• 沒有找到結果。

國立中山大學材料科學研究所 碩士論文

N/A
N/A
Protected

Academic year: 2022

Share "國立中山大學材料科學研究所 碩士論文 "

Copied!
114
0
0

加載中.... (立即查看全文)

全文

(1)

國立中山大學材料科學研究所 碩士論文

添加硼對高功能鎂銅釔合金非晶質化及奈米結晶行為之影響 Amorphization and Nanocrystallization Behavior in Mg-Cu-Y Alloyby

Adding Boron

研究生:鄭宇庭 撰 指導教授:黃志青 博士

中華民國 九十四 年 七 月

(2)

Content

Content... i

Tables List... iii

Figures List... iv

Abstract ... 1

1 Introduction and literature review... 4

1.1 Introduction……... 4

1.2 The development of Mg-based amorphous alloys...4

1.3 Influence of quaternary additions on Mg-based alloys... 5

1.4 The purpose of this study... 6

2 Background... 7

2.1 The evolution of amorphous alloys... 7

2.2 The systems of glassy metals... 8

2.3 The dominant factors for the glass forming ability... 9

2.4 The main empirical rules for the synthesization of amorphous alloys...….….... 10

2.4.1 Bonding parameter for Mg-based BMGs………... 13

2.5 Characterization of amorphous alloys…………... ... 14

2.5.1 Mechanical properties... ... 15

2.5.2 Magnetic properties………. ... 15

2.5.3 Chemical properties... 15

2.5.4 Other properties and behavior of bulk amorphous alloys... 15

2.6 Crystallization transformation of amorphous………... 16

2.6.1 The Kissinger plot... 16

2.6.2 Modified kinetics of non-isothermal crystallization kinetics…... 17

(3)

2.6.3 Isothermal analysis for the crystallization kinetics of amorphous alloys… ..20

2.6.4 The Arrhenius plot... 21

3 Experimental procedures... ... 22

3.1 Materials... 22

3.2 Synthesization methods... 22

3.2.1 Arc melting... 22

3.2.2 Melt spinning... 23

3.3 Phase identification by XRD... 24

3.4 SEM/EPMA observations………... ..24

3.5 Microhardness testing………... 24

3.6 Density measurement...25

3.7 DSC thermal stability analysis... ... 25

3.8 Microstructure characterizations... 26

4. Experimental Results... 27

4.1 Specimens prepartion... 27

4.2 X-ray diffraction analyses... 27

4.3 SEM/EPMA observations... 28

4.4 Microhardness tests………... ... 29

4.5 Density measurements... …... ... 29

4.6 Thermal analyses... ... ... 30

4.6.1 Non-isothermal analyses and kinetics... ... ...30

4.6.2 Kissinger plots for non-isothermal analyses...32

4.6.3 Modified Kissinger plots for non-isothermal analyses... ...32

4.6.4 Isothermal analyses and kinetics... ……… 34

4.7 TEM observations………. 36

5. Discussions……….. 37

(4)

5.1 The effect of the boron element………. 37

5.2 XRD analyses………. 38

5.3 Thermodynamics and kinetics analyses………. 39

6. Conclusions………. 42

References... 44

Tables... 47

Figures... 56

(5)

Table List

Table 1.1 Application filed for bulk metallic glasses... 47

Table 2.1 The classification of amorphous alloy systems... 48

Table 2.2 The annual of bulk amorphous alloy was first published... 49

Table 2.3 Summary of △Tx, Trg, γ, critical cooling rate Rc and critical section thickness Zc for typical bulk amorphous alloys... 50

Table 2.4 The exponent n for the JMA equation for different kinds of crystallization mechanisms... 51

Table 3.1 The alloy systems selected in this study... 55

Table 4.1 Summary of the density for various alloy combinations……… 53

Table 5.1 Summary of bonding parameter ∆x and δ for Mg-Cu-Y-B alloys………. 54

Table 5.2 Summary of Tg , Tx , ∆Tx and γ value for the Mg-Cu-Y-B amorphous alloys at a heating of 20 K/min……….. 55

(6)

Figures List

Figure 1.1 The atomic arrangements of crystal and amorphous alloy... 56

Figure 2.1 Schematic drawings of four quenching methods from the melt: (a) piston and anvil, (b) melt spinning, (c) melt extraction, and (d) twin roller quenching... 57

Figure 2.3 Relationship between the critical cooling rate (Rc), the maximum sample thickness (tmax) and reduced glass transition temperature (Tg/Tm) for bulk amorphous systems...58

Figure 2.4 Relationship between the critical cooling rate (Rc) and the maximum sample thickness (tmax) for bulk amorphous alloy systems...58

Figure 2.5 The parameter of γ for glass forming ability ... .59

Figure 2.6 The atomic arrangement of the different atomic size system...60

Figure 2.7 The three empirical rules was proposed by Inoue... 61

Figure 2.8 The relationship between the bond parameters, including electronegativity difference (∆X) and atomic size parameters (δ), and the widths of supercooled lliquid region (∆Tx) in Mg-Based bulk metallic glasses...62

Figure 2.9 Relationship between the Young’s modulus, tensile strength and hardness for various metallic glass systems... .63

Figure 3.1 Flow chart showing the experiment procedures...64

Figure 3.2 Photograph of the arc melting device...65

Figure 3.3 Photography of the melt spinning device... 66

Figure 3.4 The DSC trace for amorphous alloy at heating... 67

Figure 3.5 The DSC programs for (a) non-isothermal, and (b) isothermal analyses... 68

Figure 4.1 The experiment procedures for Mg-based amorphous alloys……….………….. 69

(7)

Figure 4.2 Surface appearance of melt spun Mg-Cu-Y-B amorphous alloy……… 70

Figure 4.3 Bending behavior of the Mg-Cu-Y amorphous alloy………. 70

Figure 4.4 X-ray diffraction patters for melt spun Mg65Cu25Y10-xBx alloy……….. 71

Figure 4.5 X-ray diffraction patters of the melt spun Mg65Cu25B10 alloy……….71

Figure 4.6 X-ray diffraction patters for melt spun Mg65Cu25-xY10Bx alloy……… 72

Figure 4.7 SEM/EPMA micrographs of the Mg65Cu25Y10 metallic glass: (a) BEI image (b) mapping image.……… 73

Figure 4.8 Hardness as function of B content for Mg65Cu25-XY10BX alloys……….. 74

Figure 4.9 SEM/BEI micrograph showing the shear bands in the Mg-Cu-Y-B alloy…….. 74

Figure 4.10 DSC curves of the amorphous Mg65Cu25Y10 alloy………. 75

Figure 4.11 DSC curves of the amorphous Mg65Cu22Y10B3 alloy………. 75

Figure 4.12 DSC curves of the amorphous Mg65Cu20Y10B5 alloy………. 76

Figure 4.13 DSC curves of the amorphous Mg65Cu15Y10B10 alloy……… 76

Figure 4.14 DSC curves of the amorphous Mg65Cu25-xY10Bx (x = 0, 3, 5, 10) alloy at a heating rate of 20 K/min……… 77

