Hot Rolling of the Twin-Roll Cast and Homogenized Mg-6.8Y-2.5Zn (WZ73) Magnesium Alloy Containing LPSO Structures

Abstract

In this study, hot rolling trials were conducted on a twin-roll cast and homogenized magnesium alloy Mg-6.8Y-2.5Zn (WZ73). The WZ73 contains long period stacking ordered (LPSO) structures due to the ratio of Y and Zn content. Microstructural and texture evolution depending on the different strain and strain rates were investigated, and the resulting mechanical properties were also considered. Therefore, samples were quenched in water directly after hot rolling. The results revealed that the rolling parameters significantly influence dynamic recrystallization (DRX), while continuous and twin-induced dynamic recrystallization are assumed to be the main DRX mechanisms. It was also found that high strains and strain rates are required to proceed the DRX. The resulting textures revealed that the non-basal slip of <$a$>-dislocations and <$c+a$>-dislocations is activated during hot rolling. Hot rolling results in increased strength and ductility compared to the initial twin-roll cast and homogenized state.

1. Introduction

The alloy Mg-6.8Y-2.5Zn (WZ73) belongs to those alloy systems that have special long period stacking ordered (LPSO) structures as a result of the combination of Y and Zn. LPSO structures are chemically and structurally ordered. Y and Zn atoms occupy positions on the {0001} basal plane of Mg, but the LPSO structures are stacking-ordered along the c-axis. This results in stacking periods. The 14H and 18R structures are the most frequently reported. In magnesium wrought alloys, the LPSO phases are reported to occur as bulk-shaped structures at the grain boundaries or precipitate along the basal plane of the matrix as lamellar structures [1]. The production and processing of alloys containing LPSO structures are mainly carried out by conventional casting processes and subsequent extrusion. In some studies, hot rolling was also investigated [2,3,4]. Xu et al. (2012) [4] suggested that sheets of the Mg-8.2Gd-3.8Y-1.0Zn-0.4Zr alloy exhibit an almost fully recrystallized microstructure after hot rolling with a total reduction of 96% and good mechanical properties, with an ultimate tensile strength of 403 MPa, 0.2% proof stress of 318 MPa, and an elongation at fracture of 13.7%. Dynamic recrystallization (DRX) required a high amount of deformation. Total reductions of 60% or 73% result only in incipient DRX. However, due to the high amount of rare earth elements (RE), plastic deformation is difficult and hot rolling of LPSO containing magnesium alloys is performed using small rolling reductions per pass and high processing temperatures [4,5,6]. In contrast, however, high strains and strain rates during hot deformation are favorable for a sufficiently high DRX, but lead to edge cracking during hot rolling [7]. Li et al. (2019) [8] showed that during hot rolling of a Mg-13Gd-4Y-2Zn-0.6Zr alloy, dynamically recrystallized grains spread from the grain boundaries into the coarse deformed grains. Several studies revealed that different DRX mechanisms occur during the hot deformation of LPSO containing magnesium alloys. Beside the formation of new grains along the original grain boundaries as a result of continuous dynamic recrystallization (CDRX), DRX at twin (TDRX) and kink (KDRX) boundaries provide important recrystallization mechanisms [8,9,10,11,12,13].

The hot deformation behavior and mechanisms occurring in LPSO containing magnesium alloys are widely studied. Wang et al. (2019) [14] found basal and prismatic slip to be the dominating initial plastic deformation mechanism during hard plate rolling of the Mg-8.8Gd-3.4Y-1Zn-0.8Mn magnesium alloy. In pure magnesium, basal slip and $\left\{10\overline{1}2\right\}$-twinning are the most common deformation mechanisms. Several studies reveal that those common deformation mechanisms are partially inhibited by the LPSO structures, and other deformation mechanisms became the dominant ones. Matsuda et al. (2005) [15] showed that grains without LPSO phases exhibit straight <$a$> dislocations, while grains with LPSO structures possess <$c+a$> dislocations. Therefore, it can be assumed that LPSO structures promote the activation of non-basal slip, while the basal slip is suppressed. The occurrence of twins in magnesium alloys with LPSO structures is also frequently discussed. Some studies report few or no twins [5,16,17,18], while Garces et al. (2018) [19] continue to report twinning as the dominant mechanism during hot deformation. Shao et al. (2021) [20] investigated the interaction between twinning and LPSO plates during hot rolling and found that LPSO plates cannot suppress $\left\{10\overline{1}2\right\}$-twinning in coarse grains, which exhibit a favorable orientation for twinning. However, lamellar LPSO structures seem to slow the twin growth rate. The inhibition effect of the lamellar LPSO structures depends on the lamella thickness of the magnesium matrix between adjacent LPSO lamellas. Besides this, kink deformation, which is not a common deformation mode in pure magnesium, was observed frequently in magnesium alloys containing LPSO structures [7,21,22,23].

