- Email: [email protected]
Effect of microstructure and texture on the mechanical properties of the as-extruded Mg–Zn–Y–Zr alloys
Materials Science and Engineering A 443 (2007) 248–256
Effect of microstructure and texture on the mechanical properties of the as-extruded Mg–Zn–Y–Zr alloys D.K. Xu a,b,∗ , L. Liu a , Y.B. Xu a , E.H. Han b a
Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, Shenyang 110016, China b Envionmental Corrosion Center, Institute of Metal Research, Chinese Academy of Sciences, Shenyang 110016, China Received 5 June 2006; received in revised form 11 August 2006; accepted 11 August 2006
Abstract This work mainly studied the influence of the microstructure and crystallographic texture on the mechanical properties of the as-extruded Mg–Zn–Y–Zr alloys with different Y contents. The samples were machined from thick plates obtained by extrusion and the tensile tests were performed parallel to extrusion and transverse directions, respectively. Microstructure observation firmly indicated that the grain-refining effect of icosahedral quasicrystal phase (I-phase) was superior to that of the cubic W-phase. In addition, the tensile results indicated that I-phase could effectively improve the strength (yield strength and ultimate tensile strength) of alloys. However, strengthening effect of W-phase was lower. With the quantity of W-phase increasing, the strength of alloys was degraded. It also showed that the alloys were mechanically anisotropic, i.e. the longitudinal strength was higher than that of the transverse direction. However, the ductility of the transverse direction was superior. With the increase of Mg–Zn–Y phases, the anisotropy of the ultimate tensile strength (UTS) between the longitudinal and transverse directions increased remarkably. SEM fracture observations showed that the fractures of the TD samples were characterized by the typical “woody fracture”, with a large amount of cracked Mg–Zn–Y particles (I-phase and W-phase) distributed at the bottom of dimples. With Y content increasing, the average spacing of the zonal distributed Mg–Zn–Y particles on the fracture surface became narrow, which influenced the transverse mechanical properties greatly. © 2006 Elsevier B.V. All rights reserved. Keywords: Mg alloy; Crystallographic texture; Anisotropy; Woody fracture; Mechanical properties
1. Introduction Recently, it has been reported that the mechanical properties of wrought Mg alloys are superior to those of cast Mg alloys, because the former has a finer grain structure [1]. Generally, wrought Mg alloys having a grain size less than 10 m, can be easily obtained just through primary processing such as hot rolling or extrusion [2,3]. However, these procedures generally give rise to a strong basal texture. Since the critical resolved shear stress (CRSS) of a basal plane at room temperature is much lower than that of a non-basal plane [4], the mechanical properties of wrought Mg alloys are greatly influenced by the basal texture [5–9]. At present, wrought Mg–Zn–Y–Zr alloys have attracted wide attention because they have both high yield strength and tensile strength either at room or elevated tem-
∗
Corresponding author. Tel.: 86 24 83978270; fax: 86 24 23891320. E-mail address: [email protected] (D.K. Xu).
0921-5093/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2006.08.037
peratures [10]. It has been reported [11–15] that the content variation of rare earth elements (RE) and element Y in the alloys can influence the mechanical properties greatly, which is mainly ascribed to the Orowan mechanism [16]. However, due to the addition of element Y, a large amount of Mg–Zn–Y phases (W–Mg3 Zn3 Y2 and I–Mg3 Zn6 Y) are formed at the grain boundaries of the as-cast alloys [10,17]. Therefore, when the ascast ingots were forged into thick plates, Mg–Zn–Y phases were cracked and zonal distributed along the deformation direction [18]. With Y content increasing, the quantity of the zonal distributed Mg–Zn–Y phases will increase greatly and the influence of Mg–Zn–Y phases on the mechanical properties of the alloys should be taken into consideration. Therefore, the mechanical anisotropy of the as-extruded Mg–Zn–Y–Zr alloys should not be solely depended on the crystallographic texture, and the influence of Mg–Zn–Y phases should be considered. Therefore, in the present studies, the longitudinal and transverse mechanical properties of the as-extruded Mg–Zn–Y–Zr alloys (with Y contents of 0, 1.08, 1.97and 3.08 wt.%) have
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
249
Table 1 Chemical composition of the as-extruded Mg–Zn–Y–Zr alloys Nominal alloy
Alloy I Alloy II Alloy III Alloy IV
Composition (wt %) Mg
Zn
Y
Zr
Zn (wt %)/ Y (wt %)
Bulk Bulk Bulk Bulk
5.68 5.53 5.64 5.49
0 1.08 1.97 3.08
0.78 0.83 0.73 0.82
– 5.12 2.86 1.78
been investigated and compared to determine the influence of Mg–Zn–Y phases and the crystallographic texture on the mechanical anisotropy.
Fig. 1. X-ray diffraction patterns of the alloys I–IV. (The arrows in the figure indicate the intensifying tendency of the W-phase diffraction peak) (in JPEG format).
2. Experimental procedures
3. Experimental results
The materials used in this study were the as-extruded Mg–Zn–Y–Zr magnesium alloys with different Y contents, which were prepared by special technology in magnesium alloy research department of IMR, China. Through inductively coupled plasma atomic emission spectrum (ICP-AES) apparatus, the chemical compositions of the alloys I–II were determined (listed in Table 1). The alloys were made by melting high-pure magnesium in an electric resistance furnace, and then 6.3 wt.% Zn, 2.0 wt.% Zr and different wt.% Y were added under the protection of SF6 and CO2 mixed gas. After stirring the molten alloy and keeping for 30–40 min at 710 ◦ C to homogenize it, molten alloys were cast into cylindrical ingots with 110 mm in diameter, and 500 mm in height. It has been reported [14] that the tensile properties of the as-cast Mg–Zn–Y–Zr magnesium alloys without homogenization were superior. Therefore, in this experiment no homogenized treatment was carried out to the ascast Mg–Zn–Y–Zr alloys. Then the as-cast ingots were extruded into thick plates with cross section of 14 mm × 60 mm at 390 ◦ C. The extrusion ratio was 10:1. Microstructures of the alloys I–II on the L (longitudinal)–T (transverse) plane were examined by the means of optical microscope (Axiophoto 2 image). The samples were etched with an etchant of 4 ml nitric acid and 96 ml water. Phase analysis and the (0 0 0 2) pole figures representing the basal plane were determined by D/Max 2400 X-ray diffractometer (XRD). The tensile bars with a gauge length of 25 mm and 5 mm in diameter were machined from the alloys. The axial directions of the tensile specimens were parallel to the extruded direction (ED samples) and parallel to the transverse direction (TD samples), respectively. Tensile experiments were conducted on the MTS (858.01 M) testing machine with the constant strain rate of 1 × 10−3 s−1 at room temperature. SEM (XL30-FEG-ESEM) observations using either secondary electron imaging (SEI) or Backscatter electron imaging (BEI) has been done to determine the fracture characteristics and the distribution and quantity of the cracked Mg–Zn–Y phases (I-phase and W-phase) on the fracture surfaces.