Figure 4.15 Glass transition and crystallization temperature as a function of B content for Mg65Cu25-xY10Bx alloys………. 78

Figure 4.16 DSC curves of the amorphous Mg65Cu25Y10-xBx (x = 0, 3, 5, 10) alloy at a heating rate of 20 K/min……… 79

Figure 4.17 DSC curves of the amorphous Mg65Cu25Y7B3 and Mg65Cu22Y10B3 alloys at a heating rate of 20 K/min……… 80

Figure 4.18 DSC traces showing the melting endotherms obtained from the Mg65Cu25Y10-x Bx (x = 3, 5) alloy at a heating rate of 20 K/min………... 81

Figure 4.19 Kissinger plot for melt spun alloys (a) Mg65Cu25Y10 (b) Mg65Cu22Y10B3…... 82 Figure 4.20 Kissinger plot for melt spun alloys (a) Mg65Cu20Y10B5 (b)Mg65Cu15Y10B10… 83

(8)

Figure 4.21 The extraction of the n value from the modified non-isothermal equation for Mg65Cu25Y10……… 84 Figure 4.22 The extraction of the n value from the modified non-isothermal equation for Mg65Cu22Y10B3.……….. 84 Figure 4.23 Extraction of the activation energy from the modified non-isothermal DSC scan for Mg65Cu25Y10..………... 85 Figure 4.24 Extraction of the activation energy from the modified non-isothermal DSC scan for Mg65Cu22Y10B3..……….... 85 Figure 4.25 Extraction of the activation energy from the modified non-isothermal DSC scan for Mg65Cu20Y10B5..……….... 86 Figure 4.26 Extraction of the activation energy from the modified non-isothermal DSC scan for Mg65Cu15Y10B10.………...……….86 Figure 4.27 DSC trace of isothermal annealing for Mg65Cu25Y10 amorphous alloy……… . 87 Figure 4.28 DSC trace of isothermal annealing for Mg65Cu22Y10B3 amorphous alloy…… . 87 Figure 4.29 PeakFit software for Mg65Cu25Y10….……….… 88 Figure 4.30 PeakFit software for Mg65Cu22Y10B3.……….… 88 Figure 4.31 The plot of transformed volume fraction versus time for Mg65Cu25Y10….…… 89 Figure 4.32 The plot of transformed volume fraction versus time for Mg65Cu22Y10B3.…… 89 Figure 4.33 The variation of Avrami exponent n for Mg65Cu25Y10..……… 90 Figure 4.34 The variation of Avrami exponent n for Mg65Cu22Y10B3.……….. 90 Figure 4.35 Extraction of isothermal activation energy of the first crystalline phase for the Mg65Cu25Y10 amorphous alloy…... …….…... 91 Figure 4.36 Extraction of isothermal activation energy of the first crystalline phase for the Mg65Cu22Y10B3 amorphous alloy…...…….…... 91 Figure 4.37 Extraction of isothermal activation energy of the second crystalline phase for the Mg65Cu22Y10B3 amorphous alloy.……….…….92

(9)

Figure 4.38 The TEM images of the as-melt-spun Mg65Cu22Y10B3 amorphous alloy:

(a) Bright field image and (b) Selected area diffraction pattern……….. ...93

Figure 4.39 The TEM images of the yttrium oxide particle in the as-melt-spun Mg65Cu22Y10 B3 amorphous alloy: (a) Bright field image and (b) Selected area diffraction pattern.………. ...94

Figure 4.40 The TEM images of the crystallization in the Mg65Cu22Y10B3 amorphous alloy annealed to 670 K then cooled to room temperature: (a) Bright field image, (b) Selected area diffraction pattern, and (c) EDS analysis... ………. ...95

Figure 4.41 The TEM images of the crystallization in the Mg65Cu22Y10B3 amorphous alloy annealed to 670 K then cooled to room temperature: (a) Bright field image, (b) Selected area diffraction pattern, and (c) EDS analysis... ………. ...96

Figure 4.42 The TEM images of the crystallization in the Mg65Cu22Y10B3 amorphous alloy annealed to 670 K then cooled to room temperature: (a) Bright field image, (b) Selected area diffraction pattern, and (c) EDS analysis... …………... ...97

Figure 4.43 The TEM images of the crystallization in the Mg65Cu22Y10B3 amorphous alloy annealed to 670 K then cooled to room temperature: (a) Bright field image, (b) Selected area diffraction pattern, and (c) EDS analysis...………. ...98

Figure 4.44 The TEM images of the crystallization in the Mg65Cu22Y10B3 amorphous alloy annealed to 670 K then cooled to room temperature: (a) Bright field image, (b) Selected area diffraction pattern, and (c) EDS analysis...………. ...99

Figure 4.45 The TEM images of the crystallization in the Mg65Cu22Y10B3 amorphous alloy annealed to 670 K then cooled to room temperature: (a) Bright field image, (b) Selected area diffraction pattern, and (c) EDS analysis...…... .100

Figure 5.1 Bonding parameter for the Mg-based alloys against boron content...…... .101

Figure 5.2 The overall crystallization for the Mg65Cu22Y10B3………...102

Figure 5.3 The overall crystallization for the Mg65Cu15Y10B10………. ... .102

(10)

Figure 5.4 Extraction of non-isothermal activation energy for the Mg65Cu25Y10...103 Figure 5.5 Extraction of non-isothermal activation energy for the Mg65Cu22Y10B3...103 Figure 5.5 Extraction of non-isothermal activation energy for the Mg65Cu20Y10B5...104

(11)

論文摘要

近幾十年來,多元非晶質合金由於具有一些獨特的物理特性及化學特性,已引起眾 多學者的研究與探討。而在這些多元非晶質系統中,鎂基非晶質合金因具有高強度重量 比,低玻璃轉換溫度及輕密度等特性。因此在輕金屬應用中,佔有相當大的發展潛力及 優勢。

本研究以在鎂基非晶質合金之中,具有高玻璃形成能力的Mg65Cu25Y10非晶質為母

材,藉由添加小原子尺寸的硼探討對母材熱性質、非晶質化及奈米結晶行為的影響。在 實驗流程上,首先利用小型電弧爐及高週波爐熔鍊得到厚度約 100~200 µm鎂基非晶質 薄帶,分別為Mg65Cu25-xY10Bx(x = 1, 3, 5, 10)及Mg65Cu25Y10-xBx(x = 1, 3, 5, 10)兩組系統,

再利用XRD、TEM來觀察合金的非晶質化與結晶結構及DSC恆溫、非恆溫的結晶動力學 研究與分析。

由非恆溫實驗結果發現,Mg65Cu25-xY10Bx(x = 1, 3, 5, 10)此組其結晶所需活化能經由 Kissinger方程式計算,可得知從 138 kJ/mol提高至 156 kJ/mol,表示硼的添入有增加其母