Hot rolling as an industrial processing method enables refinement of the grain structure, the fragmentation of second phases, and enhancement of the strength of magnesium alloy sheets. In order to produce magnesium sheets in an energy-efficient way, the twin-roll casting process became important to provide near-net-shaped sheets as the initial material for hot rolling. Recently, twin-roll casting (TRC) of the WZ73 alloy containing LPSO structures was introduced [24]. The twin-roll cast WZ73 alloy consists of the magnesium matrix and network-shaped LPSO structures, which are mainly located at the grain boundaries. Generally, TRC leads to the formation of a characteristic microstructure with columnar dendrites growing from the surface to the mid-thickness of the strip, as it is reported for AZ31 or ZAX210. In this work, hot rolling considering different rolling parameters (strain and strain rate per pass) of a twin-roll cast WZ73 alloy is presented. Microstructure and texture evolution are investigated in order to reveal the dominant deformation and recrystallization mechanisms. Mechanical properties of the as-rolled sheets were determined by tensile testing at room temperature.

2. Materials and Methods

Twin-roll cast sheets of a Mg-6.8Y-2.5Zn (WZ73) magnesium alloy, previously produced at the Institute of Metal Forming (IMF, hpl-Neugnadenfelder Maschinenfabrik GmbH, Neugnadenfeld, Germany) [24], were used for the investigation of this work.

Table 1. Chemical composition of WZ73 (wt. %) alloy as determined by means of optical emission spectrometry (OES).

Y	Zn	Zr	Si	Fe	Cu	Ni	Others	Mg
6.8	2.5	0.4	0.01	0.005	0.001	0.001	0.01	Balance

presents the chemical composition of the sheets after twin-roll casting measured by optical emission spectrometry (OES, SPECTRO Analytical Instruments GmbH, Kleve, Germany).

After TRC, sheets with a width of 280 mm were cut from the TRC plates and homogenized at 500 °C for 2 h. Immediately after heat treatment, the sheets with an initial thickness of 5.5 mm were hot rolled using a reversing rolling mill at the laboratory scale. The rolls with a diameter of 360 mm were preheated to 120 °C using heating mats in order to reduce temperature losses. The starting temperature of hot rolling was 480 °C, while strain and rolling speed varied according to

Table 2. Parameters for the hot rolling of the twin-roll cast and homogenized WZ73.

Parameter of Hot Rolling	Parameter Range
equivalent strain	0.43|0.52|0.63|0.75
rolling speed in m/s	1|1.5|2

. For the maximum and the minimum rolling speed, all strains were applied. For 1.5 m/s, only equivalent strains of 0.52 and 0.63 were tested in order to obtain mean equivalent strain rates for comparison. The equivalent strain (φ_v) and the resulting equivalent strain rates (${\dot{\phi }}_v$) were calculated according to Equations (1) and (2), respectively, where h is the final sheet thickness after one pass, h₀ is the initial sheet thickness, $v_{HR}$ is the rolling speed, and $l_d$ is the contact length.

$${\phi }_v=\frac{2}{\sqrt{3}}·ln\frac{h}{h_0}$$

(1)

$${\dot{\phi }}_v=\frac{v_{HR}}{l_d}·{\phi }_v$$

(2)

