislocation slip line

(b)

A

Fig. 4. Transmission electron micrographs recorded from a sample of as-extruded NZ30K alloy: (a) bright-field image, and (b) the corresponding SAED patterns, where the  0iMg. beam line was parallel to h1 1 2

ED RD Basal plane (0001)

Prismatic plane (10 1 0)

Pyramidal plane (10 1 1)

 0Þ plane, and pyramidal ð1 0 1  1Þ plane of the extruded NZ30K alloy, where ED stands for the extrusion direction Fig. 5. Pole figures of basal (0 0 0 1) plane, prismatic ð1 0 1 and RD indicates the radial direction.

F.A. Mirza et al. / Materials and Design 64 (2014) 63–73

Engineering stress, MPa

300 250 200 150 100

NZ30K GW103K

0 0

2

4

6

8

(a)

First cycle

200 100 0 -100

AM30

-200

NZ30K -300 -1.5

-1

-0.5

0

0.5

1

1.5

Strain, % 300

(b)

Mid-life cycle

200 100 0 -100

AM30

-200

NZ30K -300 -1.5

-1

-0.5

0

0.5

1

1.5

Strain, % Fig. 7. Typical stress–strain hysteresis loops of (a) first cycle and (b) mid-life cycle at a total strain amplitude of 1.2% and strain ratio of Re = 1 for the extruded AM30 and NZ30K alloys, respectively.

triangular waveform results in a constant strain rate in the course of cycling. Low cycle fatigue tests were performed at total strain amplitudes of 0.2%, 0.4%, 0.6%, 0.8%, 1.0%, and 1.2%, and at least two samples were tested at each level of the strain amplitudes. The strain-controlled tests at lower strain amplitude levels were continued up to 10,000 cycles, then the tests were changed to load control at a frequency of 50 Hz using sine waveform. For the sake of comparison, some samples of extruded RE-free AM30 Mg alloy (with a composition of 3.4 wt.% Al, 0.33 wt.% Mn, 0.16 wt.% Zn, 0.0026 wt.% Fe, 0.0006 wt.% Ni, 0.0008 wt.% Cu, and balance Mg) and RE-containing GW103K alloy (with a composition of 10 wt.% Gd, 3 wt.% Y, 0.5 wt.% Zr, and balance Mg) were also tested. The fracture surfaces of fatigued specimens were examined via scanning electron microscope (SEM) to identify fatigue crack initiation sites and propagation characteristics. The residual twins in the regions near the fracture surface were observed as well.

350

50

300

Stress, MPa

Microstructural examinations were performed using an optical microscope (OM), and scanning electron microscope (SEM) JSM6380LV equipped with Oxford energy dispersive X-ray spectroscopy (EDS) system, and transmission electron microscope (TEM, JEM-2100) operated at 200 kV. Standard metallographic preparation procedures were used to grind and polish sample surfaces, and etching was done with Vilella’s reagent containing 1 g picric acid, 5 ml HCl, and 100 ml ethanol. The texture was determined by measuring incomplete pole figures between W = 0–75° in the back reflection mode using a PANalytical X-ray diffractometer (XRD) with Cu Ka radiation at 45 kV and 40 mA and analyzed using MTEX software [49], and defocusing arising from the rotation of XRD sample holder was corrected using experimentally determined data obtained from the diffraction of magnesium powders received from Magnesium Elektron. Sub-sized tensile and fatigue samples in accordance with ASTM: E8 standard were machined with the loading axis parallel to the extrusion direction (ED). The samples had a gauge length of 25 mm (or a parallel length of 32 mm) and a width of 6 mm. The thickness of the samples was 6 mm as well. The gauge section of fatigue samples was ground progressively along the loading direction with emery papers up to a grit number of 600 to avoid the influence of machining marks on the tensile and fatigue results. The tensile tests were carried out in accordance with the ASTM: E8 standard using a computerized United tensile testing machine at a strain rate of 1  104 s1 for the extruded NZ30K Mg alloy at room temperature. The strain rate was controlled by an extensometer with a 25 mm gauge length. Strain-controlled, pull-push type fatigue tests in accordance with the ASTM: E606 standard were conducted in air at room temperature with a 25 mm extensometer using a computerized Instron 8801 fatigue testing system that was controlled by a Fast Track Low Cycle Fatigue (LCF) program at a zero mean strain (i.e., a strain ratio of Re = 1, completely reversed strain cycle) at a constant strain rate of 1  102 s1. Triangular loading waveform was applied in the cyclic deformation tests, as noted in the ASTM: E606 standard for continuous cyclic tests and generally for strain-rate sensitive materials, since the

Stress, MPa

66

10

Engineering strain, %

3. Results

Fig. 6. Typical tensile stress–strain curves of extruded NZ30K and GW103K alloys tested at a strain rate of 1  104 s1.

Table 1 Tensile properties of the extruded NZ30K and GW103K alloys obtained at a strain rate of 1  104 s1. Materials

rYS (MPa)

rUTS (MPa)

Elongation (%)

n

NZ30K GW103K

180 232

225 318

9.1 8.3

0.12 0.15

3.1. Microstructure and crystallographic texture Fig. 1 shows typical optical micrographs and SEM back-scattered electron images of NZ30K alloy in the as-extruded condition taken at different magnifications. The NZ30K alloy showed a bimodal microstructure due to the occurrence of dynamic recrystallization (DRX) in the hot extrusion process at 400 °C, which consisted of equiaxed recrystallized grains and unrecrystallized coarse grains. A significant number of larger intermetallic compounds

