Microstructure Evolution and Local Hardness of Mg–Y–Zn Alloys Processed by ECAE

Motohiro Yuasa; Ryoichi Sato; Takao Hoshino; Daisuke Ando; Yoshikazu Todaka; Hiroyuki Miyamoto; Hidetoshi Somekawa

doi:10.2320/matertrans.MT-MD2022018

Abstract

Mg–9 at%Y–6 at%Zn and Mg–2 at%Y–1 at%Zn alloys were processed by equal-channel-angular extrusion (ECAE) to investigate their microstructure evolution and local hardness. The area fraction of the kink bands in the Mg–9 at%Y–6 at%Zn alloys increased with increasing the number of ECAE passes, resulting in higher hardness. In contrast, the number of kink boundaries in the local region near the indentation was almost constant. The relationship between the microstructure factors of the kink bands and the local hardness is discussed in comparison with the forged alloy. In the Mg–2 at%Y–1 at%Zn alloys, the microstructural evolution of the α-Mg matrix phase and long-period stacking ordered (LPSO) phase by 1-pass ECAE and the increase in local hardness were discussed.

1. Introduction

Mg–Y–Zn alloys containing the long-period stacking ordered (LPSO) phase have attracted considerable attention because of their excellent mechanical properties.¹^–⁷⁾ These properties have been attributed to the deformation kink bands introduced into LPSO phases during thermomechanical wrought processing.⁶^–⁹⁾ To elucidate the relationship between deformation kink bands and mechanical properties of the LPSO phase, the plastic deformation behavior of directionally solidified LPSO single-phase specimens have been systematically investigated.⁴^,¹⁰^–¹⁶⁾ As a result, deformation kink bands can be induced in directionally solidified specimens by applying stress/strain normal to the (0001) close-packed plane.¹²⁾ A deformation kink band acts as an effective barrier to the movement of dislocations, resulting in the strengthening of the LPSO phase (= kink-band strengthening).¹³^,¹⁴⁾ The formation process and three-dimensional features of deformation kink bands have been also demonstrated.¹⁵^,¹⁶⁾

Recently, the effects of microstructural factors related to kink bands on the mechanical properties of LPSO phases have been investigated by performing micro-Vickers hardness tests on LPSO phases with kink bands formed by various wrought-processing.¹⁷^–²³⁾ The advantage of micro-Vickers hardness tests is that the hardness of target areas (microstructures) can be measured by changing the indentation load. In the caliber-rolled Mg–5 at%Y–2.5 at%Zn alloy, the average bending angle of the kink bands is larger, and the hardness of the specimens is higher.¹⁷⁾ The difference in local hardness (i.e. the hardness of the regions with kink bands subtracted from those without kink bands) is closely related to the number of kink boundaries beneath the indentation in the forged Mg–9 at%Y–6 at%Zn alloys.¹⁹⁾

Our previous study investigated the relationship between the area fraction of kink bands and strain distribution induced by various wrought-processing using numerical and experimental methods.²³⁾ The study has suggested that shear strain is one factor that effectively induces kink bands from a processing viewpoint. In the present study, we focused on equal-channel-angular extrusion (ECAE), which applies high-shear stress/strain to specimens. The effects of microstructural factors on the hardness, including that of the morphology of kink bands formed into Mg–9 at%Y–6 at%Zn and Mg–2 at%Y–1 at%Zn alloys during ECAE, were investigated by changing the indentation load in the micro-Vickers hardness tests. The microstructural factors included the area fraction of the kink bands and the number of kink boundaries formed by ECAE for the Mg–9 at%Y–6 at%Zn alloys. In the Mg–2 at%Y–1 at%Zn alloys, the factors were very complicated because of the different plastic deformation responses of the α-Mg matrix and LPSO phases. Thus, the local hardness of the α-Mg matrix phase and LPSO phase after only 1-pass of ECAE was measured to determine how each phase was strengthened.

2. Experimental Procedure

Mg–9 at%Y–6 at%Zn alloy containing almost 100% LPSO phase and Mg–2 at%Y–1 at%Zn alloy with approximately 25% volume fraction of the LPSO phase were used in this study. These two alloys were machined into cylinder bars of 10 mm diameter and 70 mm length, then subjected to ECAE at 623 K with a ram speed of 1 mm/min. The channel angle (Φ) and corner curvature angle (Ψ) are 110° and 20°, respectively. Using this ECAE die, an equivalent plastic strain of 0.76 was introduced into the alloys by a single pass. The microstructure and hardness of both the ECAEed alloys were evaluated in the transverse direction (TD) plane.

