Low temperature superplasticity of ultrafine grained Mg–9.25Zn–1.66Y alloy with an icosahedral quasicrystalline phase

Author’s Accepted Manuscript Low temperature superplasticity of ultrafine grained Mg-9.25Zn-1.66Y alloy with an icosahedral quasicrystalline phase W.J. Kim, Y.J. Yoo www.elsevier.com/locate/msea

PII: DOI: Reference:

S0921-5093(15)30176-3 http://dx.doi.org/10.1016/j.msea.2015.07.022 MSA32560

To appear in: Materials Science & Engineering A Received date: 16 April 2015 Revised date: 30 June 2015 Accepted date: 8 July 2015 Cite this article as: W.J. Kim and Y.J. Yoo, Low temperature superplasticity of ultrafine grained Mg-9.25Zn-1.66Y alloy with an icosahedral quasicrystalline p h a s e , Materials Science & Engineering A, http://dx.doi.org/10.1016/j.msea.2015.07.022 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Low temperature superplasticity of ultrafine grained Mg-9.25Zn-1.66Y alloy with an icosahedral quasicrystalline phase

W. J. Kim* and Y. J. Yoo Department of Materials Science and Engineering, Hongik University, Mapo-gu, Sangsu-dong 72-1, Seoul 121-791, Korea

∗ Corresponding author. Tel.: +82 2 320 1468; fax: +82 2 325 6116, E-mail address: [email protected] (W. J. Kim)

Abstract The ultrafine grained Mg-9.25Zn-1.66Y alloy sheet with I-phase prepared by directly applying high-ratio differential speed rolling to the as-cast microstructure exhibited excellent low-temperature superplasticity (386 - 888% at 443 - 473 K) at a moderately high strain rate of 10-3 s-1. A criterion for achieving low temperature superplasticity was proposed.

Keywords: Mechanical characterization; Magnesium alloys; Thermomechanical processing; Grain refinement; Superplasticity

1. Introduction The icosahedral quasicrystalline phase (I-phase, Mg3YZn6) in Mg-Zn-Y alloys has attracted researchers’ attention because it has high thermal stability and forms a low 1

interfacial energy and coherent interface with the matrix [1, 2]. Researchers have tried to refine the cast microstructures of the I-phase containing Mg-Zn-Y alloys through various thermomechanical processing routes [1, 2]. Superplasticity of the I-phase containing Mg–Zn–Y–(Zr) alloys with fine grains has been studied by several investigators [3-6]. There are, however, relatively limited studies on their superplastic behaviors at low temperatures below 473 K. Low temperature superplasticity (LTS), meaning superplasticity below 0.5 Tm (where Tm is the melting temperature), is attractive for the application of superplasticity in industrial forming operations in terms of energy saving and prevention of the surface oxidation of products. The purpose of this study is to explore the possibility of achieving LTS from the ultrafine grained (UFG) I-phase containing Mg-9.25Zn-1.66Y alloy, which was prepared by directly applying the high-ratio differential speed rolling (HRDSR) technique to a cast microstructure.

2. Materials and methods The starting alloy used in the present study was as-cast Mg-9.25 wt.% Zn-1.66 wt.% Y alloy (MgZn3.7Y0.49). The as-cast slab with 4 mm was rolled using conventional rolling followed by high-ratio differential speed rolling (HRDSR, with a speed ratio of 3) to a final thickness of 0.63 mm. The detailed casting and rolling procedures are available elsewhere [7]. The microstructures of the HRDSRed alloy were examined by using a field-emission transmission electron microscope (JEM 2001F) operated at 200 kV. For TEM sample preparation, the HRDSRed alloy was mechanically polished and then thinned by jet polishing (using a solution composed of 60% methyl alcohol, 30% glycerine and 10% nitric acid), followed by ion milling (BAL-TEC RES 101). 2

For the tensile testing, tensile specimens with dog-bone geometries with a 5-mm gauge length, 4-mm width, 0.6-mm thickness and 4-mm shoulder radius were cut along the rolling direction. The strain rate change (SRC) tests were conducted at 443-653 K. The elongationto-failure tests were conducted at 443- 653 K using the initial strain rates of 5  10-4 - 1  10-2 s-1. The sample heating and holding time before tensile loading was 10 min.