材熱穩定性;且∆Tx與 γ 值於Mg65Cu22Y10B3合金成分時會有最大值分別為 66 K及

0.420。而在恆溫實驗結果,經由Arrhenius 方程式計算也可推得其結晶所需活化能由原 本母材的 150 kJ/mol 提升至Mg65Cu22Y10B3的 210 kJ/mol。推測由於硼的添加阻礙的原子

間的擴散行為,進而增加第一個結晶相(Mg2Cu)產生所需克服的活化能。此結果也與XRD

所得結果相符,隨著硼元素的含量的增加,Mg2Cu將會被抑制產生。

此外添加硼元素之後,Mg65Cu25-xY10Bx非晶相的硬度值也由 220Hv提高至 320Hv, 而Mg65Cu25Y10-xBx非晶相的硬度值則由 220Hv提高至 290Hv,可推測硼原子的添入增加 了結構的緻密度使得硬度質有效的提昇。

(12)

Abstract

The glass forming ability (GFA) of the lightweight Mg65Cu25Y10 alloy has been widely studied.This alloy contains a medium sized Mg matrix (0.16 nm in atomic radius), a small sized Cu (0.14 nm) and a large sized Y (0.18 nm). The glass transition temperature Tg, supercooled temperature range ∆T, and the glass forming γ index are 425 K, 54 K, and 0.40, respectively. Follow-up studies have applied other large sized elements, such as Tb (0.178 nm) and Gd (0.18 nm), to replace Y, and resulted in satisfactory performance. It has recently suggested that the addition of the even smaller sized B (0.08 nm) in the Zr based bulk amorphous alloys to a certain level can further enhance GFA and provide the chance in fabricating larger bulk amorphous billet, partly due to the extra-small atomic size to complicate the atom arrangements and partly due to the absorption of oxygen by B. The current study extends the concept of adding B to the Mg65Cu25Y10 based alloy, including the replacement of Y or Cu by 1-10 at% (1, 3, 5, and 10%), using the drop casting and melt spinning methods. The amorphous alloys are further subject to thermal annealing at temperatures within the supercooled temperature region in DSC, and the data are analyzed in terms of their thermal stability and crystallization kinetics under the isothermal or non-isothermal DSC conditions.

In the non-isothermal analyses, the activation energy of the Mg65Cu25Y10 amorphous alloy for crystallization determined by the non-isothermal Kissinger method is 138 kJ/mol.

With increasing B content, it can be promoted to a high energy barrier against crystallization, and Mg65Cu22Y10B3 has a higher value of 156 kJ/mol. T

As for the isothermal analyses, the activation energy values of the Mg65Cu25Y10 and Mg65Cu22Y10B3 amorphous alloys for crystallization determined from the Arrhenius equation

(13)

are 150 and 200 kJ/mol, respectively. It means that boron can restrain the growth of Mg2Cu.

The Mg65Cu22Y10B3 amorphous alloy has the higher activation energy (210 kJ/mol) for crystallization, suggesting that the B diffusion away and the Cu diffusion toward the Mg2Cu nuclei are also important.

(14)

Chapter 1

Introduction and literature review

1.1 Introduction

Metallic glasses or amorphous metals, simply speaking, are referred to the systems lack of the crystalline long range order and exhibiting the liquid-like atomic arrangements as shown in Fig. 1.1 [1]. They have also been called liquid metals or non-crystalline metals alternatively.

Owing to the unique structure, metallic glass alloys show characteristic physical features such as high strength, corrosion resistance and electromagnetic properties, which are different from the corresponding crystalline alloys. The characteristics of metallic glasses are summarized in Table 1.1. Furthermore, in recent years more and more researchers worldwide discover better ways to synthesize the amorphous alloys using much lower cooling rates and new combinations of alloy compositions for glassy metals. It is believed that the metallic glasses will exhibit their significance for both basic research and application in the future.

1.2 The development of Mg-based amorphous alloys

Among a large number of amorphous alloys, Mg-based alloys have attracted much attention especially due to its high strength to weight ratio and a low glass transition temperature. Before the new Mg-based amorphous alloys with high tensile strength and good ductility were found in 1988 [2], the glass formation of Mg-based amorphous was limited to the Mg-Zn [3] and Mg-Cu [4] binary systems. However, the Zn and Cu concentrations were added to a level high as 20-40% and the resulting melt-spun ribbon would be highly brittle.

Little has been known about the Mg-based amorphous alloys with high strength.

(15)

Until 1991, Inoue et al. [5] succeeded in finding new Mg-based amorphous alloys, such as Mg-Ln-TM (Ln=lanthanide metal; TM=transition metal) systems with high tensile strength.

They also fabricated bulk metallic glass (BMG) with a diameter of 4 mm [6] by injection casting Mg-Cu-Y alloy into a Cu mold. Next year, Inoue’s group used the high-pressure die casting method and increased even more the diameter to 7 mm [7]. Their contribution opened the door to design new families of light amorphous alloy systems.

1.3 Influence of quaternary additions on Mg-based alloys

Inoue et al. [8, 9,10] have reported the relationship between thermal stability of amorphous phase and the topological short range ordering. It is thought that the increase in packing density can cause an increase in thermal stability. The follow-up investigations on the Mg-based BMG have been directed to the addition of the quaternary element to improve the glass forming ability (GFA).

Recently, Park et al. [11] reported the effect of Ag as the substituting element for Cu on the GFA of Mg-Cu-Y-Ag. Due to the large difference in atomic size between Ag and the constituting elements, the larger mixing enthalpy of Ag-Mg and Ag-Y than that of Ag-Cu can contribute to increase the atomic packing density of the liquid structure and GFA.

Men et al. [12] also investigated the effect of substitution of Y in the Mg65Cu25Y10 alloy with Gd on the GFA and crystallization behavior. Because of the difference in electronic configuration between Y (4d15s2) and Gd (4f75d16s2), it results in a certain change of the short-range order in the undercoolded liquid of Mg-Cu-Y-Gd alloys, and shows a surprising high GFA. Other cases, for example, the Mg65Cu20Zn5Y10 [13] and Mg65Cu15Ag5Pd5Y10 [14]

alloys also exhibit high GFA and enable to fabricate metallic rods with diameters of 6 and 7

(16)

mm by a Cu-mold injection casting.

1.4 The purpose of this study

From the bibliography of Zr-Al-Ni-Cu amorphous alloy, by adding the small atom beryllium to substitute the zirconium not only decreases the excess free volume but also increases the disorder arrangement. It is apparent that GFA is improved from the DSC results and other measurements.

From the same view of topology, by the addition of the boron, it can also cause a large atomic difference in atomic size in the Mg-Cu-Y-B system and will generate strongly negative

∆Hm for Mg-B and Y-B. Also, the addition of boron may fill the liquid free volume and increase the density of randomly packed liquids; both will further improves the GFA and thermal stability against crystallization.