67

F.A. Mirza et al. / Materials and Design 64 (2014) 63–73

250

Stress amplitude, MPa

Total strain amplitudes

1.2% 1.0% 0.8% 0.6% 0.4% 0.2%

200

150

100

50 1E+0

reported in Al–Zn–Mg alloys [54] and Al–4Cu alloys [55]. The basic difference was that the PFZs in the present alloy could be seen at a microscopic level under optical microscope, whereas the previous literature showed PFZs at a nanoscopic level using transmission electron microscopy (TEM). To better see the intermetallic compounds in the as-extruded alloy, EDS line scan was performed as shown in Fig. 2(a) and (b), where it showed Nd-containing particles. X-ray diffraction pattern obtained from the as-extruded NZ30K alloy as shown in Fig. 3 confirmed the presence of two types of Nd-containing particles. Besides the a-Mg peaks, metastable Mg12Nd peaks (which was simulated by the suppression of

1.4 1E+1

1E+2

1E+3

1E+4

1.2

Number of cycles, N Fig. 8. Stress amplitude vs. the number of cycles of the extruded NZ30K alloy tested at different total strain amplitudes at a strain ratio of Re = 1.

1.2% 1.0% 0.8% 0.6% 0.4% 0.2%

Plastic strain amplitude, %

Total strain amplitudes

0.6

0.4

0.2

Δ ε t /2, %

1.0 0.8 0.6 0.4 0.2 0.0 1E+1

1E+3

1E+5

1E+7

Number of cycles to failure, Nf Fig. 10. Total strain amplitude vs. the number of cycles to failure for the extruded NZ30K alloy, in comparison with the data reported in the literature for various extruded Mg alloys [13,14,23,46,70–72].

Table 2 Low cycle fatigue parameters obtained for the as-extruded NZ30K alloy. Low cycle fatigue parameters

Extruded NZ30K

Cyclic yield strength, r0 y, MPa Cyclic strain hardening exponent, n0 Cyclic strength coefficient, K0 , MPa Fatigue strength coefficient, r0 f, MPa Fatigue strength exponent, b Fatigue ductility coefficient, e0 f Fatigue ductility exponent, c

165 0.10 313 431 0.10 0.24 0.72

250

Δ σ /2 at mid-life, MPa

were mainly present along the grain boundaries in the form of dashed lines, and a large portion of these coarse intermetallic phases were broken into small particles and moved from grain boundaries to grain interiors during the hot extrusion process, as indicated by arrows in Fig. 1(a). Similar types of microstructures were also reported in extruded Mg–2.4Nd–0.6Zn–0.6Zr alloy by Wen et al. [50] and extruded Mg–3Nd–0.2Zn–Zr alloy by Ma et al. [51]. The average grain size was approximately 5–7 lm which was fairly small in comparison with the common extruded Mg alloys, such as AZ31 and AM30 [12–16]. This was due to the role of added RE elements and zirconium (Zr) where Zr mainly restricted the grain growth [17]. A similar role of grain refinement by other RE elements, e.g., gadolinium [18], yttrium [19], cerium [52], neodymium [53], has also been reported. Besides, as shown in Fig. 1(a) no twins were present in the un-deformed samples. Fig. 1(b) shows a typical SEM back-scattered electron images of the as-extruded sample where a lot of larger particles along the grain boundaries in the form of dashed lines and smaller particles within the grains could be better seen. Furthermore, it can be seen more clearly from Fig. 1(c) and (d) taken at a higher magnification that there were some regions in the vicinity of grain boundaries which were free of precipitates – a phenomenon referred to as the ‘‘precipitate free zone’’ (PFZ). Similar types of PFZs were also

0.8

AM30 [13] AZ31 [14] GW103K [23] ZK60 [46] AZ31 [70] AZ61-ED [71] AZ61-ETD [71] AM30 [72] NZ30K-present study

1E+5

200

150

100

50

Monotonic stress-strain curve Cyclic stress-strain curve

0 1E+0

0 1E+1

1E+2

1E+3

1E+4

1E+5

Number of cycles, N Fig. 9. Plastic strain amplitude vs. the number of cycles of the extruded NZ30K alloy tested at different total strain amplitudes at a strain ratio of Re = 1.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

Δ ε t /2 at mid-life, % Fig. 11. Cyclic stress–strain curves for the extruded NZ30K alloy, where the corresponding monotonic stress–strain curve is plotted for comparison.

68

F.A. Mirza et al. / Materials and Design 64 (2014) 63–73

Mg41Nd5 peaks as reported by Zhang et al. [56]) and Mg3Nd peaks were detected in the XRD pattern. A similar type of second phase (Mg12Nd) was also reported in other Mg–Nd alloys, e.g., in a gravity cast Mg–yNd–xZn–Zr alloy by Fu et al. [57], cast Mg–3Nd–0.2Zn–04Zr alloy by Fu et al. [58], and an extruded Mg–Nd–Zn–Zr alloy by Yu et al. [59]. Furthermore, fine microstructures of as-extruded NZ30K alloy were also identified by three zone-axis electron microdiffraction patterns via TEM, as shown in Fig. 4. Electron microdiffraction pattern in Fig. 4(b) shows the structure of the particle (A) (Fig. 4(a)) which was identified as cuboid-shaped (RE-rich) precipitates/particles (face-centered cubic (fcc) with a = 0.56 nm). In addition, basal plane dislocation slip lines were also observed  0i direction as indicated by an arrow in Fig. 4(a), along h1 1 2 Mg which were formed during extrusion. Fig. 5 shows the initial crystallographic textures (pole figures of  0Þ plane, and pyramidal basal (0 0 0 1) plane, prismatic ð1 0 1  1Þ plane) of the extruded NZ30K alloy evaluated using MTEX ð1 0 1 software where ED stands for the extrusion direction and RD indicates the radial direction. A relatively strong concentration (with a maximum intensity of 6.9 multiples of random distribution (MRD)) of basal (0 0 0 1) pole mainly towards the RD, along with prismatic  0Þ and pyramidal ð1 0 1  1Þ poles towards the ED, was ð1 0 1 observed. This indicates that the c-axes of most grains were aligned perpendicular to the ED. While one benefit of the RE elements added into Mg alloys was to weaken the texture, e.g., Stanford and Barnett [28] reported that microalloying with RE elements could weaken texture in the forming process, it seemed that the texture in the extruded NZ30K alloy was not sufficiently weakened, in comparison with the reported results in other RE-Mg alloys [20,23–25,27,60,61]. This could be related to the presence of PFZs (Fig. 1) and grainy Mg12Nd precipitates distributed in the intra-crystalline and along the grain boundaries as reported by Zheng et al. [62]. A similar higher texture intensity of NZ30K alloy has also been reported by Ma et al. [51].