Immediately after hot rolling, the sheets were quenched in water to suppress softening resulting from static recrystallization processes. For microstructural and texture characterization, small samples were cut from the sheets and metallographically prepared by grinding and polishing. For etching, a two-step procedure was applied: The sample was poured into a solution of nitric acid for 5 s and then into picric acid for 7 s. Optical microscopy was performed by the Keyence VHX 6000 microscope at the IMF, Freiberg, Germany. The micrographs were taken between the upper edge and the mid-thickness of the strip. A scanning electron microscopic (SEM) evaluation was performed on a ZEISS GeminiSEM 450 device at the IMF, Freiberg, Germany. Texture analysis was performed by electron backscatter diffraction (EBSD) analysis using a ZEISS GeminiSEM 450 at the IMF. EBSD was conducted with a voltage of 10 kV and a step size of 0.5 µm. AZtec data collection software (version 5.0, Oxford instruments, Abingdon, UK) was used for the processing of all orientation maps. The free MTEX MATLAB toolbox (version 5.7.0, MTEX, Ralf Hielscher, Chemnitz, Germany) [25] was used for the analysis of the EBSD data and the calculation of the pole figures of the TRC, both the homogenized and the as-rolled states. The fraction of DRXed grains was determined via EBSD. The selection of the recrystallized grains was based on two criteria: (1) The misorientation within the grain must not exceed 2° and (2) the recrystallized grain must be surrounded by large-angle grain boundaries (>15°). For tensile testing at room temperature, samples according to DIN 50125 shape H with a measured length of 80 mm were cut from the as-rolled sheets. Samples were taken parallel (0°) and transverse to the rolling direction (90°). For statistical validation, five samples of the same conditions were tested. The testing speed was 2 m/min.

3. Results and Discussion

3.1. Initial Material (Twin-Roll Cast and Homogenized WZ73 Alloy)

The microstructure of the twin-roll cast and homogenized (500 °C, 2 h) WZ73 magnesium alloy is presented in Figure 1. The microstructure consists of the magnesium matrix, which is permeated by fine, network-like LPSO phases. Previous studies revealed that the LPSO phases belong to 18R and 14H structures [24,26]. During the heat treatment at temperatures above 500 °C, the network-like LPSO phases can partially be dispersed and the alloying elements Y and Zn are dissolved in the magnesium matrix [24]. The chemical composition of the magnesium matrix was measured via EDX analysis and consisted of 3.17 wt.-% Y and 0.94 wt.-% Zn. Compared to the TRC condition, the amount of Y and Zn increased (2.2 wt.-% Y and 0.7 wt.-% Zn [24]).

Results of the texture analysis via EBSD of the twin-roll cast and homogenized WZ73 alloy were presented in [24]. The strip exhibits a basal texture with maximum intensities (1.7 mrd) tilted away from the core intensity. The c-axis of most of the crystals is oriented between 30° and 60° with regard to the sheet plane. Compared to the TRC condition [24], the heat treatment results in a slight texture weakening. Xu et al. (2017) [5] also reported a texture weakening after heat treatment in the LPSO containing the Mg-Gd-Y-Zn-Zr alloy. The WZ73 alloy offers moderate strength values of 195 MPa ultimate tensile strength and 136 MPa 0.2% proof stress and low elongation at fracture of about 3.6% in the twin-roll cast and homogenized state at room temperature (0°). Comparable mechanical properties for an as-cast Mg₉₇Y₂Zn₁ alloy was reported by Chen et al. (2014) [27].

3.2. Hot Rolling

The results of hot rolling of the twin-roll cast and homogenized WZ73 considering the resulting equivalent strain and equivalent strain rates are given in

Table 3. Results of hot rolling of the twin-roll cast WZ73 sheets (initial thickness: 5.5 mm).

