A. Microstructures of Mg–Zn–Y alloys

Figure 1 shows the XRD patterns of the as-cast experimental alloys. It is observed from Fig. 1 that the five Mg–Zn–Y alloys are mainly composed of α-Mg and Mg₁₂YZn phases, which have been reported in detail in Luo et al. ^{Reference Luo, Tang, Pan, Song and Wang6} Figure 1 also illustrates that the second phase is the LPSO (Mg₁₂YZn) phase rather than W-phase or I-phase regardless of Y/Zn atomic ratio in these alloys. This result differs from that in others studies, which showed that the phase constituent of Mg–Zn–Y alloys closely depends on Y/Zn ratio of alloys. ^{Reference Zhang, Liu, Hu, Li, Le, Bao, Zhu and Cui11,Reference Huang, Liang, Chen and Han14} In addition, the formation mechanism of LPSO in the present Mg–Zn–Y alloys is possibly attributed to that unit cell is composed of six close-packed planes of the magnesium crystal with a stacking sequence is significantly enriched by Zn and Y. ^{Reference Zhu, Morton and Nie16}

FIG. 1.

XRD patterns of the as-cast Mg–Zn–Y alloys: (a) alloy 1; (b) alloy 2; (c) alloy 3; (d) alloy 4; and (e) alloy 5.

The SEM microstructures of the as-cast Mg–Zn–Y alloys are shown in Figs. 2(a)–2(e). From the images it can be seen that the five alloys are mainly composed of a matrix of equiaxed dendrites of α-Mg and secondary phases. Combined with XRD results, the main intermetallic compound in Figs. 2(a)–2(e) is identified as the LPSO phase. It is also shown that the dendrite sizes of the first three alloys orderly decrease as the formation of the LPSO phase increases. This result indicated that different Y/Zn atomic ratios greatly affect the morphological characteristics of α-Mg matrix. Furthermore, the microstructural morphological characteristics of the as-cast alloys are almost similar at different Y/Zn atomic ratios. This finding may be mainly attributed to the formation of the same LPSO phase in the five alloys. Thus, this paper mainly investigated the alloy1, alloy 2, and alloy 3. The actual volume fraction of the LPSO phase of alloys 1, 2, and 3 are roughly 17.51, 21.83, and 22.94%, respectively by using image processing software from microstructure of the alloys. Figures 2(f)–2(h) display the SEM microstructures of the as-extruded Mg–Zn–Y alloys. It suggests that the original block and the lamellar LPSO phase have been destroyed and uniformly distributed in the α-Mg matrix. In addition, grain sizes of as-extruded alloys are considerably smaller than those of their corresponding as-cast alloys. This result illustrates that dynamic recrystallization occurs during hot extrusion. ^{Reference Inoue, Kawamura and Matsushita15} The EDS results of corresponding positions in Figs. 2(a)–2(c) are shown in Table II. In Fig. 2(a), points A1 and A2 correspond to the Mg₁₂YZn phase, and point A3 corresponds to the α-Mg matrix. EDS did not identify specific phases in Figs. 2(b) and 2(c), the compound phases are consistent with the XRD results. To sum up, phase compositions in as-cast alloys are identified as α-Mg and LPSO phase from Figs. 1 and 2, and Table II.

FIG. 2.

Microstructures of as-cast alloys: (a) alloy 1; (b) alloy 2; (c) alloy 3; (d) alloy 4; (e) alloy 5 and as-extruded alloys: (f) alloy 1; (g) alloy 2; (h) alloy 3.

TABLE II.

EDS results of the corresponding positions in Figs. 2(a)–2(c).

To further investigate the LPSO phase structure with different Y/Zn atomic ratios, it is necessary to detect whether the LPSO phase exhibits unique morphological characteristics and type in as-cast Mg–Zn–Y alloys. The bright-field TEM image of the as-cast alloy 2 and the corresponding selected area electron diffraction (SAED) patterns recorded along the $\left[ {11\overline 2 0} \right]$ direction of the selected region are shown in Fig. 3(a). It can be found that there is only one LPSO phase structure. Moreover, the SAED pattern shows periodic extra weak spots at positions n/6, where n is an integer, so the LPSO phase is identified as the 18R phase. The corresponding high resolution electron microscopy (HREM) image in Fig. 3(b) shows a period fringe spacing of 1.6 nm. This 18R structure is adequately discussed in Mg–Zn–Y alloy, reported by Zhu et al. ^{Reference Zhu, Morton and Nie16} In addition, several research results show that the LPSO phase structure of 18R-type is existed in the cast Mg–Zn–Y alloy with a Y/Zn atomic ratio of 2/1 or 4/3. ^{Reference Zhang, Liu, Hu, Li, Le, Bao, Zhu and Cui11,Reference Tang, Wu, Tang, Wang, Peng and Ding17} Therefore, as-cast Mg–Zn–Y alloys with different Y/Zn (2/1, 3/2 or 4/3) atomic ratios exhibit the same stable type of LPSO phase structure, i.e. 18R-type.