(a)

3.2. Tensile properties Fig. 6 shows the typical stress–strain curves obtained for the extruded NZ30K and GW103K alloys at a strain rate of 1  104 s1 at room temperature and the tensile properties obtained are listed in Table 1. As seen from Table 1, the present NZ30K alloy had a relatively lower strength and strain hardening exponent, but an equivalent or slightly higher ductility. In addition, the NZ30K alloy exhibited a certain degree of yield point phenomenon, in contrast to the smooth and continuous stress–strain curve of the GW103K alloy (Fig. 6). Similar yield point phenomenon on the tensile curve was also been reported for the NZ30K alloy by Ma et al. [51,63]. While being slightly inferior to the extruded GW103K alloy due to a lower RE addition, the present NZ30K in the extruded state exhibited higher yield strength and ultimate tensile strength than the alloy in the cast state [57,58,64,65]. 3.3. Hysteresis loops Fig. 7 shows typical stress–strain hysteresis loops of the first and mid-life cycles at a total strain amplitude of 1.2% and strain ratio of Re = 1 for the extruded NZ30K and AM30 alloys, respectively. It is seen from Fig. 7 that the initial slope in the ascending phase of the first cycle was the essentially same in both alloys, indicating that the 3% RE addition did not significantly affect the Young’s modulus. As seen from Fig. 7, unlike the extruded GW103K alloys [23–25], the current NZ30K alloy exhibited asymmetrical hysteresis loops in shape, which were similar to those of RE-free Mg alloys [12–15]. This asymmetrical tensile and compressive yielding phenomenon was somewhat akin to the Bauschingerlike effect [66], which was predominantly associated with the twinning–detwinning process [12–15,67]. It is also seen from Fig. 7 that the very first cycle showed a stronger Bauschinger-like effect (i.e., a flatter curve in the compressive part in the descending

(b)

Fig. 12. SEM images of overall fracture surfaces of the extruded specimens fatigued at a total strain amplitude of (a) 0.4% and (b) 1.0%.

(a)

(b)

Fig. 13. SEM micrographs of the fracture surface near crack initiation of the extruded specimens fatigued at a total strain amplitude of (a) 0.4%, and (b) 1.0%.

69

F.A. Mirza et al. / Materials and Design 64 (2014) 63–73

phase) than the mid-life cycle. The possible reason behind this is that the formation of residual twins gradually became saturated near the mid-life cycle [13]. Such an obvious asymmetry or skewness of hysteresis loops in the present NZ30K alloy was mainly attributed to the existence of stronger crystallographic texture (Fig. 5), which caused the occurrence of extension twinning in compression and detwinning in tension during cyclic deformation along the ED. Similar twinning–detwinning activities in the extruded Mg alloys under cyclic loading were also reported in [12–15]. In addition, unlike the fcc metals where the slope of hysteresis loops after the strain reversal either at the maximum or minimum stress was basically equal to the value of Young’s modulus, the pseudoelastic or nonlinear elastic behavior in both descending and ascending phases in the NZ30K alloy was still present. It appeared reasonable to consider that the pseudoelastic behavior was an inherent characteristic of Mg alloys, which was in essence associated with the superior damping capacity [25]. 3.4. Cyclic deformation response Fig. 8 shows the evolution of stress amplitudes with respect to the number of cycles at different applied strain amplitudes on a semi-log scale. As the applied total strain amplitude decreased, the stress amplitude decreased and the fatigue life of the alloy increased. Unlike the extruded GW103K alloy [23–25] where

microalloying with RE elements led to the cyclic stabilization at almost all the strain amplitudes, it seemed that cyclic stabilization was hardly achieved in the extruded NZ30K alloy as seen from Fig. 8. It exhibited a certain extent of cyclic softening at a low strain amplitude of 0.2%. As the applied strain amplitude increased from 0.4% up to 1.0%, the alloy showed basically cyclic stabilization within the first ten cycles and then cyclic hardening up to failure. At a higher strain amplitude of 1.2%, an initial cyclic hardening within the first three cycles and then small cyclic softening could be seen from Fig. 8. Again, this would also be attributed to the twinning–detwinning activities of this alloy, which will be presented later. In general, cyclic response is governed by the cyclic stability of the microstructural features, dislocation glide and multiplication, twinning, and twin–dislocation interactions. The strain hardening phenomenon mainly arose from the interactions among the dislocations that impede the motion of dislocations [68]. As the  2g extension twinbasal texture in the Mg alloys facilitated f1 0 1 ning, which caused a 86.3° rotation of basal planes [23–26,69]. Such a change in the orientation would act as a barrier to the movement of dislocations on the original slip plane and caused the formation of dislocations pile-ups, which initiated the back stress and thus resulted in the strain hardening [13]. In the low cycle fatigue tests, plastic strain amplitude has been considered as a physical quantity that results in several damaging processes and influences the internal microstructure which is closely related

(a)

(b)

(c)

(d)

PFZ

PFZ

(e)

(f)

PFZ

Fig. 14. SEM micrographs (left: secondary electron images, and right: back-scattered electron images) taken at different magnifications in the fatigue crack propagation region of the extruded specimens fatigued at a total strain amplitude of 0.4%.

