Fabrication and Characterization on High Performance Mg/Carbon-Fiber/PEEK Laminates and Nanoparticle/PEEK

(1)

國立中山大學材料科學研究所博士論文

高性能鎂/碳纖/聚二醚酮夾層及奈米粉體強化聚二醚酮複材之製備與特性分析

Fabrication and Characterization on High Performance Mg/Carbon-Fiber/PEEK Laminates and Nanoparticle/PEEK

Nanocomposites

研究生：郭木城撰指導教授：黃志青博士

中華民國九十四年一月

(2)

(3)

博碩士論文授權書

（國科會科學技術資料中心版本，93.2.6）

本授權書所授權之論文為本人在國立中山大學材料科學研究所九十三學年度第一學期取得＿博＿士學位之論文。

論文名稱：高性能鎂/碳纖/聚二醚酮夾層及奈米粉體強化聚二醚酮複材之製備與特性分析

■同意 □不同意 (政府機關重製上網)

本人具有著作財產權之論文全文資料，授予行政院國家科學委員會科學技術資料中心(或其改制後之機構)、國家圖書館及本人畢業學校圖書館，得不限地域、時間與次數以微縮、光碟或數位化等各種方式重製後散布發行或上載網路。

本論文為本人向經濟部智慧財產局申請專利（未申請者本條款請不予理會）

的附件之一，申請文號為：＿＿＿＿＿＿，註明文號者請將全文資料延後半年後再公開。

---

■同意 □不同意 (圖書館影印)

本人具有著作財產權之論文全文資料，授予教育部指定送繳之圖書館及本人畢業學校圖書館，為學術研究之目的以各種方法重製，或為上述目的再授權他人以各種方法重製，不限地域與時間，惟每人以一份為限。

上述授權內容均無須訂立讓與及授權契約書。依本授權之發行權為非專屬性發行權利。依本授權所為之收錄、重製、發行及學術研發利用均為無償。上述同意與不同意之欄位若未鉤選，本人同意視同授權。

指導教授姓名:黃志青

研究生簽名: 學號:8936805 (親筆正楷) (務必填寫) 日期:民國 94 年 1 月 25 日

1. 本授權書 (得自http://sticnet.stic.gov.tw/sticweb/html/theses/authorize.html 下載或至http://www.stic.gov.tw首頁右下方下載) 請以黑筆撰寫並影印裝訂於書名頁之次頁。

2. 授權第一項者，請確認學校是否代收，若無者，請個別再寄論文一本至台北市(106) 和平東路二段 106 號 1702 室國科會科學技術資料中心黃善平小姐。(電

(4)

LIST OF TABLES

Table 1.1 Typical properties of metal matrices for metal-metal laminates………...…94 Table 1.2 Comparison of mechanical and physical properties of various materials…….…95 Table 1.3 Values for the lattice spacing constants of the annealed isotropic samples and

other parameters of chain comformations………...……….….…96 Table 1.4 Properties of PEEK and ‘Victrex’ polyethersulphone…………...……….…...…97 Table 1.5 Solubility of PEEK at 25 ^oC………..…98 Table 1.6 Typical values for specific strength and specific stiffness of different materials along the longitudinal (or fiber reinforced) direction……….…..…99 Table 1.7 Selected properties for different types of matrix………..100 Table 1.8 Fibre properties………....101 Table 2.1 Comparison of the weight and volume percentage (wt% and vol%) of the nano

SiO2 and Al2O3 particles added in the PEEK composites. The densities for PEEK, SiO2, and Al2O3 are 1.30, 2.65, and 3.98 g/cm³, respectively…….………102 Table 3.1 Processing conditions of Mg/APC-2 laminated composites………..………….103 Table 3.2 The room temperature mechanical properties along the longitudinal (L) and

transverse (T) directions. The volume fractions of AZ31, carbon fiber, and PEEK in the resulting Mg/CF/PEEK composite are 61%, 24%, and 15% in volume, respectively………..104 Table 3.3 UTS and elongation data obtained at room temperature (25ôC), 100ôC, and 150ôC along the longitudinal and transverse directions………..……...…105 Table 3.4 The wavenumbers of the characteristic group absorptions on the AS-4 prepreg, etched CF-phase, unetched CF-phase, and etched Mg-phase………..……...…106 Table 3.5 Characteristic group absorption wave-numbers of the PEEK polymer…….….107 Table 3.6 The microhardness and tensile data of the nanoparticle-filled PEEK composites.