FIG. 3.

(a) TEM image and (b) the corresponding HREM image of as-cast alloy 2.

B. Mechanical properties of Mg–Zn–Y alloys

The comparison results of the mechanical properties of the as-cast Mg–Zn–Y alloys with different Y/Zn atomic ratios are demonstrated in Fig. 4(a). It clearly indicates that the as-cast alloys with 3/2 atomic ratio yields the highest mechanical properties, and the other cast alloys cannot be distinguished. And the mechanical properties of the alloy with 3/2 ratio increases as increasing the amount of the LPSO phase. The tensile properties of as-extruded Mg–Zn–Y alloys at room temperature are shown in Fig. 4(b). It can be followed that the as-extruded alloy with 3/2 atomic ratio exhibits good comprehensive mechanical performances, particularly YS, ultimate tensile strength (UTS), and elongation of 233 MPa, 346 MPa, and 16.1%, respectively. And then, in the three studied alloys with a decreasing Y/Zn atomic ratio, its volume fraction of the LPSO phase and tensile strength increase. The YS and UTS of the as-extruded MgZn_1.5Y₂ are raised to 258 and 356 MPa, respectively, but its elongation decrease to 11.7%. Among the alloys, the extruded alloy with 2/1 atomic ratio exhibits the weakest tensile properties.

FIG. 4.

Mechanical properties of Mg–Zn–Y alloys with different Y/Zn atomic ratio: (a) as-cast and (b) as-extruded.

It is clarified in Fig. 4(a) that the mechanical properties of the three as-cast Mg–Zn–Y alloys are not evidently different because these alloys exhibit similar microstructures and display the same type of LPSO phase structure. Compared with the mechanical properties of as-cast alloys, the mechanical properties in as-extruded alloys in Fig. 4(b) are remarkably enhanced. According to the Hall–Petch relation, ^{Reference Zhang, Wang, Du, Liu and Li18} the YS can be enhanced through grain refinement. The existence of the LPSO phase strongly enhances the refinement of Mg matrix grains after extrusion; these fine grains are contributed to improve mechanical properties. In addition, LPSO phases with plate-like shapes are aligned and distributed along the extrusion direction, which could be considered to the fiber strengthening in composite materials. Such LPSO phases likely hinder the movement of dislocations and enhance the strength of the studied alloys. Among the mechanical properties of the three as-extruded alloys, the strength increases is possibly attributed to the volume fraction of the LPSO phase increases, which is usually regarded as second-phase strengthening. What's more, a further change in Y/Zn atomic ratio results in a decrease in elongation owing to the suppression of the dislocation movement.

C. Damping capacities of Mg–Zn–Y alloys

The damping capacities of as-cast Mg–Zn–Y alloys as a function of varying strain amplitudes from 6 × 10⁻⁶ to 3 × 10⁻³ are shown in Fig. 5(a). It is apparent that damping capacities depending on strain amplitude can be divided into two parts. In the first part at low strain amplitude, strain amplitude slightly affects damping value, and the damping capacities of the studied alloys are very low. However, in the second part, the damping capacities of alloys strongly depend on strain amplitude; as strain amplitude exceeds a critical value, damping capacity rapidly increases. Through calculating the average damping value near the turning point of the curve, the corresponding critical strain values of as-cast alloys 2, 3, and 1 are 6 × 10⁻⁴, 7 × 10⁻⁴, and 9 × 10⁻⁴, respectively. Among the studied alloys, the alloys with 3/2 (Y/Zn) atomic ratio yields the highest damping capacity in low and high strain amplitude stages [Fig. 5(a)]; the alloy with 2/1 atomic ratio exhibits the least damping capacity. Additionally, the alloys with 4/3 atomic ratio presents moderate damping capacity. Figure 5(b) illustrates the strain amplitude-dependent damping capacities of as-extruded Mg–Zn–Y alloys at room temperature. This result suggests that damping capacities of as-extruded alloys differ from those of as-cast alloys [Fig. 5(a)]. The alloy with 2/1 atomic ratio exhibits the highest damping capacity in a high-strain amplitude stage; by contrast, the alloy with 3/2 atomic ratio displays a slightly more efficient damping capacity than other alloys in a low-strain amplitude stage. Meanwhile, the alloy with 4/3 atomic ratio presents the lowest damping value in low- and high-strain amplitude stages.

FIG. 5.

Strain amplitude-dependent damping capacities of Mg–Zn–Y alloys at room temperature: (a) as-cast and (b) as-extruded (f = 1 Hz).