In the Mg–9 at%Y–6 at%Zn alloy, ECAE was carried out for three passes using route B_c, in which the specimen was removed from the die and rotated by 90° between each pass.²⁴⁾ The microstructure was observed using a laser microscope, and the area fraction of the deformation kink bands was measured using the point-counting method.²⁵⁾ Vertical and horizontal lines were drawn with a space of 25 µm on the laser microscopy images, and the number of kink bands at the cross point was counted. The specimens for microstructure observation were prepared by grinding with SiC paper and polishing with diamond suspensions (up to 1 µm). The polished surfaces were then etched with nital for several seconds. The hardness of the ECAEed Mg–9 at%Y–6 at%Zn alloy was evaluated by micro-Vickers hardness tests under indentation loads of 300 gf and 10 gf. Using the local hardness with an indentation load of 10 gf, the magnitude of kink-band strengthening was estimated by subtracting the hardness in the area where kink bands were observed with laser microscopy from that in the area where kink bands were not observed.¹⁹^,²¹⁾

In the Mg–2 at%Y–1 at%Zn alloy, ECAE was performed for only one pass to investigate the microstructure formed by a simple shear of ECAE and its local hardness. The microstructure was observed using scanning electron microscopy (SEM), electron backscatter diffraction (EBSD), and transmission electron microscopy (TEM). Acceleration voltages for SEM(-EBSD) and TEM are 15 kV and 200 kV, respectively. The preparation for the SEM-EBSD observation was the same as in the case of the Mg–9 at%Y–6 at%Zn alloy (polishing with 1 µm diamond suspensions at final polishing). A cross-section polisher was then used for final preparation of the EBSD specimens at an average voltage of 6 kV. Specimens for the TEM observations were prepared by polishing to a thickness of approximately 100 µm and then thinning by ion milling. The local hardness of the Mg and LPSO phases processed by ECAE was measured using micro-Vickers hardness tests under an indentation load of 10 gf. Micro-Vickers hardness tests were performed more than 15 times for the Mg–9 at%Y–6 at%Zn and Mg–2 at%Y–1 at%Zn alloys.

3. Results and Discussion

3.1 Mg–9 at%Y–6 at%Zn alloy

Figure 1 shows the microstructure of the ECAEed Mg–9 at%Y–6 at%Zn alloys obtained by laser microscopy. Deformation kink bands were observed, as indicated by the black arrows, and the area fraction tended to increase with increasing the number of ECAE passes. The area fractions of the deformation kink bands estimated by the point-counting method were 8.2%, 10.2%, and 12.1% for the 1-pass, 2-pass, and 3-pass alloys, respectively.

Fig. 1

Microstructure of the ECAEed Mg–9 at%Y–6 at%Zn alloys obtained by laser microcopy: (a) 1 pass, (b) 2 passes, and (c) 3 passes. The black arrows indicate kink bands.

The relationship between the hardness measured with an indentation load of 300 gf and the area fraction of the kink bands is shown in Fig. 2. The results of the forged Mg–9 at%Y–6 at%Zn alloys in our previous study²²^,²³⁾ are also plotted for comparison. The hardness tended to increase with increasing area fraction of the kink bands, independent of the processing methods. Note that both the hardness and area fraction of kink bands in the ECAEed alloys were higher than those of the forged alloys. In our previous work, there was a difference in the hardness of alloys processed using different processing methods.²²⁾ The results shown in Fig. 2 demonstrate that the difference in hardness among the different processing methods can be attributed to the difference in kink band formation.

Fig. 2

Relationship between the hardness and the area fraction of kink bands in the Mg–9 at%Y–6 at%Zn alloys.

As shown in Fig. 2, the “macroscopic” hardness with an indentation load of 300 gf was strongly related to the area fraction of kink bands in the ECAEed and forged Mg–9 at%Y–6 at%Zn alloys. In other words, the higher hardness of ECAEed alloys is attributed to a higher area fraction of kink bands introduced by great shear strain through ECAE compared with that of forged alloys. As mentioned in the introduction, kink-band strengthening depends on the number of kink boundaries within a local area near an indentation in the forged Mg–9 at%Y–6 at%Zn alloy.¹⁹⁾ The magnitude of kink-band strengthening is measured by the difference in local hardness: ΔH (= H_kink − H_no-kink) obtained by subtracting the hardness in the area where kink bands were observed by laser microscopy (H_kink) from that in the area where kink bands were not observed (H_no-kink). The number of kink boundaries was counted in a local region of 10.8 µm × 10.8 µm near an indentation with a load of 10 gf. Thus, we investigated whether the strong shear strain induced by ECAE affects the number of kink boundaries in the local area and local hardness.