(17)

Chapter 2 Background

2.1 The evolution of amorphous alloys

Since the amorphous phase was first fabricated in the Au-Si system by rapid quenching from the melt in the early sixties [15], a great number of scientists have paid much attention to fabricate the new amorphous alloy systems in the bulk form. However, because of the high quenching rate higher than 107 K/s, only thin ribbons, foils or sheets with a thickness less than 100 µm could be fabricated.

Until 1976, the first bulk metallic glass, Pd-Ni-P [16] alloy, was fabricated by a simple suction casting method. More and more amorphous alloys were synthesized by various preparation methods. In 1980s, a variety of solid-state amorphization techniques, which are completely different mechanisms from rapid quenching as shown in Fig. 2.1, such as mechanical alloying, diffusion induced amorphization in multilayer, and ion beam mixing were developed. The diameter of bulk metallic glasses were also raised to the millimeter scale.

In the late 1980s, Inoue et al. [17] in Tohoku University of Japan succeeded in finding new multicomponent alloy systems consisting mainly of common metallic elements and proposed the empirical rules to form the bulk amorphous alloys with high GFA and lower critical cooling rate (Rc). According to the empirical rules, a number of bulk amorphous alloys with much lower Rc have been developed in multicomponent alloy systems and these new multicomponent alloy systems have enabled to produce the bulk amorphous alloys with a thickness range up to about 72 mm for the Pd-Cu-Ni-P alloy [18] by conventional solidification

(18)

methods.

But owing to the cost and heavyweight of the Pd alloys, recent years, the lightweight glassy metals such as the Zr, Ti, Mg, Al base alloys with the aim of meliorating the specific strength were the developing emphasis. In addition, the alloys with the mixture of amorphous and nano-crystalline phases prepared from the crystallization of the original amorphous alloys have also attracted attention due to the various characteristics [19]. It is generally believed that the amorphous or the nanocrystalline alloys will own highly developing potential in the future.

2.2 The systems of glassy metals

The systems of glassy metals component can be divided into five major categories as shown in Table 2.1 [20]. The binary alloys consisting of metal and metalloid (such as Si, B, P) are the most popular glassy metals before 1988. But the disadvantages of these binary alloys are the very high cooling rate (106-108 K/s) to be vitrified and the small production size. Until the noble metal base amorphous alloys, such as Pd-Ni-P and Pt-Ni-P alloys with lower critical cooling rates of the order of 103 K/s [21] and larger size [22], were announced, the development of amorphous alloys makes a big breakthrough. Recently, the lightweight Mg-Ce-Ni amorphous with much lower cooling rates (below 103 K/s) was proposed by Inoue et al.. This kind of glassy metal has very high tensile fracture and large GFA. After that, the Mg-TM-Ln, Ln-Al-TM, Ti-Zr-TM and Zr-Al-TM amorphous alloy systems with great GFA and unique properties have been developed gradually. More recently, researchers focused on the multicomponent alloy systems, such as the Zr [23], Ti [24], Mg [25], Al [26] base alloys, because these alloy systems could be vitrified into a bulk form by casting with a much lower cooling rate of about 10 K/s and their supercooled region is wide. Table 2.2 [9] summarizes the typical

(19)

bulk amorphous alloys systems reported since 1988 up to date, and they can be divided into the nonferrous and ferrous alloy systems.

2.3 The dominant factors for the GFA

The glass forming ability (GFA), as related to the ease of devitrification, is very crucial to understand the origins of glass formation and can act as a guideline for exploring new bulk glassy compositions. Initially, the GFA of an alloy was determined by the Rc. The high cooling rate was used to suppress the occurrence of the crystallization and to obtained the fully amorphous alloys. However, the exact value of critical cooling rare of some vitrification methods is not easily to measure. As a result, many other criteria have been proposed to reflect the GFA of the amorphous alloys in the basis of the characteristic temperature measured by differential thermal calorimetry (DSC) or differential thermal analysis (DTA).

The most extensively used are the reduced glass transition temperature, Trg[27] (= Tg/Tl, where Tg is the glass transition temperature and Tl is the liquidus temperature, respectively) and the supercooled liquid region, ∆Tx (= Tx-Tg, where Tx is the onset crystallization temperature). Lu et al. [28] also show a better correlation with GFA by Tg/Tm for BMGs (Tm is the solidus temperature), but usually we take ∆Tx and Trg as indicators of the GFA for metallic glasses. There is a clear tendency for the Rc to decrease with increasing Trg, as shown in Fig.

2.3.

From the Trg (= Tg/Tl) view of the point, Tl is dependent on composition, choosing the suitable concentration near the eutectic point so that the probability of being able to cool fast through the supercooled liquid region without any crystallization, i.e. GFA is increased. As to another parameter ∆Tx, the thermal ability also can be increased with increasing Tx. Figure 2.4

(20)

reveals that the Rc also decreases with increasing supercooled liquid region, so the synthesized size of glassy metal will be increased.

Although both Trg and Tx are used as indicators of the GFA, they sometimes show the contrasting trends versus GFA in some alloy systems [29, 30, 31]

. Therefore, a new parameter, γ (= Tx/(Tg+Tl)), is proposed recently by Lu and Liu [32]. The definition of γ value is shown in Fig. 2.5. Table 2.3 [32] also reveals Trg ,Tx,γ for various alloy systems. Furthermore the Rc (K/s) and the critical section thickness Zc (mm) can be expressed by the following equations,

Rc=5.1x1021exp(-117.19γ), (1)

and Zc=2.8x10-7exp(41γ). (2)

These two equations reveal much better accuracy than Trg, and have been successfully applied to the bulk glass formation of the Pd-Ni-Fe-P system.

2.4 The main empirical rules for the synthesization of amorphous alloys

Over the past decade, various methods developed to produce bulk metallic glasses were mostly empirical in nature, but researchers gradually began to understand that that the correct choices of elemental constituents would lead to amorphous alloys exhibiting critical cooling rates as slow as 1-100 K/s. One of the general guiding principles designing alloys to form bulk metallic glasses was proposed by Inoue. He summarized the features of alloy components for reduced instability of supercooled liquid and proposed three empirical component rules [33, 34, 35]

to fabricate metallic alloys with high GFA: (1) multicomponent consisting of more than three elements, (2) significantly different atomic size mismatches exceeding 12% among the main three constituent elements, and (3) negative heats of mixing

(21)

among the main elements. The alloys satisfying the rules and having special atomic configurations in the liquid state are significantly different from those of the corresponding crystalline phases. The empirical rules can also be explained by the thermodynamics, kinetics and structure aspects, as described below.

Firstly, from the thermodynamic point of view, it starts from the generally known relation of ∆G =∆Hmix – ∆Smix. The low ∆G value can be obtained in the cases of low ∆Hmix and large ∆Smix. The large ∆Smix is expected to be obtained in multicomponent alloys systems because ∆Smix is proportional to the number of component. Besides, the low ∆Hmix, i.e. the large negative heat of mixing among the three main constituent elements, is the main factor.

The large negative heat of mixing is referred to the fact that the distinct atoms tend to bond together. Hence if there are three different elements mixed together and the difference in atomic size is large, the atomic configuration tends to appear as high dense random packing as shown in Fig. 2.6.

