Effect of Rolling Temperature on the Microstructure and Mechanical Properties of Mg-2Zn-0.4Y Alloy Subjected to Large Strain Rolling

Abstract

Large strain rolling (LSR) has been conducted on the Mg-2Zn-0.4Y alloy. After the 1st rolling process at 250, 300, 350, and 400 °C, the alloy demonstrates a fully recrystallized microstructure. The grain size increases from 6, 8, 12, to 17 μm with an increasing rolling temperature. After the 2nd rolling process at 300 °C, twinning and shear bands were introduced. During the 3rd rolling process at 350 °C, dynamic recrystallization (DRX) was observed and resulted in a more uniform microstructure. DRX occurred because of temperature increase and large dislocation density induced by LSR. For the room temperature tensile tests, the plates rolled at 300 and 350 °C in the 1st rolling process demonstrate higher strength and lower elongation due to twinning. The one rolled at 400 °C in the 1st rolling process, shows the most uniform rolling microstructure and the best combination of strength and elongation at room temperature.

1. Introduction

Rare earth (RE) elements are widely used to improve the mechanical properties of Mg alloys [1,2,3,4,5]. Yttrium is one of the most interesting elements, because it was found able to improve the mechanical properties and the corrosion resistance of Mg alloys at the same time [6]. Lee et al. [7] pointed out that Zn/Y ratio has an effect on the formation of second phases in Mg-Zn-Y alloys. When Zn/Y ratio is ~10, Mg₇Zn₃ phase forms in the as-cast microstructure; when it is 5~7, there are α-Mg and I phase (Mg₃Zn₆Y); when it is 1.5~2, W phase (Mg₃Zn₃Y₂) appears besides α-Mg and I phase [8,9]. Shao et al. [10] observed long period stacking ordered (LPSO) structure in a Mg₉₇Zn₁Y₂ alloy subjected to hot compression. Mg-Zn-Y alloys containing I phase were found to demonstrate high strength as well as large elongation to failure at both ambient and elevated temperatures [7,11].

There is some recent progress in microstructural analysis of different materials under heat treatments [12,13,14,15]. Lu et al. [12] investigated the substitution behavior of Ag in the Ag/Ti₂AlC interaction area prepared following the wetting experiment, which was carried out at 1030 °C for 5 min. Interfacial microstructure and mechanical properties of TC4/Ti₃SiC₂ joints brazed at different temperatures were investigated by Wang et al. [13]. Moneghini et al. [14] characterized the solid dispersions of itraconazole and vitamin E TPGS prepared by microwave technology to study the effect of microwave irradiation. Panna et al. [15] utilized a hot-stage microscopy to evaluate sample morphology changes on heating. The method was found applicable to precise measuring the coefficient of thermal expansion, S, of clay raw materials.

Large strain rolling (LSR) is an efficient way to produce metal sheets with fine grain sizes [16]. Compared to other severe plasticity deformation (SPD) methods, LSR is simpler and can be applied to large parts, which makes it possible to scale-up in industry. In the previous studies [17,18,19,20,21], the microstructure of AZ31, AZ61 and AM60 Mg alloys were significantly refined through LSR. More Twinning and dynamic recrystallization (DRX) were observed compared to normal rolling, and a more uniform microstructure was achieved even in thick plates. To the best of our knowledge, LSR has not be systematically studied on Mg-Zn-Y alloys. Therefore, it is meaningful to study the microstructure evolution during the LSR of Mg-Zn-Y alloys to establish the relationship between the rolling temperature and the final strength and ductility.

In this work, the Mg-2Zn-0.4Y alloy was subjected to three processes of hot rolling. In the 1st rolling process, six passes (two large strain passes) were conducted at temperatures of 250, 300, 350, and 400 °C. In the 2nd and 3rd rolling processes, four passes of large strain rolling were performed at 300 and 350 °C, respectively. The microstructure evolution after each rolling process was investigated. The room temperature tensile properties and fractures of the rolled plates were also studied after all three rolling processes.