3.1. The microstructures of the as-extruded Mg–Zn–Y–Zr alloys XRD analysis reveals that the main phases vary with Y content, as shown in Fig. 1. It shows that with the addition of element Y, the main phases of alloy II are I-phase and ␣-Mg, whereas the main phases of alloys III and IV are I-phase, W-phase and ␣-Mg. With the increase of Y content, the diffraction peak of W-phase will be gradually intensified. It has been reported [17,19,20] that I-phase could form interdendritic eutectic pockets with ␣-Mg. Therefore, based on the optical microstructure of the as-cast alloys I–IV [21], it was very easy to distinguish W-phase from I-phase by their different morphologies. It also indicated [22] that for alloy II, I-phase was the main ternary phase, whereas for alloys III and IV, with the increasing of Y content, the quantity of W-phase gradually increased and I-phase mainly existed at the triple junctions of grain boundaries. In the previous work [11], it indicated that Zn/Y (in wt.%) ratios of I-phase (Mg3 Zn6 Y) and W-phase (Mg3 Zn3 Y2 ) are 4.38 and 1.10, respectively. By a comparison of Zn/Y ratios of the alloys, when Zn/Y ratio exceeds 4.38, it will meet the requirement to completely form I-phase. However, when Zn/Y ratio was between 1.10 and 4.38, then the quantity of Zn could not meet the requirement to completely form I-phase and some W-phase would be formed to make element Y being fully existed in the form of Mg–Zn–Y phases [11]. With the decreasing of Zn/Y ratio, more W-phases will be formed in the Mg matrix. Therefore, through checking the different Zn/Y ratios (listed in Table 1) of the alloys, the change process of the main phases can be easily understood. The microstructure observations of the as-extruded Mg–Zn–Y–Zr alloys with different Y contents are shown in Fig. 2. It shows that the grain size decreases with the addition of element Y. For alloy I (ZK60), the grain size is about 15 m, whereas the grain size of alloys II and III (with Y contents of 1.08 and 1.97 wt.%, respectively) is about 8 m. When Y content reaches to 3.08 wt.%, the grain size of alloy IV is further refined, with the size about 2–4 m. Although the grain
250
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
Fig. 2. Optical micrographs observed on the L–T plane in the samples: (a) alloy I, (b) alloy II, (c) alloy III and (d) alloy IV (in JPEG format).
size of as-cast alloys I–IV was gradually decreased with the increasing of Y content [21], the variation of grain size for the as-extruded alloys II and III did not meet this trend. With Y content increasing almost two times, the change in grain size is not much. It firmly indicates that the grain-refining effect of icosahedral phase (I-phase) is superior to that of W-phase. A zone of higher stress may form near a particle if the particle is harder compared to the matrix. Quasicrystalline phases are known for their high hardness. In addition, it has been reported [22–24] that I-phase could form strong interfaces with the matrix in various orientations, which made I-phase particles having a strong pining effect on the dislocation movement. On the other hand, it has been reported that W-phase is cubic structure [25,26]. Compared to I-phase, not all the interfaces of W-phase with the matrix may be coherent because of the limited symmetry of these phases, so W-phase could not have a strong bonding with the matrix Therefore, during the hot deformation process, the zone of high dislocation density near I-phase particles is greatly larger than that near W-phase. Humphreys [27] indicated that recrystallization originated within a zone of high dislocation density and large lattice misorientation at the particles. Therefore, the “Partial Stimulated Nucleation Effect” of I-phase is stronger during recrystallization, resulting in finer grains. However, the result also indicates that with the increasing of Y content, the quantity of W-phase increases and then the grain-refining effect of W-phase is improved, which makes the change in grain size is not much for alloys II and III. However, when Y content is further increased (alloy IV),
the grain-refining effect of W-phase is further increased. In addition, with the addition of element Y, the grain size has greatly decreased. It suggests that element Y can effectively refine grains of the Mg alloys. It has been reported that the refinement effects of element Y can be mainly ascribed to the factors as follows: (1) rare earth element Y can change solution degree of Zn, which decreases the solidus curve and shortens the time for nucleation, and then reduces the grain size [15]. (2) The formation of Mg–Zn–Y phases, especially I-phase can effectively restrain the grain growth during the dynamic recystallization process [13]. In addition, Zr can refine the grains by acting as nucleus during solidification [28]. However, comparing with the refining factor mentioned above, severe plastic deformation and dynamic recrystallization is the primary source of grain refinement in wrought alloys. Therefore, although no Y content is contained in alloy I, its grain size is also refined greatly. Fig. 1 also shows that Mg–Zn–Y phases (I-phase and W-phase) are zonal distributed along the longitudinal (extruded) direction and the quantity of Mg–Zn–Y phases distributed in the Mg matrix increases greatly with the increasing of Y content. 3.2. The determination of the crystallographic texture Fig. 3 shows the (0 0 0 2) pole figures of alloys. It is apparent by inspection that for alloys I–IV, the basal plane of most grains is parallel to the L (longitudinal)–T (transverse) plane. It also indicates that the addition of element Y can influence the
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
251
Fig. 3. The (0 0 0 2) pole figures: (a) alloys I, (b) alloy II, (c) alloy III and (d) alloy IV (in JPEG format).
{0 0 0 1} basal fiber texture. With Y content increasing, the orientation of the basal plane for alloys II–IV slightly tends to the transverse direction. However, the change of the contour lines with higher density level in (0 0 0 2) pole figures is slight. Therefore, intensities of the basal texture for the alloys are almost the same. The slight variation of the basal texture can be ascribed to the influence of Mg–Zn–Y phases (I-phase and W-phase). It has been reported [29] that recrystallization originates within a zone of high dislocation density and large lattice misorientation at the particles, and proceeds by a rapid polygonization process, resulting in grains with orientations closely related to the as-deformed structure. In Section 3.1, it has explained that during recrystallization process, I-phase or W-phase can create a zone of high dislocation and large lattice misorientation, which influences the orientation of the recrystallized grains, and then indirectly leads to the change of the basal texture.
greatly, and the strength of the ED samples is obvious higher than that of the TD samples, whereas the ductility of the TD samples is superior. Compared with the mechanical properties of the ED samples, it reveals that the alloy with no Y content, the ductility is the highest, but the strength is the lowest. When Y content is 1.08 wt.%, the strength of the alloy is the highest, but the ductility decreases greatly. However, when Y content varies from 1.97 to 3.08 wt.%, the strength shows a tendency to
3.3. The mechanical properties of the alloys The stress–strain curves of the ED and TD samples tested at room temperature are shown in Figs. 4 and 5, respectively. To describe and compare conveniently, the longitudinal and transverse mechanical properties of 0.2% proof yield stress (σ 0.2 ), ultimate tensile strength (UTS), and elongation to failure for the alloys are listed in Table 2. It shows that the variation of Y content can influence the mechanical properties of the alloys
Fig. 4. The stress–strain curves of the ED samples (in Opj format).
252
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
Table 2 The mechanical properties of the as-extruded Mg–Zn–Y–Zr alloys Condition
Alloy I Alloy II Alloy III Alloy IV
Longitudinal direction
Transverse direction
Average grain size (m)
σ 0.2 (MPa)
UTS (MPa)
Elongation (%)
σ 0.2 (MPa)
UTS (MPa)
Elongation (%)
180 200 193 185
316 345 334 324
16.3 10.8 12.3 13.1
128 117 108 72
249 261 226 180
16.4 17.2 14.2 14.6
Fig. 5. The stress–strain curves of the TD samples (in Opj format).
decrease but the ductility has a little increase. In the previous studies [11], it has been suggested that element Y can improve the mechanical properties mainly through refining the grain size and forming I-phase in the Mg matrix. However, when Y content exceeded 1.72 wt.%, the occurrence of W-phase would degrade the strength of the alloys [11]. Based on the Zn/Y ratio [11], when Zn content was about 5.65 wt.%, the alloy with Y content ranged from 1.17 to1.72 wt.%, the Zn/Y ratio was slightly higher than 4.38 and the as-extruded Mg–Zn–Y–Zr alloy could have the highest strength, because element Y mainly existed in the form of I-phase and the volume fraction of icosahedral quasicrystal I-phase reached the peak value. This phenomenon was also reported in Ref. [13]. However, the comparison of the mechanical properties of the TD samples reveals that when Y content is 1.08 wt.%, the alloy can have both the highest ultimate tensile strength (UTS) and the superior ductility. Furthermore, as Y content exceeds 1.08 wt.%, the strength of alloys III and IV decreases greatly. By a comparison of the mechanical properties of the ED and TD samples, it reveals that with Y content increasing, the mechanical anisotropy becomes more and more obvious, as shown in Fig. 6. 3.4. Fracture analysis Fig. 7 shows the fracture surfaces of the ED samples. It reveals that with Y content increasing, the semi-cleavage planes decrease gradually and the plastic dimples increase greatly. Therefore, it indicates that with Y content increasing, the failure modes of the tensile samples will change from ductile–fragile
15 8 8 2–4
Fig. 6. The anisotropy in mechanical properties (ultimate tensile strength (UTS), 0.2% proof yield stress (σ 0.2 ), and elongation to failure) between the longitudinal and transverse directions for the alloys I–IV. (in JPEG format).