From the kinetic point of view, in 1969, Turnbull [36] has studied the relation between the nucleation rate and GFA in the supercooled region. He proposed the homogeneous nucleation and growth of a crystalline phase with a spherical morphology from the supercooled liquid can be expressed by the following relationships:

( )

⎥⎥

⎢⎢

⎡ α β

= η 2

r r 3 30

T - 1 T

b -

I 10 (cm-3s-1), (3)

and

( )

⎥⎦

⎢ ⎤

⎡ β

= η

r r 2

T T - exp 1 - f 1

U 10 (cms-1), (4)

where Tr is the reduced temperature (= T/Tm; Tm is the melting point), b is shape factor and is

(22)

16π/3 for the spherical nucleus, η is viscosity and f is the fraction of nucleus sites at the growth interface. α and β are dimensionless parameters related to liquid/solid interfacial energy, σ, and can be expressed as α= (NoV)1/3 σ/∆H and β= ∆S/R. Here, No, V and R are the Avogadro number, the atomic volume and the gas constant, respectively. In these relationships, the most important parameters are η, α and β. The increase of the three parameters will decrease the I and U values, and further leading to an increase of GFA. Moreover, the increase of α and β also implies the increase in σ and ∆S and the decrease in ∆H, consistent with the interpretation of achieving high GFA derived from the thermodynamic point of view.

Furthermore, the atomic structure development also plays an important role for the mechanism of amorphous phase forming. Form the anti-Hume-Rothery criterion, the difference in atomic size ratios above 15% will induce a dense random packing in the supercooled region. The difficulty of atomic rearrangements attributes the formation of an amorphous phase and impedes the formation of a crystalline alloy.

Form the above described, the new amorphous alloys following these empirical rules must have a higher liquid/solid interfacial energy, leading to the suppression of nucleation of crystalline phase as well as the difficulty of atomic rearrangements. The latter contributes the low atomic diffusivity and high viscosity, leading to the difficulty of growing a crystalline phase. These mechanisms for the reduced instability and the formation of bulk glassy alloys are shown in Fig. 2.7.

Furthermore, there are many other empirical criterions for the glass forming ability. Shek et al. [37] have proposed that the composition of the amorphous alloys and their crystalline counterparts should possess a constant value of e/a. Fang et al. [38] have also proposed an empirical criterion between bond parametric functions that comprises of electronegativity

(23)

difference and atomic size parameters to the relationship and the widths of the supercooled liquid region of Mg-Based bulk metallic glasses. It will be discussed thoroughly in the next section.

2.4.1 Bonding parameter for Mg-based BMGs

In order to further clarify the interatomic effect on GFA and thermal stability for the BMGs, the electronegativity difference (∆x) and atomic size parameter (δ) of multicomponent have also been introduced to explain the GFA of the Mg-based alloys, which show a nearly linear relationship with GFA [38]. The relationship is summarized in Fig. 2.8. The electronegativity difference ∆x and atomic size parameter δ among the elements of a multicomponent alloy are defined by

( )

=

=

n

1 i

2 i

i x x

C

x , (5)

and

=

⎟⎠

⎜ ⎞

⎝⎛ −

=

δ n

1 i

2 i i

r 1 r

C , (6)

where Ci is the atomic percentage of the element i in the alloy; xi and ri are the Pauling electronegativity and covalent atomic radius of the element i, respectively; x and r are the arithmetical mean value of electronegativity and covalent atomic radius for a compound, respectively, which can be calculated as following equation:

=

= n

1 i

i

i x

C

x , (7) and

=

= n

1 i

i i r C

r . (8)

(24)

In general, the larger the electronegativity difference ∆x and the atomic size parameter δ would be, the higher the glass forming ability it becomes. The accuracy for judging the GFA of Mg-Cu-Y-B amorphous systems will be discussed in Chapter 5.

2.5 Characterization of amorphous alloys

The metallic glasses have various properties which are superior to the traditional crystalline materials due to their unique structure. And the unique properties are described below.

2.5.1 Mechanical properties

Due to the dense packing structure and irregular rearrangement of atoms, when the stress is applied, the amorphous alloys only allow small displacements of atoms to resist deformation without the invoke of dislocations as for the case of crystalline metals. Therefore, the amorphous alloys show higher tensile strengths than the traditional crystalline alloys, as shown in Fig. 2.9.

Moreover, the most useful characteristic of metallic glasses is their ability to be shaped and formed in the supercooled liquid region. This characteristic is related to superplastic behavior due to the Newtonian viscous flow. By utilizing the ideal superplasticity in supercooled liquid region, Zr-Al-Ni-Cu alloys [39] exhibit an extremely large elongation of 1.5 x 106 %. Furthermore, the strain rare sensitivity exponent (m-value) of amorphous alloys in supercooled region can reach 1.0 [40] under the deformation condition obeying the Newtonian flow behavior.

(25)

2.5.2 Magnetic properties

Permeability stands for the soft magnetic property of the material. Because the greater part of the magnetic materials is the pieces form, such as silicon steel lamination and wave filter, it is easy to fabricate the amorphous ribbons by melt spinning. Moreover, some amorphous alloys exhibit great permeability and minimum magnetic energy loss. Hence, the application of this kind of magnetic amorphous alloy has been very wide [41]. The typical magnetic amorphous alloy systems can be divided into two categories: TM-M and TM-TM (TM: Fe, Co, Ni, Zr, Hf, etc; M: B, C, Si, P, Ge, etc).

2.5.3 Chemical properties

The very high corrosion resistance has been observed in some amorphous alloys due to the homogeneous single-phased structure which is lack of grain boundaries, dislocations, and the other crystal defects. In addition, one can improve the corrosion resistance of amorphous alloys by adding some kinds of corrosive solute. For instance, the corrosion resistance of Zr-Al-Ni-Cu amorphous alloy increases as adding the corrosive solute, such as Nb, Ta, Ti, and Cr [42]. The other amorphous alloys, such as Pd-based as well as Fe-based [43] amorphous alloys, also have the highest corrosion resistance and they can serve as practical corrosion resistance materials.

2.5.4 Other properties and behavior of bulk amorphous alloys

For application of bulk metallic glasses as industrial materials, it is necessary to establish the bonding technology of BMGs. The major contribution in this aspect is made by Kawamura et al. [44]. They have reported that different bulk metallic glasses were successfully

(26)

welded together by friction welding. Besides, the glassy state can be retained after the processes, and no crystallization and defect was observed in the interface. It is confirmed that, with careful control of the process parameters, the joints remain amorphous and the strength is comparable to the bulk of the metallic glasses.

2.6 Crystallization transformation of amorphous alloys

The amorphous solid is thermodynamically metastable, due to the Gibbs free energy difference between the amorphous and the crystalline states, the amorphous state will transform to the crystalline state under appropriate conditions (i.e., heat treatment, irradiation, or mechanical attrition). In general, the crystallization of an amorphous alloy takes place in there modes which are polymorphous, eutectic and primary crystallization.