3. Results and discussion Figures 1 (a)-(d) show the TEM micrographs of the HRDSRed alloy. The HRDSRed alloy has fine and equiaxed grains with the sizes of 0.5-1.0 m. In the microstructure, the irregular shaped particles with sizes of 0.1-1 m (Figures 1 (a) and (b)) are the I-phase fragments broken and separated from the I-phase eutectic structure (inset of (a)), whereas the round or elliptical shaped particles with typical sizes of 100 nm (Figures 1(c) and (d)) are I-phase particles formed during the casting (solidification) process (inset of (d)). Many particles with the sizes less than 50-150 nm interact with moving dislocations or grain boundaries (indicated by arrows in Figures 1(b) and (d)), indicating the potential of these size particles for stabilizing the fine-grained microstructure at elevated temperatures. Figure 2 (a) shows the tensile elongation values of the HRDSRed alloy plotted as a function of the temperature at a given strain rate of 1  10-3 s-1. The following is inferred. First, a tensile elongation of 386 % is obtained at a temperature as low as 443 K (0.48Tm). At 473 K (0.51Tm), a large tensile elongation of 888 % is obtained. At 523 K, a significant decrease of the tensile elongation (to 230 %) is resulted. As the temperature further increases, however, the tensile elongation gradually increases (up to 350 % at 653 K). This tensile elongation behavior indicates that the HRDSRed alloy exhibits excellent LTS, but its superplastic ability sharply degrades above 523 K. Figure 2(b) shows the engineering stress3

engineering strain curves of the HRDSRed alloy at different strain rates at a given temperature of 473 K. Compared at 1  10-3 s-1, a slightly larger tensile elongation of 944 % is obtained at a lower strain rate of 5  10-4 s-1, but a notably smaller elongation of 263 % is obtained at a higher strain rate of 10-2 s-1. It is worthwhile to note that the tensile elongation behavior of the HRDSRed alloy as a function of temperature and its maximum tensile elongation value at 10-3 s-1 are similar to those of the Mg–5.00Zn–0.92Y–0.16Zr prepared by applying the combination of extrusion (an extrusion ratio of 9:1) and ECAP (8 passes) to the cast microstructure [6]. This comparison demonstrates the high efficiency of the HRDSR technique for refinement of microstructures of the I-phase containing Mg-Zn-Y alloys. Figure 3(a) shows the SRC test results conducted at various temperatures. At 443 K, the HRDSRed alloy has a high m value of ~ 0.5 at low strain rates ( 10-3 s-1). At 473 K, the high m value region extends to higher strain rates (up to 10-2 s-1). At 523 K, however, the m value at low strain rates largely decreases to 0.3. As temperature further increases to 653 K, the m value increases to ~0.5 and the high m value region extends to higher strain rates again. This variation of m value as a function of temperature can explain why the tensile elongation of the HRDSRed alloy sharply decreases at 523 K and then gradually increases again with further increase in temperature. Like the m value, flow-stress at low strain rates also suddenly changes at 523 K. While flow stress at high strain rates ( 10-2s-1) systematically decreases as the temperature increases, the flow stress at low strain rates abruptly increases at 523 K and then decreases again as the temperature further increases. A question arises: why does the m value and flow stress at low strain rates behave abnormally at 523 K? In a previous study, we demonstrated that the deformation of finegrained Mg alloys can be predicted using Eq. (1) when the contributions of grain-size 4

dependent grain boundary sliding (GBS) and grain-size independent dislocation climb creep (DCC) to the total plastic flow are simultaneously considered [8]:

D *eff    2 Deff    5   A 2    B 2   , d E b E

(1)

In this equation, d is the true grain size (=1.74  L, where L is the linear intercept grain size), E is the Young’s modulus of pure Mg [9], b is the Burger’s vector of Mg [9], and Deff and D*eff are the effective diffusion coefficients [8]. In this analysis, the A and B values were chosen to obtain the best fit to the experimental data over all experimental conditions: 3.8 108 and 7.41  107. The curve fitting results using Eq. (1) are presented in Figure 3. When the initial grain size was used (L=1.1 m [12]), good curve fitting was obtained at 443 and 473 K, but large deviation occurred at higher temperatures. At temperatures  523 K, the best fitting was obtained at L= 6 m. This grain size is similar to that (5.3 m) measured from the microstructure just before tensile loading for the SRC tests at 523 K (Figure 4(a)), indicating that rapid grain growth occurs during the sample heating and holding duration at temperatures above 523 K. Figure 4 (b) show the microstructure taken from the gauge section of the HRDSRed alloy sample after the elongation-to-failure test at 473 K -10-3s-1. The very fine grained microstructure was retained (L = 1.9 m) despite having experienced the large deformation. Degree of dispersion of the I-phase particle has increased in the gauge section compared to the as-annealed microstructure shown in Figure 4(a), indicating that superplastic flow enhances the dispersion of I-phase particles through disintegration of particle agglomerates. This particle dispersion may contribute to the stability of the UFG microstructure against grain growth. 5

A critical strain rate, c , at which transition of the deformation mechanism from grain boundary sliding (GBS) to dislocation climb creep (DCC) occurs was calculated based on the analysis of the deformation behavior of the HRDSRed alloy by Eq. (1). As the curve fitting results indicated that grain boundary diffusion (Dgb) -controlled GBS [8, 10] and pipe diffusion (Dp)-controlled DCC [8, 10] control the plastic flow at low and high strain rates, respectively, at temperatures around and less than 0.5Tm, an expression for c could be derived by setting 

DgbGBS

 D p  DCC :

 0.068 Ab  c    3  d 

7/5

 b2     0.8B 

2/5

(2)

Dgb

,

According to Eq. (2), as the temperature increases and the grain size decreases, c increases. Eq. (2) can explain the change of the slope of the strain rate-stress curve (Figure 3) as a function of temperature. At 523 K where large grain growth occurs, c decreases, leading to extension of DCC with m =0.2 to lower strain rates. As temperature further increases, c increases, leading to extension of GBS with m =0.5 to higher strain rates. Figure 5 shows a plot of d as a function of the temperature at different c . For 443 K, at d= 1.9 m (L =1.1 m),

c is near 10-3 s-1. For this reason, a moderate value of tensile elongation near 400% was obtained at 10-3 s-1. For 473 K, where d=1.9 m, c is near 10-2 s-1; thus, superplastic deformation at 10-3 s-1 is excellent. For 523 K, where d=10.4 m, c is near 10-4 s-1; thus, non-superplastic deformation occurs at 10-3 s-1.

4. Conclusions 6

The ultrafine grained I-phase containing magnesium sheet, which exhibits excellent LTS, could be fabricated using the HRDSR technique. The sheet has good thermal stability below 473 K.

Acknowledgments This research was financially supported by the Basic Science Research Program through the National Research Foundation of Korea (2013) funded by the MEST (2013R1A1A2010637).