failure to ductile failure. Especially for alloy IV, a large amount of plastic dimples and cracked particles (distributed at the bottom of the dimples) can be observed on the fracture surface, which belongs to the ductile failure. In general, the dimple is mainly caused by the existence of inclusions or second phases. It has been reported that the dimple size depends on the inclusion size [30]. Fig. 8 is the backscatter electron image (BEI) of the fracture surfaces of the ED samples. For the TD samples, the secondary electron imaging (SEI) and backscatter electron imaging of the fracture surfaces are shown in Figs. 9 and 10, respectively. Fig. 9 shows that for alloys I–IV, the fractures have the typical “woody fracture” characteristics and the fracture surfaces are covered with a large amount of shallow dimples. Fig. 10 shows that for alloys II–IV, a large amount of zonal distributed Mg–Zn–Y particles can be observed on the fracture surfaces. The microstructure observation (in Fig. 2) clearly shows that after hot extrusion processing, Mg–Zn–Y phases (I-phase and W-phase) are uniformly distributed in the Mg matrix along the extruded direction and the volume fraction of Mg–Zn–Y phases, especially for W-phase increases with Y content. Generally, the interface between the intermetallic particle and the matrix is weak, providing the opening site of a micro-crack as reported in an AZ91 alloy containing the intermetallic particles of Mg17 Al12 [31]. In this study, for alloys II–IV, a large amount of Mg–Zn–Y phase particles are closely spaced and zonal distributed along the extruded direction. Therefore, geometrically necessary dislocations are easily formed around Mg–Zn–Y particles, resulting in decohesion
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
Fig. 7. The secondary SEM fracture images of the ED samples: (a) alloy I, (b) alloy II, (c) alloy III and (d) alloy IV (in JPEG format).
Fig. 8. The backscatter SEM fracture images of the ED samples: (a) alloy I, (b) alloy II, (c) alloy III and (d) alloy IV (in JPEG format).
253
254
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
Fig. 9. The secondary SEM fracture images of the TD samples: (a) alloy I, (b) alloy II, (c) alloy III and (d) alloy IV (in JPEG format).
Fig. 10. The backscatter SEM fracture images of the TD samples: (a) alloy I, (b) alloy II, (c) alloy III and (d) alloy IV (in JPEG format).
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
of the particle/matrix interfaces, especially for W-phase/matrix interfaces, which degrades the tensile strength of the alloys and forms a large amount of dimples on the fracture surface. However, for the TD samples, the spacing and the size of the particles are very small, and the dimple size relies on the particle size; therefore, the dimples on the fractures are very small and shallow, as shown in Fig. 9. 4. Discussion The yield strength of the alloys varies with the grain size and the relationship follows the Hall–Petch equation [32,33]. σy = σ0 + kd −1/2
(1)
where d is the grain size, σ 0 and k are the experimentally derived constants. Based on the equation, alloy IV should have the highest yield strength, because its grain size is the smallest. However, Table 2 shows that the yield strength of alloy IV in both longitudinal and transverse directions is the lowest, whereas the yield strength of alloy II in both directions is the highest. Therefore, besides the grain size, the crystallographic texture and second phase also strongly influence the mechanical properties. However, the (0 0 0 2) pole figures in Fig. 3 show that the main textures of alloys I–IV are all the {0 0 0 1} basal fiber texture. Therefore, the difference of the mechanical properties can be mainly ascribed to Mg–Zn–Y phases (I-phase and W-phase). As Fig. 1 shows, I-phase and W-phase are zonal distributed along the longitudinal direction for alloys II–IV. X-ray analysis indicates that the quantity of W-phase increases with the increasing of Y content. About the influence of Mg–Zn–Y phases on the mechanical properties has also been reported in Refs. [12,17,34], but they mainly focused on the strengthening effect of I-phase. Singh et al. indicated [22] that icosahedral quasicrystal phase could form much better interfaces with Mg-matrix than crystalline phases. It has been reported [11,34] that with the volume fraction of I-phase increasing, the strength of alloys would be improved greatly. For alloy II, the Zn/Y ratio exceeds 4.38 and the main second phase is I-phase (in Fig. 1), so its yield stress (σ 0.2 ) and tensile strength (σ b ) of the ED and TD samples are more superior, with the values of 345 and 261 MPa, respectively. It was also suggested [11] that the existence of W-phase would degrade the mechanical properties of the alloys. For alloys III and IV (with Zn/Y ratio is between 1.10 and 4.38), although the main second phase includes I-phase and W-phase, the quantity of W-phase gradually increases with the increasing of Y content, which results in the strength lower than that of alloy II. In addition, it has been reported [35,36] that the embedment of the second phase particle into the ductile metal matrix can greatly improve their mechanical properties such as stiffness, tensile strength and creep resistance, but simultaneously induce high stress concentrations both with the particle and at the particle/matrix interface, which can trigger voids to nucleate by particle cracking or particle/matrix interface debonding. Fracture analysis (BEI) shows that Mg–Zn–Y particles, especially W-phase distributed on the fracture surfaces increase with Y content greatly, which firmly support the decrease of the strength
255
for alloys III and IV. For the ED samples, with the quantity of Mg–Zn–Y particles, more dimples can be observed on the fracture surface and the ductility of the alloys increases with Y content (listed in Table 2). Since Mg–Zn–Y particles are zonal distributed along the extrusion direction, the quantity of MgZn–Y particles on the transverse cross-sectional plane of the TD samples will be greatly higher than that of the ED samples. Fracture analysis (SEI) shows that for the TD samples, although many dimples can be observed on the fracture surface, the spacing of the dimples decreases with Y contents (in Fig. 9), which leads to the decrease of the ductility and the strength for alloys III and IV (listed in Table 2). In addition, due to the zonal distribution of Mg–Zn–Y particles in the longitudinal direction, the obvious “woody fracture” characteristic occurs on the fracture surfaces for the TD samples, shown in Figs. 9 and 10. By a comparison of the mechanical properties of the ED and TD samples, a very interesting phenomenon can be seen, i.e. the mechanical anisotropy increases greatly with Y content, which is shown in Fig. 6. To describe conveniently, three terms has been defined, i.e. anisotropy in yield stress, anisotropy in ultimate tensile strength (UTS) and anisotropy in elongation, which means how much are the strength and elongation in longitudinal and transverse directions with respect to each other, respectively. It shows that for alloy I (with no Y content), the anisotropy in yield stress and UTS about 29% and 21% respectively, which can be ascribed to the influence of the texture [37]. However, for alloys II–IV, with Y content increasing, the anisotropy in strength increases greatly, especially for the UTS. For alloy IV, the anisotropy in UTS is about 45%. As the discussion mentioned above, the textures of the alloys are almost the same. Therefore, the mechanical anisotropy should be mainly ascribed to the influence of the distribution and quantity of Mg–Zn–Y particles, especially W-phase. Compared with Figs. 8 and 10, it shows that with Y content increasing, the quantity of Mg–Zn–Y particles on the fracture surface of TD samples increases more remarkably than that of the ED samples. Therefore, for the TD samples, with the increase of Y content, more Mg–Zn–Y particles formed on the transverse cross-sectional plane and geometrically necessary dislocations are more closely packed around Mg–Zn–Y particles, especially for W-phase, resulting in decohesion of the particle/matrix interfaces, which greatly degrades the tensile strength of the alloys. Therefore, with Y content increasing, the mechanical anisotropy between the ED and TD increases gradually. In conclusion, although rare earth elements such as element Y can effectively strengthen the Mg–Zn–Zr system alloys, it does not mean that the mechanical properties of the alloy are proportional to Y content. In the previous reports, it suggested that the addition of about 1 wt.% Y could make the as-extruded ZK60 alloy have the superior strength [11,13]. About the explanation of the strengthening mechanism has been reported in detail in Ref. [11]. In this work, with the addition of 1 wt.% Y, the longitudinal and transverse strength of the alloy II has been greatly improved (listed in Table 2). However, the anisotropic strength for alloy II is more obvious than that of alloy I. By a comparison of microstructure, tensile properties and the anisotropy in strength of alloys II to IV, it indicates that I-phase is associated
256
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
with finer grain size, higher strength (with comparable grain size) and lesser anisotropy (with comparable texture). Therefore, to reduce the mechanical anisotropy and fully exploit its potential properties, the controlling of the distribution and quantity of the second phases and the crystallographic texture is very important.