70

F.A. Mirza et al. / Materials and Design 64 (2014) 63–73

to the strain resistance and eventually the fatigue life [14]. The change of the plastic strain amplitude (Dep/2) during cyclic deformation is shown in Fig. 9, which corresponded well to the change of the stress amplitude during cyclic deformation as shown in Fig. 8 at different total strain amplitudes. 3.5. Fatigue life and fatigue parameters The fatigue life (i.e., the number of cycles to failure, Nf) as a function of the applied total strain amplitudes (Det/2) of the extruded NZ30K alloy is plotted in Fig. 10, along with the experimental data reported in the literature for a variety of extruded Mg alloys [13,14,23,46,70–72] for comparison. Run-out data points were denoted by arrows pointing horizontally at or more than 107 cycles. The alloy showed basically a similar fatigue life to that of the RE-free extruded Mg alloys within the experimental scatter [13,14,46,70–72]. However, it should be noted that there are potentially orders of magnitude difference in fatigue lives between the upper bound and the lower bound curves in different magnesium alloys. The fatigue life parameters were evaluated according to the equation reported in refs. [12,14,15,26,66,73,74] and summarized in Table 2. It is seen that while the obtained fatigue parameters were well within the range in other fatigued Mg alloys reported in the literature [12–14,74], the cyclic strain hardening exponent n0 of NZ30K alloy was lower than that of GW103K alloy [23]. This corresponded well to the lower monotonic strain

hardening exponent n (Table 1) and cyclic softening characteristics of this alloy (Figs. 8 and 9). It should be pointed out that in evaluating the above fatigue life parameters, the run-out data for the fatigue samples without failure at or above 107 cycles were not included. Furthermore, the obtained cyclic stress–strain curve corresponding to the mid-lives of the NZ30K alloy tested at varying strain amplitudes is shown in Fig. 11 along with the monotonic stress–strain curve. It is seen that the cyclic stress–strain curve was positioned slightly lower than the monotonic one of the extruded NZ30K alloy. The obtained cyclic yield strength (r0 y) and cyclic strain hardening exponent (n0 ) of the NZ30K alloy were also slightly lower than the corresponding monotonic yield strength and strain hardening exponent, as seen from Tables 1 and 2. This could be the reason why the NZ30K alloy exhibited a certain extent of cyclic softening at a low strain amplitude of 0.2%. 3.6. Fractography Fig. 12 shows an overall view of typical fracture surfaces at a lower magnification for the specimens tested at a strain amplitude of 0.4% and 1.0%, respectively. It is seen that fatigue crack initiated basically from the specimen surface. On a close examination at a higher magnification (Fig. 13), fatigue crack indeed initiated from near-surface defects and the fracture surfaces at both strain amplitudes had similar features, i.e., isolated facet of the cleavage-like planes near the initiation site. However, the fatigue crack propagation

(a)

(b)

(c)

(d)

PFZ

(e)

(f)

PFZ PFZ

Fig. 15. SEM micrographs (left: secondary electron images, and right: back-scattered electron images) taken at different magnifications in the fatigue crack propagation region of the extruded specimens fatigued at a total strain amplitude of 1.0%.

F.A. Mirza et al. / Materials and Design 64 (2014) 63–73

400

4. Discussion

Mid-life cycle

300

Stress, MPa

200 100 0 -100 -200

NZ30K

-300

GW103K AM30

-400 -1.5

-1

-0.5

0

0.5

71

1

1.5

Strain, % Fig. 16. Typical stress–strain hysteresis loops of the mid-life cycle at a given total strain amplitude of 1.2% and strain ratio of Re = 1 for the extruded NZ30K, GW103K, and AM30 alloys, respectively.

characteristics of extruded NZ30K alloy in both strain amplitudes were quite different from those of GW103K alloy [23–25], which were basically characterized by serrated fatigue striations observed at higher magnifications (Figs. 14 and 15). It is known that the occurrence of fatigue striations was due to a repeated plastic blunting–sharpening process in face-centered cubic (fcc) materials stemming from the slip of dislocations in the plastic zone ahead of the fatigue crack tip [75]. The formation of fatigue striations in the HCP alloy was anticipated to be related to both dislocation slip and twinning in the plastic zone during fatigue crack propagation [12–15]. It should be noted that the fatigue striations were perpendicular to the crack propagation direction. A similar type of fracture surfaces was reported in a cast Mg–3Nd–0.2Zn– 0.5Zr alloy by Li et al. [29]. Furthermore, a visible PFZ was seen from the back-scattered electron images (Fig. 14(b), (d) and (f) and Fig. 15(b), (d) and (f)).