Sample Number	Rolling Speed in m/s	Initial Thickness in mm	Final Thickness in mm	$\mathrm{Equivalent} \mathrm{Strain} {{\mathit{\phi }}_{\mathit{v}}}_{ }$	$\mathrm{Equivalent} \mathrm{Strain} \mathrm{Rate} {\dot{\mathit{\phi }}}_{\mathit{v}}$ in s−1	Recrystallized Fraction in %
1	1	5.52	3.80	0.43	24.5	-
2	1	5.51	3.52	0.52	27.3	>5
3	1	5.53	3.21	0.63	30.7	12
4	1	5.52	2.89	0.75	34.3	50
5	1.5	5.52	3.53	0.52	40.9	18
6	1.5	5.54	3.22	0.63	46.0	41
7	2	5.54	3.81	0.43	49.0	21
8	2	5.53	3.52	0.52	54.8	43
9	2	5.50	3.19	0.63	61.7	73
10	2	5.55	2.91	0.75	68.4	90

. Hot rolling was carried out in one pass. The final rolling temperature (measured directly after hot rolling) was above 420 °C for all samples. Under the given rolling conditions, edge cracks arise with a maximum length of 25 mm at the highest equivalent strain–strain rate combination. Due to rolling on a rolling mill at the laboratory scale, temperature control is difficult. Rolling on the pilot plant under industrial conditions is expected to result in fewer cracks. Investigations of the hot deformation behavior [7] reveal that dynamic recrystallization occurs at deformation temperatures above 400 °C. Based on the results of the DRX kinetics, it was shown that high equivalent strain rates have a favorable effect on dynamic recrystallization processes. Furthermore, high-equivalent strains are required to obtain a high proportion of DRX. At high-equivalent strain rates (10 s⁻¹), a 50% recrystallized microstructure demands an equivalent strain of 0.7. The equivalent strain of one pass during hot rolling of the WZ73 alloy is limited to 0.75 due to the occurrence of edge cracking. However, rolling speed and therefore the resulting equivalent strain rate are much higher during hot rolling than that applied in compression testing. Hot rolling was performed with equivalent strain rates between 24.5 s⁻¹ and 68.4 s⁻¹. Previous studies revealed that higher equivalent strain rates lead to changes in the mechanisms of the DRX process and contribute to an acceleration of DRX [7]. It is therefore assumed that during hot rolling with high equivalent strain rates, DRX is promoted.

3.3. Microstructure of the as-Rolled WZ73 Alloy

The microstructure of the as-rolled WZ73 consists of deformed grains of the original microstructure and newly recrystallized grains. Depending on the deformation conditions, the amount of the recrystallized grains varied from less than 5% when hot rolling was performed at an equivalent strain of 0.52 with a strain rate of 27.3 s⁻¹ (Figure 2a), to almost completely recrystallized (90%) at a high equivalent strain of 0.75 and strain rate of 68.4 s⁻¹ (Figure 2c). Lower equivalent strains (0.43) are not sufficient to initiate dynamic recrystallization during hot rolling of the twin-roll cast and homogenized WZ73 alloy. Moreover, a high equivalent strain or strain rate alone only partially contributes to DRX (Figure 2b). Although a significant increase in the proportion of recrystallized grains can be achieved, a completely recrystallized structure is not achieved. High equivalent strain rates as well as high equivalent strains are required to obtain progress of the DRX. This is consistent with the hot forming behavior studies in [7], which showed that higher strain rates favor DRX and show increasing recrystallized fractions compared to lower strain rates. Ullmann et al. (2021) [7] presented the DRX volume fraction of a twin-roll cast and heat-treated Mg-6.8Y-2.5Zn-0.4Zr alloy after hot deformation. The deformation at 450 °C and 10 s⁻¹ results in a recrystallized fraction of 40% at an equivalent strain of 0.65. Compared to the results in this work, it can be concluded that increasing the equivalent strain rate is accompanied by an acceleration of the DRX processes. The resulting recrystallized fraction after hot rolling at an equivalent strain of 0.75 is 50% at an equivalent strain rate of 34.3 s⁻¹ and 90% at an equivalent strain rate of 68.4 s⁻¹, respectively. With increasing the equivalent strain, the refinement of LPSO phases is presumed to enhance the DRX as can be seen in the increasing amount of DRX grains [28,29]. Other studies focusing on the deformation behavior of LPSO containing magnesium alloys of the Mg-Y-Zn-system reveal that with increasing strain rate, dynamic recrystallization is delayed [11,30]. Lv et al. (2014) [11] explained this effect by the occurrence of 14H LPSO phases. The amount of lamellar 14H LPSO phases under the investigated rolling conditions in this work is low. It is therefore assumed that their influence on the DRX is small.