Acknowledgement
5. Conclusions
[1] T. Mukai, H. Watababe, K. Higashi, Mater. Sci. Technol. 16 (2000) 1314. [2] H. Somekawa, H. Watababe, T. Mukai, K. Higashi, Scripta Mater. 48 (2003) 1249. [3] H. Watababe, T. Mukai, K. Ishikawa, K. Higashi, Mater Trans. 43 (2002) 78. [4] R.E. Reed Hill, W.D. Robertson, Acta Metall. 5 (1957) 728. [5] T. Mukai, M. Yamanoi, H. Watababe, K. Higashi, Scripta Mater. 45 (2001) 89. [6] H. Watababe, A. Takara, H. Somekawa, T. Mukai, K. Higashi, Scripta Mater. 52 (2005) 449. [7] Y. Yoshida, Cisar, S. Kamado, Y. Kojima, Mater. Trans. 44 (2003) 468. [8] S.R. Agnew, J.A. Horton, T.M. Lillo, D.B. Brown, Scripta Mater. 50 (2004) 377. [9] W.J. Kim, C.W. An, Y.S. Kim, S.I. Hong, Scripta Mater. 47 (2002) 39. [10] A. Singh, A.P. Tsai, Scripta Mater. 49 (2003) 143. [11] D.K. Xu, L. Liu, Y.B. Xu, E.H. Han, J. Alloys Compd. 426 (2006) 115– 161. [12] X. Zeng, Y. Zhang, C. Lu, W. Ding, Y. Wang, Y. Zhu, J. Alloys Compd. 395 (2005) 213. [13] Y. Zhang, X. Zeng, L. Liu, C. Lu, H. Zhou, Q. Li, Y. Zhu, Mater. Sci. Eng. A 373 (2004) 320. [14] Z.P. Luo, D.Y. Song, S.Q. Zhang, J. Alloys Compd. 230 (1995) 109. [15] C. Ma, M. Liu, G. Wu, W. ding, Y. zhu, Mater. Sci. Eng. A 349 (2003) 207. [16] T. Mohri, M. Mabuchi, N. Saitao, M. Nakamura, Mater. Sci. Eng. A 257 (1998) 287. [17] D.H. Bae, S.H. Kim, W.T. Kim, Acta Mater. 50 (2002) 2343. [18] D.K. Xu, L. Liu, Y.B. Xu, E.H. Han, Mater. Sci. Eng. A 420 (2006) 322. [19] J.Y. Lee, D.H. Kim, H.K. Lim, D.H. Kim, Mater. Lett. 59 (2005) 3801. [20] D.H. Bae, Y. Kim, I.J. Kim, Mater. Lett. 60 (2006) 2190. [21] D.K. Xu, L. Liu, Y.B. Xu, E.H. Han, J. Alloys Compd. doi:10.1016/j.jallcom.2006.05.123, in press. [22] A. Singh, M. Watanabe, A. Kato, A.P. Tsai, Philos. Mag. 86 (2006) 951. [23] A. Singh, M. Nakamura, M. Watanabe, A. Kato, A.P. Tsai, Scripta Mater. 49 (2003) 417. [24] A. Singh, M. Watanabe, A. Kato, A.P. Tsai, Mater. Sci. Eng. A 385 (2004) 382. [25] A. Singh, A.P. Tsai, Scripta Mater. 49 (2003) 143. [26] A. Niikura, A.P. Tsai, A. Inoue, T. Masumoto, Philos. Mag. Lett. 69 (1994) 351. [27] F.J. Humphreys, Acta Metall. 27 (1979) 1801. [28] G. Song, D. StJohn, J. Light Met. 2 (2002) 1. [29] F.J. Humphreys, Acta Metall. 25 (1977) 1323. [30] K. Higashi, Y. Hirai, T. Ohnishi, J. Jpn. Inst. Light Met. 35 (1985) 520. [31] Y.Z. Lu, Q.D. Wang, X.Q. Zeng, W.J. Ding, C.Q. Zhai, Y.P. Zhu, Mater. Sci. Eng. A 278 (2000) 66. [32] N.J. Petch, J. Iron Steel Inst. 174 (1953) 25. [33] E.O. Hall, Proc. Phys. Soc. B 64 (1951) 747. [34] E.S. Park, S. Yi, J.B. Ok, D.H. Bae, W.T. Kim, Proceedings of the MRS Fall Meeting, Boston, MA, 2001. [35] Y. Brechet, J.D. Embury, S. Tao, L. Luo, Acta Metall. Mater. 39 (1991) 1781. [36] D.J. Lloyd, Acta Metall. Mater. 39 (1991) 59. [37] S. Kleiner, P.J. Uggowitzer, Mater. Sci. Eng. A 379 (2004) 258.
The present study investigated microstructure and the tensile properties along longitudinal and transverse direction of the as-extruded ZK60 alloys alloyed with different quantities of element Y, which leads to the following conclusions: 1. Microstructure observation shows that with Y content increasing, the grain size of the alloys decreases gradually. When Y content is 1.08 wt.%, the main second phase is Iphase. However, when Y content is 1.97 and 3.08 wt.%, the main second phases are I-phase and W-phase. It firmly indicates that the grain-refining effect of icosahedral quasicrystal phase (I-phase) is superior to that of the cubic W-phase. 2. Through comparing the mechanical properties of the ED samples, it shows that alloy II (with 1.08 wt.% Y) has the superior tensile strength. Since the Zn/Y ratio of alloy II can fully meet the requirement of completely forming I-phases, so the strengthening effect can be mainly ascribed to the strengthening of icosahedral quasicrystal I-phase. 3. The (0 0 0 2) pole figures of alloys I–IV are almost the same, which indicates that the main texture is {0 0 0 1} basal fiber textures. However, by a comparison of the difference of the anisotropy in the mechanical properties for the alloys, it suggests that besides the influence of the texture, the quantity of the zonal distributed Mg–Zn–Y particles, especially W-phase is also a main factor in determining the anisotropy in the mechanical properties of the alloys. Although the anisotropy in the ultimate tensile strength (UTS) of alloy II is more obvious than that of alloy I, both the longitudinal and transverse strength of alloy II has been greatly improved. However, alloy II has less anisotropy in strength than the other two alloys (III and IV), even though these other two alloys have comparable or finer grain sizes and comparable texture. It mainly attributes to the role of I-phase. 4. With Y content increasing, the failure modes of the tensile samples will change from ductile–fragile failure to ductile failure. The fractures of the TD samples were characterized by the typical “woody fracture”, with a large amount of cracked Mg–Zn–Y particles (I-phase and W-phase) distributed at the bottom of dimples. With Y content increasing, the average spacing of the zonal distributed Mg–Zn–Y particles on the fracture surface became narrow, which influenced the transverse mechanical properties greatly.