Upon polymorphous crystallization, the crystallization of an amorphous phase transform to a crystalline phase without any change in the composition of that phase. This transformation may produce a single compound phase or a supersaturated solid solution phase.

Upon eutectic crystallization, crystalline phases of different compositions form simultaneously. This reaction has no concentration difference across the reaction front, but takes longer than the polymorphous crystallization to proceed because the two components have to separate by diffusion into two separate phases within the crystallized region. Upon primary crystallization, a phase of the alloy constituents is first to crystallize. This dispersed primary crystallized phase coexists with the amorphous matrix and may serve as nucleation sites for the secondary or tertiary crystallization.

2.6.1 The Kissinger plot

(27)

Activation energy is one of the important parameters describing the transformation kinetics. In order to estimate the activation energy for crystallization, the Kissinger relation is frequently used for non-isothermal heating DSC analysis, and can be expressed in the form as

[45,46]

+

⎟=

⎜⎜

⎛ φ

2 p p

m

RT Q T

ln constant , (9)

where φm is the heating rate, Tp corresponds to the peak temperature of transition in DSC curves, R is the gas constant and Q is the transition activation energy. By plotting ln(φm/ Tp2) against 1/ Tp, one can derive the value of activation energy Q from the slope of the straight line plotted.

2.6.2 Modified kinetics of non-isothermal crystallization kinetics

Due to the convenience and usefulness of Kissinger and Ozawa [47] equation, Kissinger plot and Ozawa plots have been extensively used to determinate the activation energy of crystallization. As a matter of fact, these methods could not clearly interpret the crystallization of amorphous materials and the physical meaning of the activation energy because crystallization involves both the nucleation and growth processes. Furthermore, the Kissinger equation is a series expansion of Avrami equation, it is not appropriate due to the fact that the temperature changes continuously under the non-isothermal condition.

Until 1979, Matusita and Sakka [48] proposed a method for analyzing the non-isothermal crystallization kinetics on the basis of nucleation and growth processes. By using this modified equation to express the non-isothermal crystallization, the results were in agreement

(28)

with those obtained isothermally and the processes of nucleation and growth could be modeled separately from the modified plots. The details are described below:

The radius, r, of a crystal particle was expressed as

∫ ( )

= φ

⎛ −

= t T

T0 dT

RT exp Q dt U

T U r 0

0 . (10)

This integral could be evaluated to a closer approximation by using the Doyle’s p function.

( )

=

y 2 dy y

) y y exp(

p , (11)

where y equals Q/RT. Equation (11) could be written as

⎛−

⎟=

⎜ ⎞

T

0 dT

RT exp Q Q R RT

p Q . (12)

If the y value is larger than 20, equation (11) could be expressed to an approximation as

y 4567 . 0 315 . 2 ) y ( p

log =− − . (13)

Combined equations (11), (12) and (13), r could be expressed as

⎟⎠

⎜ ⎞

= φ

RT 1.052 Q - Cexp

r . (14)

Assuming at the early stage, crystal particles grow three dimensionally, the variation of crystal volume fraction, X, was expressed by

(29)

dt r dr X)N4 - dt (1

dX 2

π

= , (15) where (1-X) is the correction factor for the impingement of crystal particles, and the reduction of glass phase. Integration of equation (15) leads to –ln(1-X)=(4π/3)Nr3 and replacing with equation (14) leads to

⎟⎠

⎜ ⎞

⎛ ×

φ

= RT

3 Q 1.052 - exp CN X) - ln(1

- -3 , (16)

when the metallic glass is heated at a constant rate φ, the number of nuclei formed per unit volume is inversely proportional to the heating rate, N= No/φ, and equation (16) could be converted to

⎟⎠

⎜ ⎞

⎛ ×

φ

= RT

3 Q 1.052 - exp CN

X) - ln(1

- 0 -4 . (17)

In more general expression, equation (17) should be expressed

⎟⎠

⎜ ⎞

⎛ ×

φ

= RT

m Q 1.052 - exp CN

X) - ln(1

- 0 -n . (18)

Here, n and m are the numerical factors depending on the nucleation process and growth mechanism, respectively. The value of m=3 indicates that the growth of crystallization is three-dimensional growth. And the m=2 and m=1 are two-dimensional and one-dimensional growth, respectively. Taking the logarithm of equation (18) twice yields

[ ]= φ +

RT 1.052m Q -

nln X) - ln(1 -

ln constant . (19)

(30)

The n value could be obtained from the plot of ln[-ln(1-X)] against lnφ at a specific temperature. And the value of -1.052mQ/R also could be obtained by plotting ln[-ln(1-X)]

versus reciprocal temperature for several heating rates. Furthermore, determining the m value was to observe the change of n in the early crystallization behavior. If n did not change with heating, a large number of nuclei formed already in the specimen and n = m. If n decreased with heating indicates that nuclei formed during the heating. In this case, m<n≤m+1. Normally, the maximum value of n is 4 and the minimum value is 1.

2.6.3 Isothermal analysis for the crystallization kinetics of amorphous alloys

Because the crystallization activation energy calculated by the Kissinger equation cannot reflect the stability of the amorphous phase, the further kinetic analysis of crystallization was carried out by isothermal DSC. The apparent transformation kinetics is usually described by using the Johnson-Mehl-Avrami equation (JMA equation) [49,50,51] measured at different temperatures between Tg and Tx. Furthermore, the JMA equation is derived assuming that the following conditions apply.

(1) Constant radial growth rate;

(2) Constant density and shape of the growing nuclei;

(3) No volume change during phase transformation;

(4) No secondary crystallization.

A general form of JMA equation is given as

{ ( )

kt n

}

exp 1

X= − − . (20)

(31)

Then take the logarithm of equation (20) twice derives equation (21)

[

ln(1 X)

]

nlnk nlnt

ln− − = + , (21)

where X is the volume fraction of transformed, n is the Avrami exponent and k is a kinetic coefficient which is a function of the absolute temperature. The crystallization mechanisms and morphology of crystalline phases can be obtain from the Avrami exponent. For the same n value, it indicates the same crystallization mechanism for the alloy. The values of exponent n for different kinds of crystallization mechanism are listed in the Table 2.4.

2.6.4 The Arrhenius plot

On the other hand, the activation energy can also be evaluated by the isothermal sigmoid curves of crystallized volume fraction versus annealing time at different temperatures. This was carried out by using the following Arrhenius equation [52] to the time required for a fixed amount of transformation X at different temperatures, which is determined by the following equation:

⎟⎠

⎜ ⎞

= ⎛

RT exp Q(X) t

t(X) 0 , (22)

where t(X) is the time required to transform volume fraction X, t0 is a time constant, and Q(X) is the activation energy for transformation. It can plot t(X) against 1/RT and further obtain the activation energy from the slope. The comparison of activation energy between Kissinger plots and Arrhenius plots will be discussed in Chapter 5.