Figure 3 shows the difference in local hardness as a function of the number of kink boundaries beneath the indentation in the ECAEed Mg–9 at%Y–6 at%Zn alloys. The number of kink boundaries in the local area was almost constant, as the number of ECAE passes increased. For comparison, the results for Mg–9 at%Y–6 at%Zn alloys forged at 623 K–698 K are also shown in Fig. 3.¹⁹⁾ It should be noted that the difference in local hardness increased with the number of kink boundaries, irrespective of whether the processing method was ECAE or forging. This is consistent with previous results that kink-band strengthening is related to the number of kink bands, irrespective of the processing method.¹⁹^,²¹⁾

Fig. 3

Relationship between the difference in local hardness and the number of kink boundaries in the Mg–9 at%Y–6 at%Zn alloys.

The area fraction of kink bands by point-counting method and “macroscopic” hardness with an indentation load of 300 gf were higher in the ECAEed alloys than in the forged alloys, as shown in Fig. 2. However, the difference in local hardness and the number of kink boundaries in the ECAEed alloys were almost comparable with those in forged alloys, as shown in Fig. 3. Our previous work on kink-band strengthening via various wrought-processing techniques has suggested that multiple shear strains are effective in improving the mechanical properties.²²⁾ In the present study, ECAE was performed up to three passes using route B_c, and thus multiple shear strains were induced into the specimens. This suggests that the shear strain induced by ECAE is effective for increasing the area fraction of kink bands in the entire specimen (related to the macroscopic hardness), but is not effective for a higher density of kink boundaries in the local region (related to the local hardness).

3.2 Mg–2 at%Y–1 at%Zn alloy

As described in the previous section, micro-Vickers hardness tests with an indentation load of 10 gf enabled us to investigate the local hardness of different microstructures. Thus, we focused on ECAEed Mg–2 at%Y–1 at%Zn alloys consisting of an α-Mg matrix phase and an LPSO phase.

Figure 4 shows the microstructure of the as-cast and ECAEed Mg–2 at%Y–1 at%Zn alloys by SEM. The dark and bright regions correspond to the α-Mg and LPSO phases, respectively. The block and lamellar-shaped LPSO phases in the as-cast alloy were deformed and broken after 1-pass of ECAE. Focusing on the LPSO phase, the microstructure of the ECAEed alloys was divided into three regions: a non-deformed region, a with-kink region, and a slipped region, as indicated by black, white, and red arrows, respectively (Fig. 4(b)). The LPSO phases in the non-deformed region maintained the block-shaped LPSO phase observed in the as-cast alloy. In the with-kink region, the LPSO phases were kink-deformed. In the slipped region, the lamellar shape of the LPSO phases in the as-cast alloys broke to a plate-like shape along the shear direction of the ECAE.

Fig. 4

SEM images of Mg–2 at%Y–1 at%Zn alloys: (a) as-cast, (b) 1-pass ECAEed alloys. The black, white, and red arrows indicate a non-deformed region, a with-kink region, and a slipped region, respectively.

Figure 5 shows the local hardness of the α-Mg and LPSO phases in the non-deformed, with-kink, and slipped regions. The local hardness of the as-cast alloy is presented for comparison (Fig. 5). Note that the local hardness of the LPSO phase is defined as the region containing more than 50% of the LPSO phase beneath the indentation because it is difficult to indent only the LPSO phase even with an indentation load of 10 gf in the Mg–2 at%Y–1 at%Zn alloy, where the volume fraction of the LPSO phase is approximately 25% (Fig. 5(b), (c)). The local hardness in the non-deformed region was higher than that of the as-cast alloy. This may be due to the plastic deformation response (i.e. increase in dislocation density) associated with the material flow induced by ECAE.

Fig. 5

Local hardness of the α-Mg and LPSO phases in the Mg–2 at%Y–1 at%Zn alloys: (a) summarized the results of micro-Vickers hardness tests with an indentation load of 10 gf. The examples of an indentation of (b) Mg phase and (c) LPSO phase.