(9)

The increment percentage of the experimental data with respect to the unfilled PEEK is also included in parentheses ()………..108 Table 3.7 DSC data on the 15 nm silica filled PEEK composites, obtained from the cooling DSC runs. Tci,, Tcp,, and Tcf are referred to the initiation, peak, and finishing temperatures for PEEK crystallization, respectively. t_c is referred to the overall crystallization time………..………....109 Table 3.8 DSC data on the 30 nm silica filled PEEK composites, obtained from the cooling

DSC runs. Tci,, Tcp,, Tcf , are referred to the initiation, peak, and finishing temperatures for PEEK crystallization, respectively. t_cis referred to the overall crystallization time………..…110 Table 3.9 DSC data on the 30 nm alumina filled PEEK composites, obtained from the

cooling DSC runs. Tci,, Tcp,, and Tcf are referred to the initiation, peak, and finishing temperatures for PEEK crystallization, respectively. t_c is referred to the overall crystallization time……….………..111 Table 4.1 Summary of the room temperature mechanical properties along the longitudinal

(L) and transverse (T) directions. The volume fractions of AZ31, carbon fiber, and PEEK in the resulting Mg/CF/PEEK composite are 61%, 24%, and 15% in

volume, respectively………..…… .112 Table 4.2 Comparison of the current Mg laminated composites with other commercial

structural metallic alloys, such as AZ91 Mg, 6061 Al, Ti-6Al-4V and 1040 steel.

……….….113 Table 4.3 Comparison of the room temperature tensile properties of the current Mg

laminated composites along the longitudinal (L) and transverse (T) directions with previously reported data on the ARALL (2024Al/AF/epoxy) and CARALL (2024Al/CF/epoxy)………..114 Table 4.4 The yield stress (YS) and ultimate tensile strength (UTS) of the AZ31 alloy,

(10)

PEEK polymer, and carbon fiber at room temperature (25ôC), 100ôC, and 150ôC ………....115 Table 4.5 Comparison of the theoretical (based on ROM) and experimental UTS values on the Mg/CF/PEEK laminated composites along the longitudinal (L) and transverse (T) directions at room temperature, 100ôC, and 150ôC………...116 Table 4.6 Summary of the room temperature flexural properties along the longitudinal (L) and transverse (T) directions. The volume fractions of AZ31, carbon fiber, and PEEK in the resulting Mg/CF/PEEK composite are 61%, 24%, and 15% in volume, respectively………....117 Table 4.7 Comparison of the theoretically predicted (Theo) and experimentally measured (Exp) mechanical data. The increment percentage of the experimental data with respect to the unfilled PEEK is also included in parentheses ()………..118 Table 4.8 The mean distance L between statistically distributed nanoparticles.

, d, F, and V ]

1 ) /

[( −

=d F V_f

L f are filler diameter, packing factor (0.64 for spherical fillers) , and volume fraction, respectively………..119

(11)

LIST OF FIGURES

Fig. 1.1 Flow chart of the conducting research………..………...120 Fig. 2.1 Microstructure of the as-received AZ31 Mg alloy. The grain size is about 34 µm in average……….….………...………....121 Fig. 2.2 TEM micrographs showing the shapes of the nano particles in the resulting PEEK nanocomposites: (a) SiO₂ (30 nm)and (b) Al₂O₃ (30 nm)…………..………….122 Fig. 2.3 Schematic drawing of the Mg based laminated composite, layers 1, 3 and 5 are Mg and layers 2 and 4 are APC-2 (with 4 foils). The longitudinal direction is indicated….…………..………123 Fig. 2.4 The geometry and dimensions of tensile test spcimen of Mg/CF/PEEK laminated composite, laps shown in figure are the copper laps adhered…..……… ……...124 Fig. 2.5 The geometry and dimensions of flexural test…..………...125 Fig. 2.6 The geometry of specimen for T-Peel test……..……….126 Fig. 2.7 Vacuum hot-press for the fabrications of SiO₂ and Al₂O₃ particulates filled PEEK nanocomposites……..………..…127 Fig. 2.8 Fabrication of SiO₂ or Al₂O₃ particulates filled PEEK nanocomposite, (a) molding,