This work was supported by National Science Fund of China (NSFC) project under grant No. 50431020. References
Effect of microstructure and texture on the mechanical properties of the as-extruded Mg–Zn–Y–Zr alloys D.K. Xu a,b,∗ , L. Liu a , Y.B. Xu a , E.H. Han b a
Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, Shenyang 110016, China b Envionmental Corrosion Center, Institute of Metal Research, Chinese Academy of Sciences, Shenyang 110016, China Received 5 June 2006; received in revised form 11 August 2006; accepted 11 August 2006
Abstract This work mainly studied the influence of the microstructure and crystallographic texture on the mechanical properties of the as-extruded Mg–Zn–Y–Zr alloys with different Y contents. The samples were machined from thick plates obtained by extrusion and the tensile tests were performed parallel to extrusion and transverse directions, respectively. Microstructure observation firmly indicated that the grain-refining effect of icosahedral quasicrystal phase (I-phase) was superior to that of the cubic W-phase. In addition, the tensile results indicated that I-phase could effectively improve the strength (yield strength and ultimate tensile strength) of alloys. However, strengthening effect of W-phase was lower. With the quantity of W-phase increasing, the strength of alloys was degraded. It also showed that the alloys were mechanically anisotropic, i.e. the longitudinal strength was higher than that of the transverse direction. However, the ductility of the transverse direction was superior. With the increase of Mg–Zn–Y phases, the anisotropy of the ultimate tensile strength (UTS) between the longitudinal and transverse directions increased remarkably. SEM fracture observations showed that the fractures of the TD samples were characterized by the typical “woody fracture”, with a large amount of cracked Mg–Zn–Y particles (I-phase and W-phase) distributed at the bottom of dimples. With Y content increasing, the average spacing of the zonal distributed Mg–Zn–Y particles on the fracture surface became narrow, which influenced the transverse mechanical properties greatly. © 2006 Elsevier B.V. All rights reserved. Keywords: Mg alloy; Crystallographic texture; Anisotropy; Woody fracture; Mechanical properties
1. Introduction Recently, it has been reported that the mechanical properties of wrought Mg alloys are superior to those of cast Mg alloys, because the former has a finer grain structure [1]. Generally, wrought Mg alloys having a grain size less than 10 m, can be easily obtained just through primary processing such as hot rolling or extrusion [2,3]. However, these procedures generally give rise to a strong basal texture. Since the critical resolved shear stress (CRSS) of a basal plane at room temperature is much lower than that of a non-basal plane [4], the mechanical properties of wrought Mg alloys are greatly influenced by the basal texture [5–9]. At present, wrought Mg–Zn–Y–Zr alloys have attracted wide attention because they have both high yield strength and tensile strength either at room or elevated tem-
∗
Corresponding author. Tel.: 86 24 83978270; fax: 86 24 23891320. E-mail address: [email protected] (D.K. Xu).
0921-5093/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2006.08.037
peratures [10]. It has been reported [11–15] that the content variation of rare earth elements (RE) and element Y in the alloys can influence the mechanical properties greatly, which is mainly ascribed to the Orowan mechanism [16]. However, due to the addition of element Y, a large amount of Mg–Zn–Y phases (W–Mg3 Zn3 Y2 and I–Mg3 Zn6 Y) are formed at the grain boundaries of the as-cast alloys [10,17]. Therefore, when the ascast ingots were forged into thick plates, Mg–Zn–Y phases were cracked and zonal distributed along the deformation direction [18]. With Y content increasing, the quantity of the zonal distributed Mg–Zn–Y phases will increase greatly and the influence of Mg–Zn–Y phases on the mechanical properties of the alloys should be taken into consideration. Therefore, the mechanical anisotropy of the as-extruded Mg–Zn–Y–Zr alloys should not be solely depended on the crystallographic texture, and the influence of Mg–Zn–Y phases should be considered. Therefore, in the present studies, the longitudinal and transverse mechanical properties of the as-extruded Mg–Zn–Y–Zr alloys (with Y contents of 0, 1.08, 1.97and 3.08 wt.%) have
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
249
Table 1 Chemical composition of the as-extruded Mg–Zn–Y–Zr alloys Nominal alloy
Alloy I Alloy II Alloy III Alloy IV
Composition (wt %) Mg
Zn
Y
Zr
Zn (wt %)/ Y (wt %)
Bulk Bulk Bulk Bulk
5.68 5.53 5.64 5.49
0 1.08 1.97 3.08
0.78 0.83 0.73 0.82
– 5.12 2.86 1.78
been investigated and compared to determine the influence of Mg–Zn–Y phases and the crystallographic texture on the mechanical anisotropy.
Fig. 1. X-ray diffraction patterns of the alloys I–IV. (The arrows in the figure indicate the intensifying tendency of the W-phase diffraction peak) (in JPEG format).
2. Experimental procedures
3. Experimental results
The materials used in this study were the as-extruded Mg–Zn–Y–Zr magnesium alloys with different Y contents, which were prepared by special technology in magnesium alloy research department of IMR, China. Through inductively coupled plasma atomic emission spectrum (ICP-AES) apparatus, the chemical compositions of the alloys I–II were determined (listed in Table 1). The alloys were made by melting high-pure magnesium in an electric resistance furnace, and then 6.3 wt.% Zn, 2.0 wt.% Zr and different wt.% Y were added under the protection of SF6 and CO2 mixed gas. After stirring the molten alloy and keeping for 30–40 min at 710 ◦ C to homogenize it, molten alloys were cast into cylindrical ingots with 110 mm in diameter, and 500 mm in height. It has been reported [14] that the tensile properties of the as-cast Mg–Zn–Y–Zr magnesium alloys without homogenization were superior. Therefore, in this experiment no homogenized treatment was carried out to the ascast Mg–Zn–Y–Zr alloys. Then the as-cast ingots were extruded into thick plates with cross section of 14 mm × 60 mm at 390 ◦ C. The extrusion ratio was 10:1. Microstructures of the alloys I–II on the L (longitudinal)–T (transverse) plane were examined by the means of optical microscope (Axiophoto 2 image). The samples were etched with an etchant of 4 ml nitric acid and 96 ml water. Phase analysis and the (0 0 0 2) pole figures representing the basal plane were determined by D/Max 2400 X-ray diffractometer (XRD). The tensile bars with a gauge length of 25 mm and 5 mm in diameter were machined from the alloys. The axial directions of the tensile specimens were parallel to the extruded direction (ED samples) and parallel to the transverse direction (TD samples), respectively. Tensile experiments were conducted on the MTS (858.01 M) testing machine with the constant strain rate of 1 × 10−3 s−1 at room temperature. SEM (XL30-FEG-ESEM) observations using either secondary electron imaging (SEI) or Backscatter electron imaging (BEI) has been done to determine the fracture characteristics and the distribution and quantity of the cracked Mg–Zn–Y phases (I-phase and W-phase) on the fracture surfaces.