References [1] J.Y. Lee, D. H. Kim, H. K. Lim, D. H. Kim, Mater. Lett. 59 (2005) 3801. [2] D. K. Xu, E. H. Han, Prog. Nat. Sci.: Mater. Inter. 2 (2012) 364. [3] M. Y. Zheng, S. W. Xu, K. Wu, S. Kamado, Y. Kojima, Mater. Lett. 61 (2007) 4406. [4] W. N. Tang, R. S. Chen, E. H. Han, J. Alloys compd. 477 (2009) 636. [5] D. H. Bae, Y. Kim, I. J. Kim, Mater. Lett. 60 (2006) 2190. [6] S. W. Xu, M. Y. Zheng, S. Kamado, K. Wu, Mater. Sci. Eng. A549 (2012) 60-68. [7] T. Y. Kwak, H. K. Lim, S. H. Han, W. J. Kim, Scripta Mater. (2015) doi:10.1016/j.scriptamat.2015.03.004 . [8] W. J. Kim, M. J. Kim, J. Y. Wang, Mater. Sci. Eng. A527 (2009) 322.

[9] H.J. Frost, M.F. Ashby, Deformation-mechanism Maps, Pergamon Press, Oxford (1982) (Chapter 6). 7

[10] O. D. Sherby, J. Wadsworth, Prog. Mater. Sci. 33 (1989) 169.

Figure captions Figure 1. (a)-(d) TEM images of the HRDSRed alloy at different locations. The inset in (a) and (d) show SEM images of the cast microstructure.

Figure 2. (a) Tensile elongations of the HRDSRed alloy and the ECAPed Mg-5Zn-0.92Y0.16Zr alloy [6] given as a function of testing temperature at 10-3s-1. (b) Engineering stressengineering strain curves of the HRDSRed alloy at different strains rates at 473 K.

Figure 3. SRC test results for the HRDSRed alloy at various temperatures. The dotted curves represent the predictions by Eq. (1).

Figure 4. (a) The microstructure observed just before tensile loading for the SRC tests at 523 K. (b) The microstructures taken from the gauge section of the HRDSRed alloy after the tensile test at 473 K -10-3 s-1. Microstructural observation was made on the longitudinal sections.

Figure 5. A plot of d as a function of the temperature at different c , which was calculated by Eq. (2).

8

1 m

500 nm

1 m

500 nm

Figure 1. (a)-(d) TEM images of the HRDSRed alloy at different locations. The inset in (a) and (d) show SEM images of the cast microstructure.

9

900

HRDSRed alloy (Cast + HRDSR) Mg-5.00Zn-0.92Y-0.16Zr (Extrusion+ECAP) [11]

120

Engineering stress (MPa)

Tensile elongation (%)

1000 800 700 600 500 400 300 200 100

0 400 425 450 475 500 525 550 575 600 625 650 675

-2

110

10 /s

Temp. : 473 K

-3

10 /s

100

-4

5 x10 /s

90 80 70 60 50 40

Elong. % = 888, 473 K/1 x 10-3 s-1

30 20 10 0

0

1

2

3

4

5

6

7

8

9

10

Engineering strain

Temperature (K)

(a)

(b)

Figure 2. (a) Tensile elongations of the HRDSRed alloy and the ECAPed Mg-5Zn-0.92Y0.16Zr alloy [6] given as a function of testing temperature at 10-3s-1. (b) Engineering stressengineering strain curves of the HRDSRed alloy at different strains rates at 473 K.

10

-1

-1

Strain rate (s )

10

443 K 473 K 523 K 573 K 623 K 653 K

-2

10

-3

10

m = 0.5 -4

10

-5

10

0

10

1

2

10

10

3

10

Flow stress (MPa) Figure 3. SRC test results for the HRDSRed alloy at various temperatures. The dotted curves represent the predictions by Eq. (1).

11

(a)

(b)

20μm

20μm

Figure 4. (a) The microstructure observed just before tensile loading for the SRC tests at 523 K. (b) The microstructures taken from the gauge section of the HRDSRed alloy after the tensile test at 473 K -10-3 s-1. Microstructural observation was made on the longitudinal sections.

12

11 -4

10 /s

10

-3

10 /s

9

-2

10 /s

8

d (m)

7 6 5 4 3 2 1 430

440

450

460

470

480

490

500

510

520

530

Temperature (K) Figure 5. A plot of d as a function of the temperature at different c , which was calculated by Eq. (2).

13