Figure 3 shows the influence of the strain rate on the dynamically recrystallized fraction exemplary at samples hot rolled with an equivalent strain of 0.63. As a result of the selected rolling speed, the resulting equivalent strain rate varies from 30.7 s⁻¹ to 46.0 s⁻¹ and further to 61.7 s⁻¹. The increase in the equivalent strain rate simultaneously causes an increase in the recrystallized fraction from 12% to 41% and further to 73%. It is assumed that different recrystallization and/or deformation mechanisms are involved at higher equivalent strain rates. Based on the literature [1,6,7,12,21] as well as on previous studies, the most important deformation mechanisms in LPSO containing magnesium alloys are twinning, kink deformation, and basal <$a$> slip. Some studies report the activation of non-basal slip, such as <$a$> slip on non-basal planes [9,31,32]. Dynamic recrystallization is predominantly driven by continuous dynamic recrystallization (CDRX), twin-induced dynamic recrystallization (TDRX), and kink-induced dynamic recrystallization (KDRX) [1,8,12].

Figure 4 shows a detailed view of the as-rolled WZ73 alloy to reveal the mechanisms of the dynamic recrystallization. Under the given rolling conditions, two main mechanisms occurred: CDRX and TDRX. The continuous dynamic recrystallization is responsible for the formation of new grains along the original grain boundaries, while forming the characteristic “necklace” structure (Figure 4a) [11,33]. Besides this, twin-induced dynamic recrystallization is an important mechanism. TDRX is particularly evident in the early stages of deformation and twinning is known to act as an important deformation mechanism even at higher equivalent strain rates, thus supporting DRX (Figure 4b). Comparable results can be found in [34,35], for example, for other alloy systems.

Kink-induced dynamic recrystallization, which describes the dynamic recrystallization on kinked bands [36], occurs rather less often under the present rolling conditions. The reason for this is the low proportion of lamellar LPSO structures within the Mg matrix. These do not occur in the TRC state, but only precipitate during cooling, especially with previous heat treatment at high temperatures. The cooling during rolling in one pass was a maximum of 60 °C, so lamellar LPSO phases can only be found in a few grains (Figure 5c). However, the network-like LPSO structures are broken up as a result of the deformation during hot rolling (Figure 5a). The refinement and uniform distribution of the LPSO structures (Figure 5b) can thus contribute to DRX in the course of particle-stimulated nucleation (PSN) [37,38].