Figure 6 shows the relationship between damping capacities and temperature of as-cast and as-extruded Mg–Zn–Y alloys measured at a constant strain of 3 × 10⁻⁴, a heating rate of 3 °C/min, and a vibration frequency of 1 Hz. Figure 6(a) indicates that the presence of a damping peak P in alloy 1 and alloy 3 at 235 and 112 °C, respectively. Two damping peaks are found in the alloy with 3/2 atomic ratio at 108 and 220 °C. These peaks have been commonly observed in Mg alloys with high damping capacities. ^{Reference Hu, Wu and Zheng19} According to some related studies, ^{Reference Hu, Wu, Zheng, Gan and Wang20–Reference Chowdhury, Mari and Schaller22} when the temperature of damping peak is about 100 °C, it is generally called the dislocation peak (P ₁). When the temperature of damping peak is approximately 200 °C, it is named the grain boundary peak (P ₂). Hence, the damping peak of the as-cast alloy with 2/1 atomic ratio is the grain boundary peak, the damping peak of the as-cast alloy with 4/3 atomic ratio is the dislocation peak, and the alloy with 3/2 atomic ratio is characterized as a dislocation peak and grain boundary peak. As can been seen in Fig. 6(a), when temperature is lower than 325 °C, the damping capacities of these Mg–Zn–Y alloys are independent of the temperature, and damping values decrease as Y/Zn atomic ratio decreases. When temperature is higher than 325 °C, the damping capacities of the experimented alloys rapidly increase as temperature increases. Figure 6(b) suggests that the damping capacities of the as-extruded alloys are almost the same, and no damping peaks are detected when temperature is lower than 300 °C. When temperature is higher than 300 °C, the damping value monotonically increases as temperature increases; these as-extruded alloys also possess a high damping background.

FIG. 6.

Temperature-dependent damping capacities of Mg–Zn–Y alloys at ε = 3 × 10⁻⁴ and f = 1 Hz: (a) as-cast and (b) as-extruded.

Damping capacities of as-extruded alloys are lower than those of as-cast alloys [Figs. 5(a) and 5(b)]. Figures 6(a) and 6(b) illustrates that as-extruded Mg–Zn–Y alloys exist a high damping background and do not contain damping peaks, while the as-cast Mg–Zn–Y alloys contain damping peaks.

In general, many factors, including the interactions between dislocations and point defects (impurity atoms, vacancies, and second-phase particles, among others), contribute to damping capacities of Mg and its alloy. Studies on Mg alloys have shown that dislocation damping is the main damping mechanism. ^{Reference Wan, Wang, Wang and Yang23,Reference Jun24} According to the classic Granato–Lücke (G–L) model, ^{Reference Granato and Lucke25,Reference Granato and Lucke26} the damping (Q⁻¹) can be described as follows:

where $Q_0^{ - 1}$ and $Q_{\rm{h}}^{ - 1}$ are strain-amplitude independent damping and dependent damping, respectively; ρ and L _C are the dislocation density and the average distance between weak pining points, respectively; F _B, L _N, E, and b are the binding force between dislocations and solute atoms, the average distance between strong pining points, the elastic modulus, and the Burgers vector, respectively. Then, F _B/bE slightly changes as Y/Zn atomic ratio decreases due to the unique phase composition and similar grain sizes in alloys. Hence, Eq. (3) can be altered and simplified as follows:

The values of the parameters of Mg–Zn–Y alloys according to Eq. (5) are shown in Table III. It suggests obviously that the as-cast alloy with 3/2 atomic ratio owns a higher value of C₁ and a lower value of C₂ than other alloys. Meanwhile, C₂ is related to the weak points mainly influenced by solid solution atoms from Eq. (6). Hence, the as-cast alloy with 3/2 atomic ratio possesses the fewest amounts of solid solution atoms. As illustrated in Fig. 7, the number of Y atoms decreases and the number of Zn atoms increases in a matrix with decreasing Y/Zn atomic ratios. Moreover, the alloy with 3/2 atomic ratio relatively contains the fewest solute atoms and the largest L _C because Y and Zn atoms are combined and placed in proximity to grain boundaries. Combined with the TEM images, as-cast Mg–Zn–Y alloys with different Y/Zn atomic ratio are composed of the same type of LPSO phase structure. Therefore, it can be assumed that the impurity atoms are possibly the main factor affecting the damping capacity of the studied alloys. As discussed above, it is demonstrated in Fig. 5(a) that the as-cast alloy with 3/2 atomic ratio exhibits the better damping value in low- and high-strain amplitude stages because of a larger L _C value, a higher value of C₁ and a lower value of C₂ than other alloys. In addition, the effect of LPSO phase on damping capacities of Mg–Zn–Y alloys may be explained by the following aspects. For one thing, the formation of LPSO phase of Mg–Zn–Y alloys cause a decrease in the amount of solute atoms and add to the interaction between the dislocation and point obstacles, which maybe induce friction in releasing additional energy, thereby contributing positive effects on the internal friction of the alloys. For another, the difference of thermal expansion coefficient between LPSO phase and α-Mg matrix may cause the generation of long and parallel dislocations from our previous study. ^{Reference Wang, Lu, Wei, Huang and Pan10} These long dislocations are likely to move in response to stress and thus play a key role in energy dissipation in the damping process. What's more, the LPSO phase, which also maybe resembles a rough fiber, shows more interfaces with the matrix and moves back and forth more easily to comply with the vibration stress. The previous studies ^{Reference Wang, Lu, Wei, Huang and Pan10,Reference Qin, Wang, Chen, Lu and Pan13} show that the LPSO phase structure is also conducive to damping capacities, which can explain the damping capacity of the alloy with 4/3 atomic ratio is close to that of alloy 2 due to the slightly increase of the content of LPSO phase. Moreover, the lower damping capacity of as-extruded alloys is attributed to the dislocation tangles formed during extrusion. From Fig. 5(b), the as-extruded alloy 1 exhibited the highest internal friction at the high-strain amplitude stage possibly due to the fewest strong pinning points by grain boundaries and second phase.