3.1. The microstructures of the as-extruded Mg–Zn–Y–Zr alloys XRD analysis reveals that the main phases vary with Y content, as shown in Fig. 1. It shows that with the addition of element Y, the main phases of alloy II are I-phase and ␣-Mg, whereas the main phases of alloys III and IV are I-phase, W-phase and ␣-Mg. With the increase of Y content, the diffraction peak of W-phase will be gradually intensified. It has been reported [17,19,20] that I-phase could form interdendritic eutectic pockets with ␣-Mg. Therefore, based on the optical microstructure of the as-cast alloys I–IV [21], it was very easy to distinguish W-phase from I-phase by their different morphologies. It also indicated [22] that for alloy II, I-phase was the main ternary phase, whereas for alloys III and IV, with the increasing of Y content, the quantity of W-phase gradually increased and I-phase mainly existed at the triple junctions of grain boundaries. In the previous work [11], it indicated that Zn/Y (in wt.%) ratios of I-phase (Mg3 Zn6 Y) and W-phase (Mg3 Zn3 Y2 ) are 4.38 and 1.10, respectively. By a comparison of Zn/Y ratios of the alloys, when Zn/Y ratio exceeds 4.38, it will meet the requirement to completely form I-phase. However, when Zn/Y ratio was between 1.10 and 4.38, then the quantity of Zn could not meet the requirement to completely form I-phase and some W-phase would be formed to make element Y being fully existed in the form of Mg–Zn–Y phases [11]. With the decreasing of Zn/Y ratio, more W-phases will be formed in the Mg matrix. Therefore, through checking the different Zn/Y ratios (listed in Table 1) of the alloys, the change process of the main phases can be easily understood. The microstructure observations of the as-extruded Mg–Zn–Y–Zr alloys with different Y contents are shown in Fig. 2. It shows that the grain size decreases with the addition of element Y. For alloy I (ZK60), the grain size is about 15 m, whereas the grain size of alloys II and III (with Y contents of 1.08 and 1.97 wt.%, respectively) is about 8 m. When Y content reaches to 3.08 wt.%, the grain size of alloy IV is further refined, with the size about 2–4 m. Although the grain
250
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
Fig. 2. Optical micrographs observed on the L–T plane in the samples: (a) alloy I, (b) alloy II, (c) alloy III and (d) alloy IV (in JPEG format).
size of as-cast alloys I–IV was gradually decreased with the increasing of Y content [21], the variation of grain size for the as-extruded alloys II and III did not meet this trend. With Y content increasing almost two times, the change in grain size is not much. It firmly indicates that the grain-refining effect of icosahedral phase (I-phase) is superior to that of W-phase. A zone of higher stress may form near a particle if the particle is harder compared to the matrix. Quasicrystalline phases are known for their high hardness. In addition, it has been reported [22–24] that I-phase could form strong interfaces with the matrix in various orientations, which made I-phase particles having a strong pining effect on the dislocation movement. On the other hand, it has been reported that W-phase is cubic structure [25,26]. Compared to I-phase, not all the interfaces of W-phase with the matrix may be coherent because of the limited symmetry of these phases, so W-phase could not have a strong bonding with the matrix Therefore, during the hot deformation process, the zone of high dislocation density near I-phase particles is greatly larger than that near W-phase. Humphreys [27] indicated that recrystallization originated within a zone of high dislocation density and large lattice misorientation at the particles. Therefore, the “Partial Stimulated Nucleation Effect” of I-phase is stronger during recrystallization, resulting in finer grains. However, the result also indicates that with the increasing of Y content, the quantity of W-phase increases and then the grain-refining effect of W-phase is improved, which makes the change in grain size is not much for alloys II and III. However, when Y content is further increased (alloy IV),
the grain-refining effect of W-phase is further increased. In addition, with the addition of element Y, the grain size has greatly decreased. It suggests that element Y can effectively refine grains of the Mg alloys. It has been reported that the refinement effects of element Y can be mainly ascribed to the factors as follows: (1) rare earth element Y can change solution degree of Zn, which decreases the solidus curve and shortens the time for nucleation, and then reduces the grain size [15]. (2) The formation of Mg–Zn–Y phases, especially I-phase can effectively restrain the grain growth during the dynamic recystallization process [13]. In addition, Zr can refine the grains by acting as nucleus during solidification [28]. However, comparing with the refining factor mentioned above, severe plastic deformation and dynamic recrystallization is the primary source of grain refinement in wrought alloys. Therefore, although no Y content is contained in alloy I, its grain size is also refined greatly. Fig. 1 also shows that Mg–Zn–Y phases (I-phase and W-phase) are zonal distributed along the longitudinal (extruded) direction and the quantity of Mg–Zn–Y phases distributed in the Mg matrix increases greatly with the increasing of Y content. 3.2. The determination of the crystallographic texture Fig. 3 shows the (0 0 0 2) pole figures of alloys. It is apparent by inspection that for alloys I–IV, the basal plane of most grains is parallel to the L (longitudinal)–T (transverse) plane. It also indicates that the addition of element Y can influence the
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
251
Fig. 3. The (0 0 0 2) pole figures: (a) alloys I, (b) alloy II, (c) alloy III and (d) alloy IV (in JPEG format).
{0 0 0 1} basal fiber texture. With Y content increasing, the orientation of the basal plane for alloys II–IV slightly tends to the transverse direction. However, the change of the contour lines with higher density level in (0 0 0 2) pole figures is slight. Therefore, intensities of the basal texture for the alloys are almost the same. The slight variation of the basal texture can be ascribed to the influence of Mg–Zn–Y phases (I-phase and W-phase). It has been reported [29] that recrystallization originates within a zone of high dislocation density and large lattice misorientation at the particles, and proceeds by a rapid polygonization process, resulting in grains with orientations closely related to the as-deformed structure. In Section 3.1, it has explained that during recrystallization process, I-phase or W-phase can create a zone of high dislocation and large lattice misorientation, which influences the orientation of the recrystallized grains, and then indirectly leads to the change of the basal texture.
greatly, and the strength of the ED samples is obvious higher than that of the TD samples, whereas the ductility of the TD samples is superior. Compared with the mechanical properties of the ED samples, it reveals that the alloy with no Y content, the ductility is the highest, but the strength is the lowest. When Y content is 1.08 wt.%, the strength of the alloy is the highest, but the ductility decreases greatly. However, when Y content varies from 1.97 to 3.08 wt.%, the strength shows a tendency to
3.3. The mechanical properties of the alloys The stress–strain curves of the ED and TD samples tested at room temperature are shown in Figs. 4 and 5, respectively. To describe and compare conveniently, the longitudinal and transverse mechanical properties of 0.2% proof yield stress (σ 0.2 ), ultimate tensile strength (UTS), and elongation to failure for the alloys are listed in Table 2. It shows that the variation of Y content can influence the mechanical properties of the alloys
Fig. 4. The stress–strain curves of the ED samples (in Opj format).
252
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
Table 2 The mechanical properties of the as-extruded Mg–Zn–Y–Zr alloys Condition
Alloy I Alloy II Alloy III Alloy IV
Longitudinal direction
Transverse direction
Average grain size (m)
σ 0.2 (MPa)
UTS (MPa)
Elongation (%)
σ 0.2 (MPa)
UTS (MPa)
Elongation (%)
180 200 193 185
316 345 334 324
16.3 10.8 12.3 13.1
128 117 108 72
249 261 226 180
16.4 17.2 14.2 14.6
Fig. 5. The stress–strain curves of the TD samples (in Opj format).
decrease but the ductility has a little increase. In the previous studies [11], it has been suggested that element Y can improve the mechanical properties mainly through refining the grain size and forming I-phase in the Mg matrix. However, when Y content exceeded 1.72 wt.%, the occurrence of W-phase would degrade the strength of the alloys [11]. Based on the Zn/Y ratio [11], when Zn content was about 5.65 wt.%, the alloy with Y content ranged from 1.17 to1.72 wt.%, the Zn/Y ratio was slightly higher than 4.38 and the as-extruded Mg–Zn–Y–Zr alloy could have the highest strength, because element Y mainly existed in the form of I-phase and the volume fraction of icosahedral quasicrystal I-phase reached the peak value. This phenomenon was also reported in Ref. [13]. However, the comparison of the mechanical properties of the TD samples reveals that when Y content is 1.08 wt.%, the alloy can have both the highest ultimate tensile strength (UTS) and the superior ductility. Furthermore, as Y content exceeds 1.08 wt.%, the strength of alloys III and IV decreases greatly. By a comparison of the mechanical properties of the ED and TD samples, it reveals that with Y content increasing, the mechanical anisotropy becomes more and more obvious, as shown in Fig. 6. 3.4. Fracture analysis Fig. 7 shows the fracture surfaces of the ED samples. It reveals that with Y content increasing, the semi-cleavage planes decrease gradually and the plastic dimples increase greatly. Therefore, it indicates that with Y content increasing, the failure modes of the tensile samples will change from ductile–fragile
15 8 8 2–4
Fig. 6. The anisotropy in mechanical properties (ultimate tensile strength (UTS), 0.2% proof yield stress (σ 0.2 ), and elongation to failure) between the longitudinal and transverse directions for the alloys I–IV. (in JPEG format).