(b) fabricated PEEK nanocomposite………..………..128 Fig. 2.9 The geometry and dimensions of tensile test specimen of particulates filled PEEK nanocomposite, laps shown in figure are the copper laps adhered………...….129 Fig. 3.1 Mg/APC-2 laminate with 4 plies of APC-2 prepreg sandwiched by two AZ31 Mg sheets formed at 400ôC under 1.1 MPa (sample 1 in Table 3.1)……….……...130 Fig. 3.2 Mg/APC-2 laminate with 4 plies of APC-2 prepreg sandwiched by two AZ31 Mg sheets formed at 400ôC under 0.7 MPa (sample 2 in Table 3.1)…….……….…131 Fig. 3.3 Mg/APC-2 laminate with 4 plies of APC-2 prepreg sandwiched by two AZ31 Mg sheets formed at 400ôC under 0.7 MPa (sample 3 in Table 3.1)………....…..…132

(12)

Fig. 3.4 Mg/APC-2 laminate with 4 plies of APC-2 prepreg sandwiched by two AZ31 Mg sheets formed at 400 ôC under 1.6 MPa (sample 4 in Table 3.1)…….………....133 Fig. 3.5 Mg/APC-2 laminate with 3 plies of APC-2 prepreg sandwiched by two AZ31 Mg sheets formed at 400ôC under 1.0 MPa (sample 5 in Table 3.1)………..………134 Fig. 3.6 Mg/APC-2 laminate with 2 plies of APC-2 prepreg each layer and laminated in the stacking sequence of Mg/ APC-2/Mg/APC-2/Mg formed at 400ôC under 0.7 MPa (sample 6 in Table 3.1)………...………...…135 Fig. 3.7 Mg/APC-2 laminate with 2 plies of APC-2 prepreg each layer and laminated in the stacking sequence of Mg/ APC-2/Mg/APC-2/Mg formed at 400ôC under 1.4 MPa (sample 7 in Table 3.1)………..…………..…………...………….…136 Fig. 3.8 (a) Room temperature tensile stress strain curve of the Mg/CF/PEEK Mg based laminated composite along the longitudinal direction, and (b) extraction of the Young’s modulus of the Mg/CF/PEEK Mg based laminated composite…...137 Fig. 3.9 (a) Room temperature tensile stress strain curve of the Mg/CF/PEEK Mg based laminated composite along the transverse direction, and (b) extraction of the Young’s modulus of the Mg/CF/PEEK Mg based laminated composite…...138 Fig. 3.10 Tensile stress strain curve of the Mg/CF/PEEK Mg based laminated composite at 100ôC along the (a) longitudinal and (b) transverse directions………139 Fig. 3.11 Tensile stress strain curve of the Mg/CF/PEEK Mg based laminated composite at 150ôC along the (a) longitudinal and (b) transverse directions………....140 Fig. 3.12 SEM micrographs of the room-temperature fractured specimens of Mg/CF/PEEK laminated composite, taken from the longitudinal specimens……….141 Fig. 3.13 SEM micrographs of the room-temperature fractured specimens of Mg/CF/PEEK laminated composite, taken from the transverse specimens………....142 Fig. 3.14 SEM micrographs of the fractured specimens of Mg/CF/PEEK laminated

composite loaded at 100^oC, taken from the longitudinal specimens, showing (a)