3.4. Texture of the as-Rolled WZ73 Alloy

The c/a ratio of magnesium is 1.622, close to the ideal value of 1.633, so that hot rolling of magnesium preferentially results in crystallographic textures in which the basal plane is aligned parallel to the sheet plane [39,40,41,42]. The resulting rolling temperature depends on the rolling conditions, in particular thermal treatment, reduction per pass (strain), and strain rate. Figure 6 shows the (0001) and (10$\overline{1}$0) pole figures of the initial state compared to the as-rolled samples rolled with an equivalent strain of 0.52, 0.63, and 0.75, with nearly equal resulting equivalent strain rates between 27.3 s⁻¹ and 34.3 s⁻¹. The hot rolling at lower equivalent strain (0.52) results in a deformed microstructure with original grains elongated in the rolling direction. Dynamic recrystallization hardly occurs. Compared to the twin-roll cast and homogenized initial state, the texture becomes more intense during hot rolling. The resulting rolling texture has a basal character with basal pole splitting, whereby the maxima are tilted away from the core intensity. In addition, the intensity of the texture increases to 6.7 m.r.d. (multiple of random distribution). The texture also exhibits a basal pole broadening in the transverse direction (TD). Increasing equivalent strain hardly leads to a change in the texture if the equivalent strain rate remains almost the same (Figure 6a,b). Only the position of the intensity maxima varies around the core intensity. Basal pole splits also occur in other magnesium alloys during hot deformation, for example ZEK100 [43] or ZAX210 [44], and are often attributed to the dominating recrystallization mechanisms. Since the microstructure analysis in Figure 4 shows that besides CDRX, twin-induced dynamic recrystallization also occurs, which is known to contribute to a basal pole split, it is also assumed that in this study, DRX has a significant influence on texture development during hot rolling of the WZ73 magnesium alloy. Higher equivalent strains result in a higher proportion of recrystallized grains. This leads to a slight weakening of the texture (5.6 m.r.d.) in the course of the incipient recrystallisation. Figure 6c shows the texture at an equivalent strain of 0.63. The intensity maxima are evenly distributed. The sharp maximum of the deformed texture weakens in the course of the DRX. Increasing the equivalent strain to 0.75 and the associated increase in the proportion of recrystallized grains then causes the texture to become sharper again. The investigations show that CDRX is significantly involved in the formation of the new microstructure, which means that there are hardly any changes to the initial texture. These results are in good agreement with the literature showing that when hot rolling Mg alloys, CDRX does not contribute to significant texture change [45,46]. The (10$\overline{1}0)$ prismatic pole figure reveals an almost six-fold symmetry of the basal plane, where the texture components appear as slight peaks located at the circumference with no preferential alignment. This indicates that non-basal slip of <$a$>-dislocations and of <$c+a$>-dislocations is activated. Comparable results are presented by Steiner et al. (2017) [47] during the rolling of AZ31 in different temperature regimes and Ritzo et al. (2020) [48] during hot deformation of AZ31.

The results show that the equivalent strain rate has a considerable influence on the texture development if the strain remains constant. As can be seen from Figure 7, increasing the equivalent strain rate at an equivalent strain of 0.63 results in texture weakening. The resultant texture still has a basal character with several intensity maxima deflected away from the core intensity. However, as the equivalent strain rate increases, the intensity decreases to 3.4 m.r.d. at 68.4 s⁻¹. It is assumed that at higher equivalent strain rates, TDRX and KDRX dominate the dynamic recrystallization behavior more strongly and thus cause a weakening of the texture. At the same time, the (10$\overline{1}$0) pole figure indicates that non-basal slip is active. These results are in accordance with the literature. As presented in Matsuda et al. (2005) [49], the occurrence of LPSO structures leads to an increasing critical resolved shear stress (CRSS) for basal slip and consequently promotes the activation of non-basal slip systems. Kim et al. (2015) [17] found that basal and pyramidal slips are the main deformation modes of the magnesium matrix in LPSO containing magnesium alloys. Based on these results, a combination of high equivalent strain (>0.5) and high equivalent strain rate (>50 s⁻¹) is to be preferred for hot deformation of the WZ73 alloy, especially with regard to processing in several rolling passes.

3.5. Mechanical Properties of the as-Rolled WZ73 Alloy

Mechanical properties of the as-rolled WZ73 magnesium alloy at room temperature in the 0°- (RD) and 90°- (TD) directions for an equivalent strain of 0.63 compared to the TRC and the homogenized condition are shown in

Table 4. Results of tensile testing at room temperature for TRC, homogenized, and hot rolled conditions tested in RD (0°) and TD (90°), average value of five samples.

Condition	0° (RD)			90° (TD)
	UTS in MPa	YS (0.2%) in MPa	A in %	UTS in MPa	YS (0.2%) in MPa	A in %
TRC	248 ± 2.5	195 ± 3.2	0.9 ± 0.3	255 ± 1.8	203 ± 2.6	0.9 ± 0.2
500 °C, 2 h	195 ± 3.3	136 ± 1.8	3.6 ± 0.8	210 ± 3.4	127 ± 2.4	5.9 ± 1.1
0.63|30.7 s−1	292 ± 2.9	246 ± 1.4	3.0 ± 0.5	290 ± 2.9	202 ± 3.1	7.6 ± 1.2
0.63|46.0 s−1	285 ± 5.0	231 ± 2.5	6.1 ± 1.6	279 ± 2.0	198 ± 1.2	10.3 ± 1.2
0.63|61.7 s−1	269 ± 4.1	191 ± 3.7	12.4 ± 2.1	266 ± 0.8	181 ± 1.9	13.4 ± 1.0