failure to ductile failure. Especially for alloy IV, a large amount of plastic dimples and cracked particles (distributed at the bottom of the dimples) can be observed on the fracture surface, which belongs to the ductile failure. In general, the dimple is mainly caused by the existence of inclusions or second phases. It has been reported that the dimple size depends on the inclusion size [30]. Fig. 8 is the backscatter electron image (BEI) of the fracture surfaces of the ED samples. For the TD samples, the secondary electron imaging (SEI) and backscatter electron imaging of the fracture surfaces are shown in Figs. 9 and 10, respectively. Fig. 9 shows that for alloys I–IV, the fractures have the typical “woody fracture” characteristics and the fracture surfaces are covered with a large amount of shallow dimples. Fig. 10 shows that for alloys II–IV, a large amount of zonal distributed Mg–Zn–Y particles can be observed on the fracture surfaces. The microstructure observation (in Fig. 2) clearly shows that after hot extrusion processing, Mg–Zn–Y phases (I-phase and W-phase) are uniformly distributed in the Mg matrix along the extruded direction and the volume fraction of Mg–Zn–Y phases, especially for W-phase increases with Y content. Generally, the interface between the intermetallic particle and the matrix is weak, providing the opening site of a micro-crack as reported in an AZ91 alloy containing the intermetallic particles of Mg17 Al12 [31]. In this study, for alloys II–IV, a large amount of Mg–Zn–Y phase particles are closely spaced and zonal distributed along the extruded direction. Therefore, geometrically necessary dislocations are easily formed around Mg–Zn–Y particles, resulting in decohesion
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
Fig. 7. The secondary SEM fracture images of the ED samples: (a) alloy I, (b) alloy II, (c) alloy III and (d) alloy IV (in JPEG format).
Fig. 8. The backscatter SEM fracture images of the ED samples: (a) alloy I, (b) alloy II, (c) alloy III and (d) alloy IV (in JPEG format).
253
254
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
Fig. 9. The secondary SEM fracture images of the TD samples: (a) alloy I, (b) alloy II, (c) alloy III and (d) alloy IV (in JPEG format).
Fig. 10. The backscatter SEM fracture images of the TD samples: (a) alloy I, (b) alloy II, (c) alloy III and (d) alloy IV (in JPEG format).
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
of the particle/matrix interfaces, especially for W-phase/matrix interfaces, which degrades the tensile strength of the alloys and forms a large amount of dimples on the fracture surface. However, for the TD samples, the spacing and the size of the particles are very small, and the dimple size relies on the particle size; therefore, the dimples on the fractures are very small and shallow, as shown in Fig. 9. 4. Discussion The yield strength of the alloys varies with the grain size and the relationship follows the Hall–Petch equation [32,33]. σy = σ0 + kd −1/2
(1)
where d is the grain size, σ 0 and k are the experimentally derived constants. Based on the equation, alloy IV should have the highest yield strength, because its grain size is the smallest. However, Table 2 shows that the yield strength of alloy IV in both longitudinal and transverse directions is the lowest, whereas the yield strength of alloy II in both directions is the highest. Therefore, besides the grain size, the crystallographic texture and second phase also strongly influence the mechanical properties. However, the (0 0 0 2) pole figures in Fig. 3 show that the main textures of alloys I–IV are all the {0 0 0 1} basal fiber texture. Therefore, the difference of the mechanical properties can be mainly ascribed to Mg–Zn–Y phases (I-phase and W-phase). As Fig. 1 shows, I-phase and W-phase are zonal distributed along the longitudinal direction for alloys II–IV. X-ray analysis indicates that the quantity of W-phase increases with the increasing of Y content. About the influence of Mg–Zn–Y phases on the mechanical properties has also been reported in Refs. [12,17,34], but they mainly focused on the strengthening effect of I-phase. Singh et al. indicated [22] that icosahedral quasicrystal phase could form much better interfaces with Mg-matrix than crystalline phases. It has been reported [11,34] that with the volume fraction of I-phase increasing, the strength of alloys would be improved greatly. For alloy II, the Zn/Y ratio exceeds 4.38 and the main second phase is I-phase (in Fig. 1), so its yield stress (σ 0.2 ) and tensile strength (σ b ) of the ED and TD samples are more superior, with the values of 345 and 261 MPa, respectively. It was also suggested [11] that the existence of W-phase would degrade the mechanical properties of the alloys. For alloys III and IV (with Zn/Y ratio is between 1.10 and 4.38), although the main second phase includes I-phase and W-phase, the quantity of W-phase gradually increases with the increasing of Y content, which results in the strength lower than that of alloy II. In addition, it has been reported [35,36] that the embedment of the second phase particle into the ductile metal matrix can greatly improve their mechanical properties such as stiffness, tensile strength and creep resistance, but simultaneously induce high stress concentrations both with the particle and at the particle/matrix interface, which can trigger voids to nucleate by particle cracking or particle/matrix interface debonding. Fracture analysis (BEI) shows that Mg–Zn–Y particles, especially W-phase distributed on the fracture surfaces increase with Y content greatly, which firmly support the decrease of the strength
255
for alloys III and IV. For the ED samples, with the quantity of Mg–Zn–Y particles, more dimples can be observed on the fracture surface and the ductility of the alloys increases with Y content (listed in Table 2). Since Mg–Zn–Y particles are zonal distributed along the extrusion direction, the quantity of MgZn–Y particles on the transverse cross-sectional plane of the TD samples will be greatly higher than that of the ED samples. Fracture analysis (SEI) shows that for the TD samples, although many dimples can be observed on the fracture surface, the spacing of the dimples decreases with Y contents (in Fig. 9), which leads to the decrease of the ductility and the strength for alloys III and IV (listed in Table 2). In addition, due to the zonal distribution of Mg–Zn–Y particles in the longitudinal direction, the obvious “woody fracture” characteristic occurs on the fracture surfaces for the TD samples, shown in Figs. 9 and 10. By a comparison of the mechanical properties of the ED and TD samples, a very interesting phenomenon can be seen, i.e. the mechanical anisotropy increases greatly with Y content, which is shown in Fig. 6. To describe conveniently, three terms has been defined, i.e. anisotropy in yield stress, anisotropy in ultimate tensile strength (UTS) and anisotropy in elongation, which means how much are the strength and elongation in longitudinal and transverse directions with respect to each other, respectively. It shows that for alloy I (with no Y content), the anisotropy in yield stress and UTS about 29% and 21% respectively, which can be ascribed to the influence of the texture [37]. However, for alloys II–IV, with Y content increasing, the anisotropy in strength increases greatly, especially for the UTS. For alloy IV, the anisotropy in UTS is about 45%. As the discussion mentioned above, the textures of the alloys are almost the same. Therefore, the mechanical anisotropy should be mainly ascribed to the influence of the distribution and quantity of Mg–Zn–Y particles, especially W-phase. Compared with Figs. 8 and 10, it shows that with Y content increasing, the quantity of Mg–Zn–Y particles on the fracture surface of TD samples increases more remarkably than that of the ED samples. Therefore, for the TD samples, with the increase of Y content, more Mg–Zn–Y particles formed on the transverse cross-sectional plane and geometrically necessary dislocations are more closely packed around Mg–Zn–Y particles, especially for W-phase, resulting in decohesion of the particle/matrix interfaces, which greatly degrades the tensile strength of the alloys. Therefore, with Y content increasing, the mechanical anisotropy between the ED and TD increases gradually. In conclusion, although rare earth elements such as element Y can effectively strengthen the Mg–Zn–Zr system alloys, it does not mean that the mechanical properties of the alloy are proportional to Y content. In the previous reports, it suggested that the addition of about 1 wt.% Y could make the as-extruded ZK60 alloy have the superior strength [11,13]. About the explanation of the strengthening mechanism has been reported in detail in Ref. [11]. In this work, with the addition of 1 wt.% Y, the longitudinal and transverse strength of the alloy II has been greatly improved (listed in Table 2). However, the anisotropic strength for alloy II is more obvious than that of alloy I. By a comparison of microstructure, tensile properties and the anisotropy in strength of alloys II to IV, it indicates that I-phase is associated
256
D.K. Xu et al. / Materials Science and Engineering A 443 (2007) 248–256
with finer grain size, higher strength (with comparable grain size) and lesser anisotropy (with comparable texture). Therefore, to reduce the mechanical anisotropy and fully exploit its potential properties, the controlling of the distribution and quantity of the second phases and the crystallographic texture is very important.