(32)

Chapter 3

Experimental procedures

3.1 Materials

The elements used in this study were purchased from the WESIS company, Taipei, Taiwan. The purity of magnesium is 99.99%, of copper is 99.999%, of yttrium is 99.9% and of boron is 99.5%. The experimental flow chart and constituent contents of the alloys prepared in this study are shown in Table 3.1 and Fig. 3.1, respectively.

3.2 Synthesis methods

3.2.1 Arc melting

In order to prepare the multicomponent amorphous alloys, the arc melting method was first adopted in this study. Pure elemental pieces are used in this route. The applied device is the Centorr Series 5 Bell Jar Single Arc Furnace, as shown in Fig. 3.2. The maximum operating temperature is 3500oC.

Before the starting of arc melting process, because crystallization will result from the oxygen impurity, the argon gas needs to be continuously purged into the furnace to flush the oxygen and maintain a positive argon pressure to ensure a low oxygen vapor pressure environment inside the furnace. Cold water also needs to be poured into and flowing through the bottom of the Cu mold to increase the heat exchange rate. Finally, the chosen component placed on the Cu mold will be melt by the high voltage arc under a Ti getter argon atmosphere.

It should be repeated for more than four times to confirm the homogeneous mixing of

(33)

composition. Besides, the cooling rate of arc process is not fast enough to form a fully amorphous alloys, so it needs to further conduct the drop casting or melt spinning process for complete vitrification of the alloys.

3.2.2 Melt spinning

To avoid the generation of magnesium vapor and the lose of constituent accuracy, the drop casting is not adopted in this study. The last metallic glass alloys will be produced by a single-roller melt spinning method. This method is generally recognized as a convenient way to form the amorphous alloys.

First, the Cu-Y specimen (made by arc melting), magnesium pieces, and the boron element are put into the low carbon steel crucible, and the crucible needs to spray a thin layer of BN to avoid the interaction between alloy and crucible at high temperatures. Next, similar to the arc melting process, the gas flushing step was performed several times before melt spinning. Then the crucible is placed in an induction furnace controlled by the high frequency system, and argon atmosphere is maintained to 2 atm in order to decrease the generation of magnesium vapor.

After complete melting, the liquid alloy will be poured onto the surface of Cu wheel. The wheel is rotated with a high speed of 25 m/s (20 Hz) in order to reach the high cooling rate.

The resulting specimen is thin ribbon type with ~10 mm in width and ~100 µm in thickness.

Small parts of ribbons are sampled for latter property characterizations.

3.3 Phase identifications by XRD

(34)

The vitrification degree of the ribbon structure fabricated by melt spinning can be first identified by the SIEMENS D5000 X-ray diffractometer (XRD) at room temperature. Filtered Cu Kα radiation operated at 40 kV and 30 mA, and 0.02 mm graphite monochrometer are employed. The diffraction angle covers from 2θ = 10o to 80o at scanning rate of 1o per 30 second. Note that a layer of alloys made by the melting spinning around 100~200 µm needs to be ground away to remove the possible oxidized outer surface. After the heat treatments, it also needs to run the XRD to identify the phase of crystallization.

3.4 SEM/EPMA observations

Because of the smallest molecular weight of boron, SEM/EPMA (scanning electron microscope/electron probe microanalyzer) is selected to measure the quantity of chosen components. The sliced specimens from the ribbon need to be coated with a thin gold layer gold, and the JEOL 8900R EPMA equipped with Wavelength Dispersive X-ray Spectrometer (WDS) is used for the qualitative analyses.

3.5 Microhardness testing

The sample hardness testing was conducted using a SHIMADZU HMV-2000 Vicker’s microhardness tester. Because the restrict of ribbon thickness, the samples were only grounded by silicon carbide abrasive papers (#800 and #1000) with water before the hardness measurement. In this testing, the parameter was carried out using a load of 200-500g with 10 second duration, and the hardness value of each sample was average from 10 datum points.

The microhardness data can also be used to investigate whether the boron would cause a dense atomic structure and compared with the Mg-Cu-Y amorphous alloy.

(35)

3.6 Density measurement

Form the simple and reliable points of view, the well-know principle of Archimedes is used as an appropriate means to measure the specimen density. In this technique, the specimen contained in a vessel is weighted in air to record a value of mass W1. Then, the specimen is weighted in a liquid of known density β to record a second value of mass W2. Therefore, the density of amorphous alloy can be evaluated by the following equation :

β

= −

ρ W W /

W

2 1

1 . (23)

The atomic configurations of amorphous alloys are different from those of the crystalline alloys. It is expected that the short-range atomic configuration changes the density of the new amorphous alloys. Normally, the densities would increase when the amorphous phase is transformed through structural relaxation or crystallization.

3.7 DSC thermal stability analysis

The differential scanning calorimeter SETARAM DSC131 is used in this study. It can examine the thermal stability and determine the values of Tg, Tx, Tm by the heating process, as shown in Fig. 3.4. Other data such as nucleation, growth behavior, and activation energy can also be estimated from the isothermal or non-isothermal analysis. The heating rate is set to be 10-40oC/min and each run begins from 40oC and ends at 600oC in this study. Normally, the DSC programs for non-isothermal and isothermal analyses are set as shown in Fig. 3.5.

Moreover, in order to prevent oxidation and magnesium vaporization at high temperatures, samples are placed in stainless steel crucibles and tightly sealed with a nickel ring. The

(36)

pre-existing air in the chamber is flushed by pure Ar and the positive Ar pressure is maintained wile heating.

3.8 Microstructure characterizations

Because the X-ray results are all referred to the averaging measurements for the bulk volume and the SEM results are limited by the image resolution, the spatial distribution between nanocrystalline and amorphous phase needs to rely on the transmission electron microscopy (TEM). Microstructure observations in this study are conducted using the JEOL 3010 analytical TEM operated at 200-300 keV with Link EXL-Ⅱ Energy Dispersive X-ray analysis, (EDS). And the EDS system is used to identify the composition of the observed phase.

The TEM samples were first punched mechanically and then ground by silicon carbide abrasive papers (#1200, #1500, and #2000) with water. Next, the TEM foils need to be adhered to a mesh 200 Cu net. Finally, the TEM foil is thinned by the Gatan 691 precision ion polishing system to produce high quality TEM specimens.

(37)

Chapter 4 Experimental Results

4.1 Specimen preparations

Since the boiling point of Mg is lower than the melting point of Cu, Y and B, respectively, the master ingot of Cu-Y binary alloy was first prepared by arc melting under a Ti-gettered argon atmosphere. Then the ingot was melted with pure Mg and B in an induction furnace under a purified argon atmosphere. Finally, glassy ribbon samples were obtained by rapid solidification of their melts on a single copper roller, as shown in Fig. 4.1. During these processes, how to keep the accuracy of constituent quantity plays a major role on the GFA of the resulting amorphous alloy. Therefore, the crucible is designed to have a cap and to keep minimum the evaporation of the pure Mg component.