Notably, the local hardness of the α-Mg phase in the with-kink and slipped regions was higher than that in the non-deformed region, as indicated by the dashed line in Fig. 5(a). Figure 6(a) shows the magnified microstructures of the α-Mg phases in the with-kink region. Deformation kink bands were observed in the α-Mg phase near the kink-deformed LPSO phase, as indicated by the black arrows. The contrast of the α-Mg phase, where kink bands were observed, was brighter than that of the phase where kink bands were not observed, indicating that the brighter area contained Y and/or Zn. Hagihara et al.²⁶⁾ have reported that kink band formation is induced in the Mg–0.6 at%Y–0.2 at%Zn single crystal (Mg matrix solid solution) containing a high-density LPSO nanoplate, resulting in yield stress comparable to that of LPSO single-phase alloys. Thus, the higher local hardness in the with-kink region than in the non-deformed region can be attributed to the kink-deformed α-Mg phase containing thin plate-shaped LPSO phases. In the slipped region, α-Mg grains approximately 2 µm in size were observed, as shown in Fig. 6(b). The fine α-Mg grains could be responsible for the higher local hardness of the α-Mg phase in the slipped region than in the non-deformed region.

Fig. 6

Magnified microstructures in the ECAEed Mg–2 at%Y–1 at%Zn alloys: (a) SEM image in the with-kink region, showing the deformation kink bands in the α-Mg phase containing thin plate LPSO phase as indicated by the black arrows, and (b) IPF map of α-Mg phase in the slipped region obtained by EBSD method. The black contrast corresponds to the LPSO phase. The white arrows indicate the dynamic-recrystallized α-Mg grains.

As indicated by the chain line in Fig. 5(a), the LPSO phase in the slipped region exhibited almost the same (rather slightly higher) hardness as the LPSO phase in the with-kink region, even though no kink bands were observed in the slipped region, as indicated by the red arrows in Fig. 4(b). The magnified SEM and TEM images are shown in Fig. 7. As indicated by the black arrows in Fig. 7(a), the kink-deformed LPSO phases with low rotation angle exist. In the TEM image (Fig. 7(b)), some kink boundaries are observed in the kink-deformed LPSO phase. Hagihara et al.²⁷⁾ have proposed that a kink-band boundary is regarded as a hierarchical structure, in which many small kink bands accumulate in a local region, resulting in preventing the motion of dislocations to the same degree as a general grain boundary. The TEM image in Fig. 7(b) also shows the multiple kink boundaries at a distance of a few µm in the kink-deformed LPSO phase. These results suggest that these kink boundaries act as obstacles to the motion of basal dislocations,¹³^,¹⁴^,²⁷^,²⁸⁾ leading to a local hardness in the slipped region equivalent to that in the with-kink region.

Fig. 7

Deformed microstructures in the slipped region of the ECAEed Mg–2 at%Y–1 at%Zn alloys: (a) SEM image showing the kink deformed LPSO phases with low rotation angle as indicated by the black arrows, and (b) TEM image of the kink deformed LPSO phases with low rotation angle. The white arrows indicate the kink boundaries.

4. Conclusion

In this study, Mg–9 at%Y–6 at%Zn and Mg–2 at%Y–1 at%Zn alloys were processed by equal-channel-angular extrusion by applying strong shear stress/strain to the specimens. The relationship between the microstructure and hardness was evaluated. In the Mg–9 at%Y–6 at%Zn alloys, the increase in the area fraction of the kink bands corresponded to an increase in the number of ECAE passes. Compared to the forged alloys, the higher hardness of the ECAEed alloys is attributed to the area fraction of the kink bands. The magnitude of the increase in local hardness related to kink-band strengthening was similar to that of the forged alloys despite the strong shear strain of ECAE. The microstructure of the 1-pass ECAEed Mg–2 at%Y–1 at%Zn alloys was divided into three regions: non-deformed, with kink, and slipped regions. The mechanism of the increase in the hardness of the α-Mg matrix and LPSO phases in the non-deformed, with-kink, and slipped regions was investigated using SEM-EBSD and TEM.

Acknowledgments

This work was supported by the JSPS Scientific Research on Innovative Areas “MFS Materials Science in Nos. 18H05475 and 18H05477”.

REFERENCES

Corresponding author

Register with J-STAGE for free!