(a)

In polycrystalline alloys, the mechanical properties are directly related to the presence of crystallographic textures which dominate the orientation of slip or twinning planes and directions relative to the externally applied stress, thus affect the yield asymmetry between tension and compression [26,76,77]. Since  2g extension twinning is a key deformation mechanism in f1 0 1 Mg alloys [12–16,78], the combination of twinning and crystallographic textures in wrought Mg alloys is mainly responsible for the tension–compression yield asymmetry which is commonly observed in nearly all RE-free wrought Mg alloys [12–15,21,26]. It was also reported by Jain et al. [76] that the yield asymmetry is sensitive to textures. As presented earlier in Section 3.1, the texture of the extruded NZ30K alloy was not sufficiently weakened by the addition of RE elements (Fig. 5) as compared to the GW103K alloy [23], along with the presence of PFZs (Fig. 1) and grainy Mg12Nd precipitates distributed in the intra-crystalline and along the grain boundaries. The basic difference between the NZ30K alloy and GW103K [23] was the amount of RE elements, i.e., only 3 wt.% Nd vs. 10 wt.% Gd and 3 wt.% Y. As the RE content increased, the texture sharpness generally decreased [79]. The differences in both the type and amount of RE elements were thus expected to have a strong influence on the texture formation which plays an important role in the yield asymmetry of the NZ30K alloy (Fig. 7). To better see the difference, the stress–strain hysteresis loop of the mid-life cycle of the extruded NZ30K alloy is re-plotted in Fig. 16, in conjunction with that of both RE-free Mg alloy (AM30) and RE-containing GW103K alloy. It can be clearly seen that the hysteresis loop of NZ30K alloy exhibited a strong skewed asymmetrical shape similar to that of the RE-free AM30 alloy, unlike GW103K alloy (Fig. 16). The ratio of the compressive-to-tensile yield stress was 0.55 for both AM30 and NZ30K alloys, as compared to the GW103K alloy where the ratio was almost 1 [23]. This can again be attributed to the lower amount of RE element in NZ30K alloy. On the other hand, as seen from Fig. 17 which shows the distribution of deformation twins in the area near the fracture surface in the fatigued samples of AM30, GW103K and NZ30K

(b)

(c)

Fig. 17. Optical micrographs in the area near the fracture surface at a strain amplitude of 1.2%, showing the distribution of deformation twins in the fatigued samples of (a) AM30, (b) as-extruded GW103K, and (c) as-extruded NZ30K alloys, respectively.

72

F.A. Mirza et al. / Materials and Design 64 (2014) 63–73

alloys, respectively, the grain size of the RE-containing alloys (NZ30K and GW103K) was smaller than that of the RE-free AM30 alloy (Fig. 17(b), (c) vs. (a)). However, in the extruded NZ30K alloy the smaller grain size did not seem to affect the hysteresis loops considerably, as compared with the GW103K alloy [23]. The most striking feature of the NZ30K alloy was the presence of PFZs along the grain boundaries (as shown in Figs. 1, 14, 15 and 17). This, in conjunction with the absence of a high density of nano-sized and bamboo-leaf/plate-shaped b0 (Mg7(Gd, Y)) precipitates, increased the tendency of twinning in the NZ30K alloy as indicated by the dense residual twins near the fracture surface area (Fig. 17(c)), which subsequently led to the asymmetric tension–compression hysteresis loops (Figs. 7 and 16). The above results indicated that the effect of 3 wt.% Nd on the overall cyclic deformation behavior of the NZ30K alloy was not so strong. The change in the deformation behavior, as compared with the high RE-containing GW103K alloy, was mainly related to (i) the presence of PFZs (Fig. 1), (ii) grainy Mg12Nd precipitates distributed in the intra-crystalline and along the grain boundaries (Fig. 1), and (iii) the stronger texture (Fig. 5). The coupled role of these factors was responsible for the hysteresis loop asymmetry (Figs. 7 and 16), the decrease in the stress amplitude (Fig. 8) and the increase in the plastic strain amplitude at a given total strain amplitude (Fig. 9), while a similar fatigue life was obtained among the NZ30K, GW103K, and AM30 alloys within the experimental scatter. 5. Conclusions Strain-controlled low cycle fatigue tests were conducted on an extruded NZ30K alloy with varying total strain amplitudes at room temperature. The following conclusions can be drawn from this investigation: (1) The microstructure of the NZ30K alloy consisted of a bimodal microstructure with equiaxed recrystallized grains and unrecrystallized coarse grains with an average grain size of about approximately 5–7 lm. A large number of smaller second-phase particles were present inside the grains and larger particles along the grain boundaries accompanied by the presence of characteristic precipitate free zone (PFZ). (2) Unlike the higher RE-containing GW103K alloy, the NZ30K alloy exhibited basically asymmetrical hysteresis loops. This was predominantly due to the presence of the relatively stronger crystallographic texture with c-axes perpendicular to the extrusion direction, which promoted heavily the twinning–detwinning activities during cyclic deformation. (3) The cyclic stabilization was barely achieved even at the lower strain amplitudes. The extruded NZ30K alloy exhibited cyclic softening at low strain amplitudes and cyclic hardening at high strain amplitudes. However, the fatigue life of this alloy was observed to be equivalent to that of other Mg alloys. (4) SEM examinations revealed that fatigue cracks initiated from the near-surface defects and the initiation site contained isolated facet of the cleavage-like planes. The propagation was basically characterized by serrated fatigue striations. The PFZ was further confirmed from the backscattered electron images.

Acknowledgements The authors would like to thank the Natural Sciences and Engineering Research Council of Canada (NSERC) and AUTO21 Network

of Centres of Excellence for providing financial support. This investigation involves part of Canada-China-USA Collaborative Research Project on the Magnesium Front End Research and Development (MFERD). The authors also thank Professor A.A. Luo from Ohio State University (formerly with General Motors Research and Development Center) for the supply of extruded AM30 magnesium alloy. One of the authors (X.Q. Zeng) would like to give thanks to Ministry of Science and Technology of China (MOST) and Natural Science Foundation of China (NSFC) for their financial support (Project No. 2011BAE22B02, No. 2011DFA50907, No. 51171113). One of the authors (D.L. Chen) is also grateful for the financial support by the Premier’s Research Excellence Award (PREA), NSERC-Discovery Accelerator Supplement (DAS) Award, Automotive Partnership Canada (APC), Canada Foundation for Innovation (CFI), and Ryerson Research Chair (RRC) program. The authors would like to thank Dr. R. Tandon and Dr. B. Davies (Magnesium Elektron) for supplying magnesium powders for the defocusing calibration. The authors would also like to thank Messrs. A. Machin, Q. Li, C. Ma, J. Amankrah and R. Churaman for easy access to the laboratory facilities of Ryerson University and their assistance in the experiments.