(13)

the carbon fiber broken in the APC-2 prepreg, and (b) the correlation of the fracture positions in the Mg phase and the carbon fibers………143 Fig. 3.15 SEM micrographs of the fractured specimens of Mg/CF/PEEK laminated

composite loaded at 150^oC , taken from the longitudinal specimens, showing (a) the carbon fiber broken in the APC-2 prepreg, and (b) the broken positions of the carbon occurring at different places, indicating the ductile fracture behavior in the Mg matrix……….………144 Fig. 3.16 SEM micrographs of the fractured specimens of Mg/CF/PEEK laminated

composite loaded at 100^oC, taken from the transverse specimens, showing (a) the interface de-attachment fracture and dimples in the Mg phase and, (b) the

interface delamination and the microcrack in the APC-2 phase………..145 Fig. 3.17 SEM micrographs of the fractured specimens of Mg/CF/PEEK laminated

composite loaded at 150^oC, taken from the transverse specimens, showing (a) the interface fracture behavior and, (b) the de-attachment behavior between the PEEK resin and the carbon fiber……….146 Fig. 3.18 (a) Room temperature flexural stress strain curve of the Mg/CF/PEEK Mg based laminated composite along the longitudinal direction, and (b) extraction of the flexural modulus of the Mg/CF/PEEK Mg based laminated composite………..147 Fig. 3.19 (a) Room temperature flexural stress strain curve of the Mg/CF/PEEK Mg based laminated composite along the transverse direction, and (b) extraction of the flexural modulus of the Mg/CF/PEEK Mg based laminated composite………..148 Fig. 3.20 Photographs of the fractured Mg/CF/PEEK laminated composites for the (a)

longitudinal and (b) transverse configurations of carbon fibers. The loading direction is indicated………...……….149 Fig. 3.21 Typical peeling test results for the Mg/CF/PEEK laminated composites along the (a) longitudinal and (b) transverse directions………..……150

(14)

Fig. 3.22 OM micrographs taken from the peel-tested specimens with the longitudinal configurations of carbon fibers: (a) Mg layer without CrO3 etching, (b) Mg layer with CrO₃ etching, and (c) APC-2 layer. The lighter-contrasted PEEK resin adhered on the Mg phase is evident in (b) and on carbon fibers in (c), and the broken carbon fibers stuck on the Mg phase in (b)……….….151 Fig. 3.23 Chemical structure of the PEEK polymer………..………..152 Fig. 3.24 FT-IR spectra on the (a) AS-4 prepreg, (b) etched CF-phase peeled from the

Mg/APC-2 laminated composite, (c) unetched CF-phase peeled from the Mg/APC-2 laminated composite, and (d) etched Mg-phase peeled from the

Mg/APC-2 laminated composite………..153 Fig. 3.25 Variations of the microhardness of the nanocomposites as a function of the

nanoparticle content in wt%……….………154 Fig. 3.26 Variations of the (a) Young’s modulus E, (b) ultimate tensile stress UTS, and (c) tensile failure elongation e of the nanocomposites as a function of the particle content in wt%……….155 Fig. 3.27 SEM/EDS elemental mapping (Si or Al) for the composites with: (a) 5 wt% SiO2, (b) 5 wt% Al₂O₃, (c) 7.5 wt% SiO₂, and (d) 7.5 wt% Al₂O₃………...………....156 Fig. 3.28 TEM micrographs showing the distribution of the nano particles: (a) 2.5 wt%

SiO₂ (15 nm)and (b) 5 wt% SiO₂ (15 nm)……….……….157 Fig. 3.29 TEM micrographs showing the distribution of the nano particles: (a) 2.5 wt%

SiO₂ (30 nm)and (b) 5 wt% SiO₂ (30 nm)……….……….158 Fig. 3.30 TEM micrographs showing the distribution of the nano particles: (a) 2.5 wt%

Al₂O₃ (30 nm)and (b) 5 wt% Al₂O₃ (30 nm)………..….159 Fig. 3.31 TEM micrographs showing the distribution of the nano particles: (a) 2.5 wt%

SiO₂ (15 nm)and (b) 2.5 wt% Al₂O₃ (30 nm)………..………160 Fig. 3.32 X-ray diffraction patterns of the PEEK nanocomposites filled with 30 nm (a) SiO2