2. Experimental Procedures

The chemical composition of the Mg-2Zn-0.4Y alloy was measured using inductively coupled plasma mass spectrometry (Optima 8300DV ICP-OES, PerkinElmer, Waltham, MA, USA) and the result is shown in

Table 1. Chemical composition of the Mg-2Zn-0.4Y alloy in wt.%.

Alloy	Zn	Y	Mn	Fe	Mg
Mg-2Zn-0.4Y	1.94	0.32	0.0021	0.0059	Balance

. Before rolling, the cast ingots were homogenized at 400 °C for 10 h. In this work, there are three rolling processes in total. In the 1st rolling process, six passes of rolling were performed on the ingots of 41 mm thickness at four temperatures of 250, 300, 350, and 400 °C and reductions of 15%, 15%, 17%, 13%, 33%, and 44%, respectively. These four plates are referred to as plates A, B, C, and D, respectively, in this paper. In the 2nd rolling process, four passes of rolling were conducted at 300 °C with heavy-reductions of 25%, 27%, 32%, and 33%, respectively. In the 3rd rolling process, four passes of rolling were conducted at 350 °C with the same reductions as the 2nd rolling process. The final thickness of the plates is about 1.5 mm. The detailed temperatures and the reductions of the rolling process are given in

Table 2. Temperatures and reductions of rolling.

Rolling Process	Temperature/°C	Reduction/%
		1st Pass	2nd Pass	3rd Pass	4th Pass	5th Pass	6th Pass
1st	250	15	15	17	13	33	44
	300	15	15	17	13	33	44
	350	15	15	17	13	33	44
	400	15	15	17	13	33	44
2nd	300	25	27	32	33
3rd	350	25	27	32	33

.

Tensile specimens were machined into dog bone shape with a gauge length of 25 mm and width of 20 mm. Tensile tests were performed on a Shimadzu AG-X (Shimadzu, Tokyo, Japan) universal testing machine at room temperature and a crosshead speed of 1 mm/min. For these tests, the tensile load was applied parallel to the rolling direction (RD) of the plates.

Microstructure was observed using Leica DM4M (Leica, Wizral, Germany) optical microscope (OM). The samples for OM analyses were prepared using standard technique of grinding with SiC abrasive paper and polishing with diamond paste of 1 μm and 0.5 μm. Afterwards, they were etched in a solution of 6 g picric acid, 5 mL glacial acetic acid, 100 mL ethanol, and 100 mL distilled water. In this paper, all the OM figures were taken at the cross-section of RD and normal direction (ND), and the RD of the samples were placed along the horizontal direction. Grain sizes were measured using the public domain software ImageJ (version 1.52a, Bethesda, MD, USA) with the intercept method based on the ASTM standard E112-13 (Standard Test Methods for Determining Average Grain Size) [22,23,24]. On each sample, five test lines were placed randomly and intercepted with at least 20 grains depending on the grain sizes. Second phases in the alloy and the fractures of the tensile specimens were analyzed using Zeiss Ultra Plus (Jena, Germany) field emission scanning electron microscope (SEM).

3. Results and Discussion

The microstructure of the Mg-2Zn-0.4Y alloy after 400 °C/10 h homogenization is displayed in Figure 1a. The average grain size of this alloy is about 30 μm. The alloy is composed of α-Mg and some secondary phases, as shown in Figure 1b. Energy dispersive spectrometry (EDS) result show that these are MgZnY ternary phases. Lee et al. [7] found that the alloy consists of α-Mg and I phases when the Zn/Y ratio is between 5 to 7 in the Mg-Zn-Y alloys. The composition of I phase is about Mg₃Zn₆Y [8,9], which is in good agreement with the EDS result of this work. Therefore, the current Mg-2Zn-0.4Y alloy is composed of α-Mg and some I phase.

In the first rolling process, the Mg-2Zn-0.4Y alloy was rolled six passes with reductions of 15%, 15%, 17%, 13%, 33% and 44%. The rolling was conducted at four temperatures of 250, 300, 350, and 400 °C, and the corresponding plates are designated as plates A, B, C, and D, respectively. The rolling microstructure is displayed in Figure 2. According to this figure, all four plates have a fully recrystallized microstructure. The rolling temperature was found to have a significant effect on the grain size. It increases with an increasing rolling temperature. The average grain sizes in plates A, B, C, and D shown in Figure 2a–d is about 6, 8, 12, and 17 μm, respectively. When the rolling was performed at 250 and 400 °C, plates A and D show a uniform microstructure. For plates B and C, which were rolled at 300 and 350 °C, the microstructure has a mixture of fine recrystallized grains and relatively large grains due to grain growth.