TABLE III.

Values of the parameters of Mg–Zn–Y alloys with different Y/Zn atomic ratios.

FIG. 7.

Schematic illustration of the solid solution atoms distribution in Mg–Zn–Y alloys with different Y/Zn atomic ratios: (a) 2/1; (b) 3/2; and (c) 4/3.

The damping peaks of the as-cast Mg–Zn–Y alloys as shown in Fig. 6(a) are similar to those detected in other as-cast Mg alloys. ^{Reference Hu, Zhang, Zheng and Wu27} The grain boundary peak of the as-cast alloy with 2/1 atomic ratio is attributed to the aggregation of fewer atoms at grain boundaries that alloy easy sliding at a low vibration strain of 3 × 10⁻⁴ at this temperature. The dislocation peak of the as-cast alloy with 4/3 atomic ratio is attributed to the free movement of dislocations in α-Mg phase while the more second phase aggregates at grain boundaries and make them difficult to slide. The as-cast alloy with 3/2 atomic ratio displays P₁ and P₂ damping peaks due to few solid solution atoms in Mg matrix and grain boundaries. This result is well coincided with previously discussed findings. Damping peaks in as-extruded alloys from Fig. 6(b) are not observed possibly due to the following two reasons. On the one hand, the severely broken LPSO phase structure impedes the movement of dislocation after extrusion, which suppressing the appearance of peak P1. On the other hand, the obvious reduction of grain size after extrusion greatly deteriorates the slipping of grain boundary, which hampering the appearance of P2 peak.

The comprehensive comparison results of the damping capacities and mechanical properties of alloys prepared at different Y/Zn atomic ratio is demonstrated in Fig. 8. It suggests that the studied alloy with 3/2 atomic ratio exhibits excellent comprehensive properties both in as-cast and as-extruded states, which possesses a UTS of 346 MPa and still remains a certain damping capacity (Q⁻¹ > 0.01) in extrusion. Indeed, the usually contradiction of damping capacities and mechanical properties is possibly addressed. This result may be attributed to the following reasons. On the one hand, the alloy with 3/2 atomic ratio contains less solute atoms, which shows excellent internal friction according to G–L theory. On the other hand, the alloy with 3/2 ratio has the middle content of LPSO phase, the strength increases is possibly attributed to the volume fraction of the LPSO phase increases, which is usually regarded as second-phase strengthening. The elongation of alloys decreases owing to the increase of LPSO phase contents suppress the dislocation movement. So the alloy with 3/2 ratio exhibited good comprehensive mechanical properties. In addition, the formation of LPSO phase of Mg–Zn–Y alloys cause a decrease in the amount of solute atoms and add to the interaction between the dislocation and point obstacles, which maybe induce friction in releasing additional energy, thereby contributing positive effects on the internal friction of the alloys. And the alloy with 3/2 ratio possesses near volume fraction of the LPSO phase of the alloy with 4/3 alloy, which shows the close strength. Consequently, the studied alloy with 3/2 atomic ratio exhibited excellent comprehensive damping and mechanical properties. However, the two other alloys exhibit relatively poor comprehensive performances. To sum up, tailoring of the Y/Zn atomic ratios in Mg–Zn–Y alloys may be an important approach for obtaining the balanced optimization of damping and mechanical properties. It is potentially hoped that the results will contribute to the basis of high strength and high damping structural material designs of magnesium alloys for antivibration and noise-reduction applications.

FIG. 8.

Comprehensive performances of Mg–Zn–Y alloys with different Y/Zn atomic ratios.