References [1] Pollock TM. Weight loss with magnesium alloys. Science 2010;328:986–7. [2] Murray J, King D. Oil’s tipping point has passed. Nature 2012;481:433–5. [3] Zheng XY, Lee H, Weisgraber TH, Shusteff M, DeOtte J, Duoss EB, et al. Ultralight, ultrastiff mechanical metamaterials. Science 2014;344:1373–7. [4] Nie JF, Zhu YM, Liu JZ, Fang XY. Periodic segregation of solute atoms in fully coherent twin boundaries. Science 2013;340:957–60. [5] Ash C, Culotta E, Fahrenkamp-Uppenbrink J, Malakoff D, Smith J, Sugden A, et al. Once and future climate change. Science 2013;341:473. [6] McNutt M. Climate change impacts. Science 2013;341:435. [7] Reyes-Fox M, Steltzer H, Trlica MJ, McMaster GS, Andales AA, LeCain DR, et al. Elevated CO2 further lengthens growing season under warming conditions. Nature 2014;510:259–62. [8] Luo AA. Magnesium: current and potential automotive applications. JOM 2002;54:42–8. [9] Kainer KU. Magnesium – alloys and technology. Cambridge: Wiley-VCH; 2003. [10] Liu XB, Chen RS, Han EH. Effects of ageing treatment on microstructures and properties of Mg–Gd–Y–Zr alloys with and without Zn additions. J Alloys Compd 2008;465:232–8. [11] Bae DH, Lee MH, Kim KT, Kim WT, Kim DH. Application of quasicrystalline particles as a strengthening phase in Mg–Zn–Y alloys. J Alloy Compd 2002;342:445–50. [12] Lin XZ, Chen DL. Strain controlled cyclic deformation behavior of an extruded magnesium alloy. Mater Sci Eng, A 2008;496:106–13. [13] Begum S, Chen DL, Xu S, Luo AA. Strain-controlled low-cycle fatigue properties of a newly developed extruded magnesium alloy. Metall Mater Trans A 2008;39:3014–26. [14] Begum S, Chen DL, Xu S, Luo AA. Low cycle fatigue properties of an extruded AZ31 magnesium alloy. Int J Fatigue 2009;31:726–35. [15] Begum S, Chen DL, Xu S, Luo AA. Effect of strain ratio and strain rate on low cycle fatigue behavior of AZ31 wrought magnesium alloy. Mater Sci Eng, A 2009;517:334–43. [16] Fan CL, Chen DL, Luo AA. Dependence of the distribution of deformation twins on strain amplitudes in an extruded magnesium alloy after cyclic deformation. Mater Sci Eng, A 2009;519:38–45. [17] Peng Z-K, Zhang X-M, Chen J-M, Xiao Y, Jiang H. Grain refining mechanism in Mg–9Gd–4Y alloys by zirconium. Mater Sci Technol 2005;21:722–6. [18] Li WP, Zhou H, Li ZF. Effect of gadolinium on microstructure and rolling capability of AZ31 alloy. J Alloys Compd 2009;475:227–32. [19] Fang XY, Yi DQ, Wang B, Luo WH, Gu W. Effect of yttrium on microstructures and mechanical properties of hot rolled AZ61 wrought magnesium alloy. Trans Nonferrous Met Soc China 2006;16:1053–8. [20] Yang F, Lv F, Yang XM, Li SX, Zhang ZF, Wang QD. Enhanced very high cycle fatigue performance of extruded Mg–12Gd–3Y–0.5Zr magnesium alloy. Mater Sci Eng, A 2011;528:2231–8. [21] Yin SM, Yang HJ, Li SX, Wu SD, Yang F. Cyclic deformation behavior of asextruded Mg–3%Al–1%Zn. Scripta Mater 2008;58:751–4. [22] Patel HA, Chen DL, Bhole SD, Sadayappan K. Low cycle fatigue behavior of a semi-solid processed AM60B magnesium alloy. Mater Des 2013;49:456–64. [23] Mirza FA, Chen DL, Li DJ, Zeng XQ. Low cycle fatigue of a rare-earth containing extruded magnesium alloy. Mater Sci Eng, A 2013;575:65–73. [24] Mirza FA, Chen DL, Li DJ, Zeng XQ. Effect of strain ratio on cyclic deformation behavior of a rare-earth containing extruded magnesium alloy. Mater Sci Eng, A 2013;5:250–9.