In the 2nd rolling process, plates A-D were rolled four passes at 300 °C with high reductions of 25%, 27%, 32% and 33%, respectively. The microstructure is shown in Figure 3. Twinning can be seen in all four plates, among which plate A has the least number of twinning and plate D has the most of it. Compared with the microstructure after the 1st rolling process shown in Figure 2a–c, plates A, B, and C experienced grain growth during this rolling process, and twinning formed in the large grains. The microstructure of plates B and C also became more uniform. In Figure 3d, besides twins, large numbers of shear bands can be seen plate D. These shear bands are at about 45° to the rolling direction.

In the 3rd rolling process, plates A-D were further rolled four passes at 350 °C with heavy reductions of 25%, 27%, 32%, and 33%, respectively. The microstructure is displayed in Figure 4. Compared with the microstructure after the 2nd rolling process shown Figure 3, some twinning and recrystallized grains can be observed in plates A and B. More fine recrystallized grains occur in plate C. Plate D has the largest grain sizes and the most uniform microstructure. Almost no twinning can be observed in plate D.

Temperature change during the rolling process can be calculated using the following Equation (1) [25]:

$$\Delta \mathrm{T}=\Delta T_p+\Delta T_f-\Delta T_R$$

(1)

where $\Delta \mathrm{T}$ is the total temperature change during rolling, $\Delta T_p$ is the temperature rise due to plastic work, $\Delta T_f$ is the temperature rise due to friction between the rolls and the plate, and $\Delta T_R$ is the temperature drop due to contact between the rolls and the plate [25]. Using the above equation, Su et al. [26] estimated the temperature change of AZ31 sheets rolled at the speed of 15 m/min with reductions of 15%, 23%, 30%, and 37%, respectively. The results are shown in the following

Table 3. Temperature change for low speed rolling with different rolling reductions [26].

Reduction	15%	23%	30%	37%
ΔTp	23.1	39.6	56.3	75.5
ΔTf	1.8	3.7	6.0	8.9
ΔTR	36.6	42.5	47.3	52.0
ΔT	−11.7	0.8	15.0	32.4

[26]. When the reduction is 15%, the value of ΔT is negative, which means there is temperature drop; when it is 23%, the temperature almost remains the same; when it is about 30% and 37%, the temperature increases 15.0 and 32.4 °C, respectively. In the 1st rolling process of the current work, two passes with large strains of 33% and 44% were applied after four passes with strains under 20%. Due to the temperature increase with large strain, all four Mg-2Zn-0.4Y alloy plates demonstrate a fully recrystallized fine grain microstructure as shown in Figure 2, including plate A, which was rolled at the lowest temperature of 250 °C. In the 2nd and 3rd rolling processes, four passes with large strains of 25%, 27%, 32%, and 33% were conducted continuously at 300 and 350 °C, respectively. Deformation heat accumulated during rolling and the actual temperature of the plates can be higher than 400 °C. Thus, DRX and grain growth occurred in the rolled plates resulting in a uniform rolling microstructure as can be seen in Figure 3 and Figure 4.