. After twin-roll casting, the sheets exhibit brittle material behavior with medium strength characteristics (Figure 8). The inhomogeneous solidification structure and the network-like formation of the LPSO phases mainly contribute to the mechanical properties in the TRC state. Heat treatment increases the ductility of the material. This can be attributed to the homogenization of the microstructure and the dissolution of the network-like structure of the LPSO phases. At the same time, there is a random texture, so that the mechanical properties exhibit only slight anisotropy. Hot rolling of the twin-roll cast and homogenized WZ73 magnesium alloy results in improved mechanical properties. At a low equivalent strain rate (30.7 s⁻¹), the sheets offer high strength values in the 0°-direction, more precisely 292 MPa ultimate tensile strength and 246 MPa 0.2% proof stress. The elongation at fracture is 3%. In the 90° direction, the elongation at fracture is 7.6% and is accompanied by a lower 0.2% proof stress (202 MPa). The differences in the mechanical properties in the 0° and 90° directions can be attributed to the texture, which exhibits a basal pole broadening in TD. Therefore, more crystallites display a favorable orientation for deformation under tensile stress. The anisotropy of the mechanical properties decreases with an increasing equivalent strain rate, which can be ascribed to the texture weakening.

The development of the mechanical properties independent of the equivalent strain rate can be attributed to the increasing area of recrystallized grains resulting in a higher ductility. Increasing the equivalent strain rate for a constant equivalent strain of 0.63 causes an increase in elongation at fracture in the 90° direction from 7.6 to 13.4%. The strength and ductility of the WZ73 alloy increased significantly after hot rolling when compared with the initial state (twin-roll cast and homogenized). Hot rolling results in texture sharpening. Nevertheless, although this is not a classic, strongly pronounced basal texture, some grains are oriented in such a way that the Schmid factor for basal slip becomes insignificant. Itoi et al. (2013) [29] describe the same for the LPSO phases, which align themselves with their basal planes parallel to the sheet plane in the course of deformation and reinforce this effect. Both thus contribute to the strengthening of the alloy.

Comparable mechanical properties for a hot rolled magnesium alloy with LPSO structures (Mg₉₈Y₁Zn₁) are presented by Itoi et al. (2013) [29]. The increasing strength of the WZ73 alloy after hot rolling can be attributed to the strengthening effect of the LPSO structures. LPSO structures affect the deformation mode during tensile testing [29] and contribute to strength improvement by reinforcing characteristics due to their unique crystal structure, the interface properties, and the magnesium matrix or kink band formation [50].

4. Conclusions

In this work, hot rolling of the twin-roll cast and homogenized WZ73 alloy was conducted, which was immediately followed by water quenching, for the purpose of investigating the influence of different equivalent strain and equivalent strain rates on its microstructure, texture, and mechanical properties. The microstructure and texture analysis revealed that high equivalent strains (>0.5) and high equivalent strain rates (>50 s⁻¹) are required to attain advanced dynamic recrystallization. The DRX is substantially supported by continuous dynamic recrystallization (CDRX) and twin-induced dynamic recrystallization (TDRX). Both contribute to the formation of a texture with basal character with a basal pole split or tilted maximum away from the core intensity during hot rolling. It is assumed that the predominant deformation modes are not the non-basal slip of <$a$>-dislocations and <$c+a$>-dislocations. However, increasing the equivalent strain for almost constant equivalent strain rates does not lead to a change in the resulting texture. Texture weakening only occurs at high equivalent strain rates. It is assumed that TDRX and KDRX dominate the dynamic recrystallization at higher equivalent strain rates. Strength and ductility of the WZ73 alloy increase significantly after hot rolling when compared with the initial state. Hot rolling at a high equivalent strain (0.63) and high equivalent strain rate (61.7 s⁻¹) results in mechanical properties with low anisotropy because of a significant texture weakening.