F.A. Mirza et al. / Materials and Design 64 (2014) 63–73 [25] Mirza FA, Chen DL, Li DJ, Zeng XQ. Cyclic deformation behavior of a rare-earth containing extruded magnesium alloy: effect of heat treatment. Metall Mater Trans A (2013) (submitted for publication). [26] Mirza FA, Chen DL. In: Zhang S, Zhao DL, editors. Aerospace materials handbook. New York: CRC Press, Taylor & Francis; 2013. [27] Hirsch J, Al-Samman T. Superior light metals by texture engineering: optimized aluminum and magnesium alloys for automotive applications. Acta Mater 2013;61:818–43. [28] Stanford N, Barnett MR. Effect of composition on the texture and deformation behaviour of wrought Mg alloys. Scr Mater 2008;58:179–82. [29] Li ZM, Wang QG, Luo AA, Fu PH, Peng LM, Wang YG, et al. High cycle fatigue of cast Mg–3Nd–0.2Zn magnesium alloys. Metall Mater Trans A 2013;44A:5202–15. [30] Yang Y, Liu YB, Qin SY, Fang Y. High cycle fatigue properties of die-cast magnesium alloy AZ91D with addition of different concentrations of cerium. J Rare Earths 2006;24:591–5. [31] Dong J, Liu WC, Song X, Zhang P, Ding WJ, Korsunsky AM. Influence of heat treatment on fatigue behavior of high-strength Mg–10Gd–3Y alloy. Mater Sci Eng, A 2010;725:6053–63. [32] Yang Y, Liu YB. High cycle fatigue characterization of two die-cast magnesium alloys. Mater Charact 2008;59:567–70. [33] Bayani H, Saebnoori E. Effect of rare earth elements addition on thermal fatigue behaviors of AZ31 magnesium alloy. J Rare Earths 2009;27:255–8. [34] Nascimento L, Yi SB, Bohlen J, Fuskova L, Letzig D, Kainer KU. High cycle fatigue behaviour of magnesium alloys. Procedia Eng 2010;2:743–50. [35] Xu YL, Zhang K, Li XG, Lei J, Yang YS, Luo TJ. High cycle fatigue properties of die-cast magnesium alloy AZ91D-1%MM. Trans Nonferrous Met Soc China 2008;18:306–11. [36] Liu W-C, Dong J, Zhang P, Jin L, Peng T, Zhai CQ, et al. Fatigue behavior of hotextruded Mg–10Gd–3Y magnesium alloy. J Mater Res 2010;25(4):773–83. [37] Zhu R, Cai XT, Wu YJ, Liu LL, Ji WQ, Hua B. Low-cycle fatigue behavior of extruded Mg–10Gd–2Y–0.5Zr alloys. Mater Des 2014;53:992–7. [38] Yin SM, Li SX. Low cycle fatigue behaviors of an as-extruded Mg–12%Gd–3%Y– 0.5%Zr alloy. J Mater Sci Technol 2013;29(8):775–80. [39] Wang FH, Dong J, Jiang YY, Ding WJ. Cyclic deformation and fatigue of extruded Mg–Gd–Y magnesium alloy. Mater Sci Eng, A 2013;561:403–10. [40] Wu L, Yang Z, Xia W, Chen Z, Yang L. The cyclic softening and evolution of microstructures for Mg–10Gd–2.0Y–0.46Zr alloy under low cycle fatigue at 573 K. Mater Des 2012;36:47–53. [41] Ma CY, Che X, Zhao B, Chen LJ. Low-cycle fatigue behavior of hot-extruded Al4% Mg-xCe alloys. J Northeast Univ 2007;28:144–9. ˇ anadija M, Guo XF, Lanc D, Yang WP, Brnic´ J. Low cycle fatigue and [42] C mechanical properties of magnesium alloy Mg–6Zn–1Y–0.6Ce–0.6Zr at different temperatures. Mater Des 2014;59:287–95. [43] Xiong Y, Yu Q, Jiang YY. Multiaxial fatigue of extruded AZ31B magnesium alloy. Mater Sci Eng, A 2012;546:119–28. [44] Xiong Y, Yu Q, Jiang YY. An experimental study of cyclic plastic deformation of extruded ZK60 magnesium alloy under uniaxial loading at room temperature. Int J Plasticity 2014;53:107–24. [45] Xiong Y, Jiang YY. Fatigue of ZK60 magnesium alloy under uniaxial loading. Int J Fatigue 2014;64:74–83. [46] Yu Q, Zhang JX, Jiang YY, Li QZ. An experimental study on cyclic deformation and fatigue of extruded ZK60 magnesium alloy. Int J Fatigue 2012;36(1):47–58. [47] Yu Q, Zhang JX, Jiang YY, Li QZ. Effect of strain ratio on cyclic deformation and fatigue of extruded AZ61A magnesium alloy. Int J Fatigue 2012;44:225–33. [48] Zhang J, Yu Q, Jiang YY, Li QZ. An experimental study of cyclic deformation of extruded AZ61A magnesium alloy. Int J Plasticity 2011;27(5):768–87. [49] Bachmann F, Hielscher R, Schaeben H. Texture analysis with MTEX – free and open source software toolbox. Solid State Phenom 2010;160:63–8. [50] Wen LH, Ji ZS, Li XL. Effect of extrusion ratio on microstructure and mechanical properties of Mg–Nd–Zn–Zr alloys prepared by a solid recycling process. Mater Charact 2008;59:1655–60. [51] Ma L, Peng LM, Jiang HY, Ding WJ. Development of high strength and ductility Mg–Nd–Zn based wrought alloy. In: Poole WJ, Kainer KU, editors. Proceedings of 9th international conference on magnesium alloys and their applications; 2012. p. 999–1005. [52] Yu K, Rui ST, Song JM, Li WX, Guo L. Effects of grain refinement on mechanical properties and microstructures of AZ31 alloy. Trans Nonferrous Met Soc China 2008;18:s39–43. [53] Li MZ, Wang YQ, Li C, Liu XG, Xu BS. Effects of neodymium rich rare earth elements on microstructure and mechanical properties of as cast AZ31 magnesium alloy. Mater Sci Technol 2011;27:1138–42. [54] Chaudhuri B, Das PP. Effect of different variables on the precipitate free zones adjacent to grain boundry region in conventionally quenched and aged

[55]

[56]

[57] [58]

[59]

[60]

[61]

[62]

[63]

[64]

[65]

[66] [67]

[68] [69]