The characteristic of BMG is identified as having a mirror like luster on the surface. As shown in Figs. 4.2 to 4.3, the melt spun Mg-Cu-Y-B alloys also have this distinguishing characteristic. Furthermore, the difference in the bending ductility of melt spun alloys can also be differentiated in terms of critical bending angle for fracture. For ductile ribbons, the bending angles can be as high as 180 o or the ribbons can be fully folded. For brittle ribbons, the bonding angles would be less than 90 o. The Mg-Cu-Y parent metallic glass exhibits ductile behavior with a 150o bending angle. With increasing B content, the samples perform more and more brittle tendency and become easily splintered into pieces.

4.2 X-ray diffraction analyses

Figure 4-3 shows the X-ray diffraction pattern of the melt spun Mg65Cu25Y10-xBx (x = 0,

(38)

1, 3, 5) alloys with Y replaced by B. It is apparent that the Mg65Cu25Y10-xBx alloys exhibit an almost fully amorphous phase except for a couple of small peaks superimposed on the amorphous diffused hump located at 2θ ~35o and 37o in the alloy with 5 at.% boron content.

The tiny peaks are identified to be the Mg2Cu crystalline phase with an orthorhombic structure plus minor MgB4, as shown in Fig. 4.4. For a more boron addition to 10at% so as to completely substitute yttrium, or the Mg65Cu25Y10, only the sharp Mg2Cu peaks are visible in the XRD curve, as shown in Fig. 4.5. However, from the Fig. 4.6, only board diffused amorphous hump occurs and no obvious crystalline peak is seen in the melt spun Mg65Cu25-xY10Bx (x = 0, 1, 3, 5, 10) alloys with Cu replaced by B. The above XRD results imply that the partial substitution of copper and yttrium by boron has the different effects on the GFA.

Furthermore, in order to identify the different crystallization phases from the Mg65Cu25-xY10Bx and Mg65Cu25Y10-xBx two systems, the as-melt-spun amorphous ribbons were annealed at various temperatures in DSC and then characterized by X-ray diffraction, as presented in later sections.

4.3 SEM/EPMA observations

In order to confirm the absence of any crystalline phase over the whole sample, a SEM/BEI composition image taken from the polished surface of the Mg-Cu-Y metallic glass is shown in Fig. 4.7 (a). It can be seen that no contrast of precipitation is seen over the whole micrograph. Furthermore, from the SEM/EDS observation, the elements of Cu and Y are uniformly spread in the Mg matrix, as shown in Fig. 4.7 (b). Hence by using melt spinning to solidify, the high cooling rate and uniform composition can be obtained.

(39)

4.4 Microhardness tests

The sample hardness is measured using the HMV-2000 microhardness tester. Ten positions are chosen randomly for each sample, then the Vickers hardness (HV) are obtained by indenting each sample at a load of 500 g and measuring the indentation size by optical microscopy.

Microhardness results of the Mg65Cu25-xY10Bx and Mg65Cu25Y10-xBx amorphous alloy systems are shown in Fig. 4.8. It is apparent that the HV is sensitive to the addition of boron.

With increasing the boron content, higher HV data would be obtained. But the Mg65Cu25-xY10Bx alloys have the higher Hv value than the Mg65Cu25Y10-xBx amorphous alloy.

The highest HV of 321 is obtained for Mg65Cu20Y10B5, which is significant higher than the value of 220 for the parent Mg65Cu25Y10 alloy. The atomic arrangement and whether the nanocrystalline phase exist may be the reason of different for Hv.

Furthermore, when the sample is applied a compression by the indenter, the plastic deformation morphology can be seen in the form of the pile-up of the semi-circular shear bands, as shown in Fig. 4.9. In a crystalline material, these bands extend a short way along the grain boundary then stop when they run into a crystal, but in amorphous substance these bands just keep on growing at stress point. Because the element of copper and yttrium are ductile in the Mg-Cu-Y-B system, lots of shear bands appeared but each one was small and thin. These bands appear to increase to overall toughness by interfering with each other so that no one band can extend into a long crack.

4.5 Density measurements

(40)

Due to the improvement of the hardness in the Mg-Cu-Y-B amorphous alloys, the atomic arrangement is of concern. The simplest method to detect whether the volume is changed is to measure the density. Table 4.1 shows the density results obtained by means of the Archimedean theorem. The Mg65B25Y10 based alloy has the smallest density value 3.09 g/cm3. The density values of Mg65Cu25-xY10Bx (x = 3,5) systems are 3.46 and 3.24 g/cm3, respectively. And the Mg65Cu25Y10-xBx (x = 3,5) systems have the higher density values 3.81 g/cm3 and 3.74 g/cm3, respectively.

4.6 Thermal analyses

4.6.1 Non-isothermal analyses and kinetics

In order to understand the origin of the high thermal stability and excellent glass-forming ability, it is very important to clarify the crystallization behavior of the supercooled liquid. By using the DSC instrument, it is sensitive to detect the glass transition and the nanocrystallization.

In the Mg65Cu25-xY10Bx alloy systems, from the Figs. 4.10 to 4.13, all the samples exhibit a distinct glass transition, followed by a wide supercooled region and then exothermic reactions due to crystallization. Figure 4.14 shows with increasing x from 0 to 3, the glass transition temperature is nearly constant at 410 K but the crystallization temperature increases slightly from 468 to 476 K. Therefore, the ∆Tx value also slightly increases from 58 K at x = 0 to 66 K at x = 3. With further increasing x to the 5 and 10 at%, ∆Tx becomes smaller with the addition of B content, 50 K for 5 at% B and 45 K for 10 at%. Figure 4.15 shows the variations of Tg, Tx against the B content. It is suggested that the B substitution for Cu is of no harm for the GFA of Mg-Cu-Y amorphous alloy.

參考文獻

相關文件

The first row shows the eyespot with white inner ring, black middle ring, and yellow outer ring in Bicyclus anynana.. The second row provides the eyespot with black inner ring

國立高雄師範大學數學教育研究所碩士論文。全國博碩士論文資訊網 全國博碩士論文資訊網 全國博碩士論文資訊網,

港大學中文系哲學碩士、博士,現 任香港中文大學人間佛教研究中心

It is considered that the significant difference in the mechanical properties is due to the discrepancy in the deformation and fracture mechanisms between bulk metallic glasses and

The performances and properties of the resulting PEEK nanocomposites were examined in terms of tensile loading, hardness, dynamic mechanical analysis (DMA), thermal

本研究是以鑄造、粉末冶金、噴覆成型製程將奈米級強化物添加入鎂合金內,且各 製程產生的塊材再經過擠型,針對破壞韌性及其它機械性質作分析。而以強化相的添加 範圍

莊雅芬 國立中山大學 海洋生物科技暨資源學系碩士班丙組 胡介申 國立中山大學 海洋生物科技暨資源學系碩士班 林右晟 國立中山大學 海洋生物科技暨資源學系碩士班 楊士明

In this research, one solid state processing technique, friction stir processing, is applied to modify the AZ61 magnesium alloy billet and to incorporate 5-10 vol% nano-sized ceramic