Acknowledgement
5. Conclusions
[1] T. Mukai, H. Watababe, K. Higashi, Mater. Sci. Technol. 16 (2000) 1314. [2] H. Somekawa, H. Watababe, T. Mukai, K. Higashi, Scripta Mater. 48 (2003) 1249. [3] H. Watababe, T. Mukai, K. Ishikawa, K. Higashi, Mater Trans. 43 (2002) 78. [4] R.E. Reed Hill, W.D. Robertson, Acta Metall. 5 (1957) 728. [5] T. Mukai, M. Yamanoi, H. Watababe, K. Higashi, Scripta Mater. 45 (2001) 89. [6] H. Watababe, A. Takara, H. Somekawa, T. Mukai, K. Higashi, Scripta Mater. 52 (2005) 449. [7] Y. Yoshida, Cisar, S. Kamado, Y. Kojima, Mater. Trans. 44 (2003) 468. [8] S.R. Agnew, J.A. Horton, T.M. Lillo, D.B. Brown, Scripta Mater. 50 (2004) 377. [9] W.J. Kim, C.W. An, Y.S. Kim, S.I. Hong, Scripta Mater. 47 (2002) 39. [10] A. Singh, A.P. Tsai, Scripta Mater. 49 (2003) 143. [11] D.K. Xu, L. Liu, Y.B. Xu, E.H. Han, J. Alloys Compd. 426 (2006) 115– 161. [12] X. Zeng, Y. Zhang, C. Lu, W. Ding, Y. Wang, Y. Zhu, J. Alloys Compd. 395 (2005) 213. [13] Y. Zhang, X. Zeng, L. Liu, C. Lu, H. Zhou, Q. Li, Y. Zhu, Mater. Sci. Eng. A 373 (2004) 320. [14] Z.P. Luo, D.Y. Song, S.Q. Zhang, J. Alloys Compd. 230 (1995) 109. [15] C. Ma, M. Liu, G. Wu, W. ding, Y. zhu, Mater. Sci. Eng. A 349 (2003) 207. [16] T. Mohri, M. Mabuchi, N. Saitao, M. Nakamura, Mater. Sci. Eng. A 257 (1998) 287. [17] D.H. Bae, S.H. Kim, W.T. Kim, Acta Mater. 50 (2002) 2343. [18] D.K. Xu, L. Liu, Y.B. Xu, E.H. Han, Mater. Sci. Eng. A 420 (2006) 322. [19] J.Y. Lee, D.H. Kim, H.K. Lim, D.H. Kim, Mater. Lett. 59 (2005) 3801. [20] D.H. Bae, Y. Kim, I.J. Kim, Mater. Lett. 60 (2006) 2190. [21] D.K. Xu, L. Liu, Y.B. Xu, E.H. Han, J. Alloys Compd. doi:10.1016/j.jallcom.2006.05.123, in press. [22] A. Singh, M. Watanabe, A. Kato, A.P. Tsai, Philos. Mag. 86 (2006) 951. [23] A. Singh, M. Nakamura, M. Watanabe, A. Kato, A.P. Tsai, Scripta Mater. 49 (2003) 417. [24] A. Singh, M. Watanabe, A. Kato, A.P. Tsai, Mater. Sci. Eng. A 385 (2004) 382. [25] A. Singh, A.P. Tsai, Scripta Mater. 49 (2003) 143. [26] A. Niikura, A.P. Tsai, A. Inoue, T. Masumoto, Philos. Mag. Lett. 69 (1994) 351. [27] F.J. Humphreys, Acta Metall. 27 (1979) 1801. [28] G. Song, D. StJohn, J. Light Met. 2 (2002) 1. [29] F.J. Humphreys, Acta Metall. 25 (1977) 1323. [30] K. Higashi, Y. Hirai, T. Ohnishi, J. Jpn. Inst. Light Met. 35 (1985) 520. [31] Y.Z. Lu, Q.D. Wang, X.Q. Zeng, W.J. Ding, C.Q. Zhai, Y.P. Zhu, Mater. Sci. Eng. A 278 (2000) 66. [32] N.J. Petch, J. Iron Steel Inst. 174 (1953) 25. [33] E.O. Hall, Proc. Phys. Soc. B 64 (1951) 747. [34] E.S. Park, S. Yi, J.B. Ok, D.H. Bae, W.T. Kim, Proceedings of the MRS Fall Meeting, Boston, MA, 2001. [35] Y. Brechet, J.D. Embury, S. Tao, L. Luo, Acta Metall. Mater. 39 (1991) 1781. [36] D.J. Lloyd, Acta Metall. Mater. 39 (1991) 59. [37] S. Kleiner, P.J. Uggowitzer, Mater. Sci. Eng. A 379 (2004) 258.
The present study investigated microstructure and the tensile properties along longitudinal and transverse direction of the as-extruded ZK60 alloys alloyed with different quantities of element Y, which leads to the following conclusions: 1. Microstructure observation shows that with Y content increasing, the grain size of the alloys decreases gradually. When Y content is 1.08 wt.%, the main second phase is Iphase. However, when Y content is 1.97 and 3.08 wt.%, the main second phases are I-phase and W-phase. It firmly indicates that the grain-refining effect of icosahedral quasicrystal phase (I-phase) is superior to that of the cubic W-phase. 2. Through comparing the mechanical properties of the ED samples, it shows that alloy II (with 1.08 wt.% Y) has the superior tensile strength. Since the Zn/Y ratio of alloy II can fully meet the requirement of completely forming I-phases, so the strengthening effect can be mainly ascribed to the strengthening of icosahedral quasicrystal I-phase. 3. The (0 0 0 2) pole figures of alloys I–IV are almost the same, which indicates that the main texture is {0 0 0 1} basal fiber textures. However, by a comparison of the difference of the anisotropy in the mechanical properties for the alloys, it suggests that besides the influence of the texture, the quantity of the zonal distributed Mg–Zn–Y particles, especially W-phase is also a main factor in determining the anisotropy in the mechanical properties of the alloys. Although the anisotropy in the ultimate tensile strength (UTS) of alloy II is more obvious than that of alloy I, both the longitudinal and transverse strength of alloy II has been greatly improved. However, alloy II has less anisotropy in strength than the other two alloys (III and IV), even though these other two alloys have comparable or finer grain sizes and comparable texture. It mainly attributes to the role of I-phase. 4. With Y content increasing, the failure modes of the tensile samples will change from ductile–fragile failure to ductile failure. The fractures of the TD samples were characterized by the typical “woody fracture”, with a large amount of cracked Mg–Zn–Y particles (I-phase and W-phase) distributed at the bottom of dimples. With Y content increasing, the average spacing of the zonal distributed Mg–Zn–Y particles on the fracture surface became narrow, which influenced the transverse mechanical properties greatly.
This work was supported by National Science Fund of China (NSFC) project under grant No. 50431020. References