Besides the temperature increase, LSR also induces high density of dislocations and deformation twins in the Mg-2Zn-0.4Y alloy. Dudamell et al. [27] studied the severe plastic deformation behavior of AZ31 magnesium alloy using a split Hopkinson pressure bar. Significant grain refinement was observed due to grain subdivision by the formation of geometrically necessary boundaries. These boundaries evolved from dislocation boundaries during the plastic deformation within the original grains [28]. Guo et al. [20] investigated the deformation energy distribution during the LSR of the AZ31 Mg alloy. It was pointed out that there is very small proportion of deformation energy stored in twinning. However, the deformation twins can act as preferential nucleation sites for the DRX grains [29,30]. Thus, large numbers of dislocations produced by large strain provide driving force for the nucleation and grain growth of the DRX process, and most of the deformation energy during the LSR is released in this way. Second phases have a great effect on the DRX behavior of Mg alloys subjected to LSR. According to Figure 2, Figure 3 and Figure 4, the rolling microstructure of the Mg-2Zn-0.4Y alloy in this study is quite homogeneous due to the existence of the I phase. These second phases can facilitate the DRX process through particle stimulated nucleation mechanism. Perez-Prado [18] also observed a quite uniform DRX microstructure in the large strain-rolled AM60 alloy with Al₂Mg₃ and Al₆Mn phases. On the other hand, in the AZ31 alloy with very few second phases, DRX was constricted in shear bands during the large strain hot rolling process [19,20].

Tensile tests were performed at room temperature and a crosshead speed of 1 mm/min on plates A, B, C, and D after all three rolling processes. The engineering stress-strain curves are shown in Figure 5. The yield strength, tensile strength and elongation of the four rolled plates are summarized in

Table 4. Summary of tensile properties of plates A, B, C, and D.

Material	Yield Strength/MPa	Tensile Strength/MPa	Elongation/%
plate A	158.1	244.3	9.6
plate B	175.9	266.7	8.0
plate C	162.8	269.4	7.7
plate D	143.5	243.9	12.3

. In Figure 5, plates B and C, which were rolled at intermediate temperatures of 300 and 350 °C, respectively, in the 1st rolling process, demonstrate similar flow curves with high strength and low elongation. While plates A and D, which were rolled at 250 and 400 °C, respectively, in the 1st rolling process, have similar levels of lower flow stress. Among these four plates, plates B and C have elongations of about 9%, while plate D has the largest elongation of about 12%. According to the final rolling microstructure shown in Figure 4, plates B and C demonstrate more twins compared to plates A and D. Twin boundaries can be obstacles for the dislocations during the deformation. A localization of stress can also occur at twinning, which causes early failure of the tensile specimens. In Figure 4, plate A has much less twins and almost no twins can be observed in plate D. Thus, the large elongation of plate D benefits from the lower twinning and dislocation density caused by DRX during the LSR [30].

Fractures of the tensile specimens of plates A, B, C, and D are displayed in Figure 6. According to this figure, the fracture of plate A consists of fine and equiaxial dimples. Plate B has a fracture of elongated dimples and tear ridges, corresponding to a slight decrease of elongation shown in Figure 5. Plate C shows quasi-cleavage facets in the fracture and has a deteriorated elongation due to the high fraction of twinning. The fracture of plate D is composed of large and deep dimples indicating the occurrence of large and uniform deformation prior to fracture because of the homogenous microstructural characteristics.

In summary, among these four plates, Plate A, which was rolled at 400 °C in the 1st rolling process, shows the most uniform final rolling microstructure and the best combination of strength and elongation at room temperature.

4. Conclusions

The Mg-2Zn-0.4Y alloy experienced three processes of large strain hot rolling. After the 1st rolling process, plates A, B, C, and D, which was rolled at 250, 300, 350, and 400 °C, respectively, demonstrate a fully recrystallized microstructure. The grain sizes increase from 6, 8, 12, to 17 μm with an increasing temperature of rolling. After the 2nd process of rolling at 300 °C, deformation twins can be observed in all four plates. After the 3rd rolling process at 350 °C, plates B and C still demonstrate large numbers of twins, very few twins exist in plate A, and plate D has a fully recrystallized microstructure and the largest grain sizes. Due to the temperature increase, as well as the large densities of dislocations and twinning induced by LSR, nucleation and growth of the DRX grains occurred in these Mg-2Zn-0.4Y plates, resulting in a uniform rolling microstructure. Plates B and C demonstrate higher strength and lower elongation compared with plates A and D. This is because there is more twinning in plates B and C. The twin boundaries act as obstacles for moving dislocations, and a localization of stress at twinning can causes early failure of the tensile specimens. Plate A, which was rolled at 400 °C in the 1st rolling process, shows the most uniform final rolling microstructure and the best combination of strength and elongation at room temperature.