[70]

[71]

[72]

[73]

[74]

[75] [76]

[77]

[78] [79]

73

AI–Zn–Mg alloys. In: Symposium on non-ferrous metals technology: national metallurgical laboratory (CSIR). Jamshedpur; 1968. Rateick RG, Griffith RJ, Hall DA, Thompson KA. Relationship of microstructure to fatigue strength loss in anodised aluminium–copper alloys. Mater Sci Technol 2005;21:1227–35. Zhang C, Luo AA, Peng LM, Stone DS, Chang YA. Thermodynamic modeling and experimental investigation of the magnesium–neodymium–zinc alloys. Intermetallics 2011;19(11):1720–6. Fu PG, Peng LM, Jiang HY, Ma L, Zhai CQ. Chemical composition optimization of gravity cast Mg–yNd–xZn–Zr alloy. Mater Sci Eng, A 2008;496:177–88. Fu PG, Peng LM, Jiang HY, Chang JW, Zhai CQ. Effects of heat treatments on the microstructures and mechanical properties of Mg–3Nd–02Zn–0.4Zr (wt.%) alloy. Mater Sci Eng, A 2008;486:183–92. Yu K, Li W-X, Wang R-C. Mechanical properties and microstructure of as-cast and extruded Mg–(Ce, Nd)–Zn–Zr alloys. J Cent South Univ Technol 2005;12(5):499–502. Mirza FA, Chen DL, Li DJ, Zeng XQ. Effect of rare earth elements on deformation behavior of an extruded Mg–10Gd–3Y–0.5Zr alloy during compression. Mater Des 2013;46:411–8. Al-Samman T. Modification of texture and microstructure of magnesium alloy extrusions by particle-stimulated recrystallization. Mater Sci Eng, A 2013;560:561–6. Zheng FY, Wu YJ, Peng LM, Li XW, Fu PH, Ding WJ. Microstructures and mechanical properties of friction stir processed Mg2.0Nd0.3Zn1.0Zr magnesium alloy. J Magnesium Alloys 2013;1(2):122–7. Ma L, Mishra RK, Peng LM, Luo AA, Ding WJ, Sachdev AK. Texture and mechanical behavior evolution of age-hardenable Mg–Nd–Zn extrusions during aging treatment. Mater Sci Eng, A 2011;529:151–5. Dong J, Zheng XW, Zhang ZY, Wang YX, Jin L, Ding WJ. Plastic formability and auto demonstration parts of Mg–Nd–Zn–Zr magnesium alloy. Mater Sci Forum 2011;690:455–8. Li ZM, Luo AA, Wang QG, Peng LM, Fu PH, Wu GH. Effects of grain size and heat treatment on the tensile properties of Mg–3Nd–0.2Zn (wt%) magnesium alloys. Mater Sci Eng, A 2013;564:450–60. Dieter GE. Mechanical metallurgy. SI Metric ed. New York: McGraw-Hill Inc.; 1986. Wu L, Jain A, Brown DW, Stoica GM, Agnew SR, Clausen B, et al. Twinning– detwinning behavior during the strain-controlled low-cycle fatigue testing of a wrought magnesium alloy. Acta Mater 2008;56:688–95. Yoo MH, Morris JR, Ho KM, Agnew SR. Non-basal slip systems in HCP metals and alloys: source mechanisms. Mater Sci Eng, A 2001;A319–A321:87–92. Liu BY, Wang J, Li B, Lu L, Zhang XY, Shan ZW, et al. Twinning-like lattice reorientation without a crystallographic twinning plane. Nature Comm 2014;5:10. Article number 3297. Lv F, Yang F, Duan QQ, Yang YS, Wu SD, Li SX, et al. Fatigue properties of rolled magnesium alloy (AZ31) sheet: influence of specimen orientation. Int J Fatigue 2011;33:672–82. Jordon JB, Gibson JB, Horstemeyer MF, El-Kadiri H, Baird JC, Luo AA. Effect of twinning, slip, and inclusions on the fatigue anisotropy of extrusion-textured AZ61 magnesium alloy. Mater Sci Eng, A 2011;528:6860–71. Luo TJ, Yang YS, Tong WH, Duan QQ, Dong XG. Fatigue deformation characteristic of as-extruded AM30 magnesium alloy. Mater Des 2010;31:1617–21. Chowdhury SH, Chen DL, Bhole SD, Powidajko E, Weckman DC, Zhou Y. Fiber laser welded AZ31 magnesium alloy: the effect of welding speed on microstructure and mechanical properties. Metall Mater Trans A 2012;43A:2133–47. Patel H, Chen DL, Bhole SD, Sadayappan K. Cyclic deformation and twinning in a semi-solid processed AZ91D magnesium alloy. Mater Sci Eng, A 2010;528:208–19. Laird C. Fatigue crack propagation, ASTM STP 415. West Conshohocken, PA: ASTM International; 1967. Jain J, Poole WJ, Sinclair CW, Gharghouri MA. Reducing the tensioncompression yield asymmetry in a Mg–8Al–0.5Zn alloy via precipitation. Scr Mater 2010;62:301–4. Yin SM, Wang CH, Diao YD, Wu SD, Li SX. Influence of grain size and texture on the yield asymmetry of Mg–3Al–1Zn Alloy. J Mater Sci Technol 2011;27:29–34. Xin YC, Wang MY, Zeng Z, Nie MG, Liu Q. Strengthening and toughening of magnesium alloy by 10–12 extension twins. Scr Mater 2012;66:25–8. Stanford N, Atwell D, Barnett MR. The effect of Gd on the recrystallisation, texture and deformation behaviour of magnesium-based alloys. Acta Mater 2010;58:6773–83.