(15)

and (b) Al₂O₃ particles……….161 Fig. 3.33 DSC thermalgrams of the pristine PEEK during nonisothermal crystallization at

different cooling rates………..162 Fig. 3.34 DSC thermalgrams of the nanocomposites during nonisothermal crystallization at

different cooling rates: (a) 2.5 wt% 15 nm silica/PEEK (b) 2.5 wt% 30 nm

silica/PEEK, and (c) 2.5 wt% 30 nm alumina/PEEK………..…163 Fig. 3.35 DSC thermalgrams of the nanocomposites during nonisothermal crystallization at

silica/PEEK, and (c) 5.0 wt% 30 nm alumina/PEEK………..………164 Fig. 3.36 DSC thermalgrams of the nanocomposites during nonisothermal crystallization at

silica/PEEK, and (c) 7.5 wt% 30 nm alumina/PEEK………..165 Fig. 3.37 DSC thermalgrams of the nanocomposites during nonisothermal crystallization at

different cooling rates: (a) 10.0 wt% 15 nm silica/PEEK (b) 10.0 wt% 30 nm silica/PEEK, and (c) 10.0 wt% 30 nm alumina/PEEK………..……..166 Fig. 3.38 DSC thermalgrams of pristine PEEK upon heating showing the melting peak. All

the heating rates are 10 ôC/min. Before heating up to 410 ôC, the specimen was cooled from 410 to 50 ôC at different cooling rates shown in the figures……...167 Fig. 3.39 DSC thermalgrams of the nanocomposites upon heating showing the melting peak:

(a) 2.5 wt% 15 nm silica/PEEK, (b) 2.5 wt% 30 nm silica/PEEK, and (c) 2.5 wt%

30 nm alumina/PEEK. All the heating rates are 10 ôC/min. Before heating up to 410 ôC, all the specimens were cooled from 410 to 50 ôC at different cooling rates shown in the figures………...…………..168 Fig. 3.40 DSC thermalgrams of the nanocomposites upon heating showing the melting peak:

30 nm alumina/PEEK. All the heating rates are 10 ^oC/min. Before heating up to

(16)

410 ^oC, all the specimens were cooled from 410 to 50 ^oC at different cooling rates shown in the figures……….………169 Fig. 3.41 DSC thermalgrams of the nanocomposites upon heating showing the melting peak:

30 nm alumina/PEEK. All the heating rates are 10 ôC/min. Before heating up to 410 ôC, all the specimens were cooled from 410 to 50 ôC at different cooling rates shown in the figures………...………..170 Fig. 3.42 DSC thermalgrams of the nanocomposites upon heating showing the melting peak:

(a) 10.0 wt% 15 nm silica/PEEK, (b) 10.0 wt% 30 nm silica/PEEK, and (c) 10.0 wt% 30 nm alumina/PEEK. All the heating rates are 10 ôC/min. Before heating up to 410 ôC, all the specimens were cooled from 410 to 50 ôC at different cooling rates shown in the figures………...……….171 Fig. 3.43 The typical effect of filler content on peak crystallization temperature, T_cp, of

PEEK nanocomposites at a cooling rate of 5 ^oC/min: (a) filler content in terms of weight percent, wt%, and (b) filler content in terms of volume percent, vol%...172 Fig. 3.44 The typical effect of filler content on melting temperature, Tm, of PEEK

nanocomposites at a cooling rate of 5 ^oC/min: (a) filler content in terms of weight percent, wt%, and (b) filler content in terms of volume percent, vol%………...173 Fig. 3.45 Overall crystallization time versus filler content at various cooling rates: (a) 15

nm silica/PEEK, (b) 30 nm silica/PEEK, and (c) 30 nm alumina/PEEK……....174 Fig. 3.46 The effect of filler content and dimension on the overall crystallization of the

PEEK chain segments at a cooling rate of 5^oC/min: (a) filler content in terms of weight percent, wt%, and (b) filler content in terms of volume percent, vol%...175 Fig. 3.47 Absolute crystallinity versus cooling rate: (a) 15 nm silica/PEEK, (b) 30 nm

silica/PEEK, and (c) 30 nm alumina/PEEK……….176 Fig. 3.48 The effects of filler content and dimension on the crystallinity of the PEEK chain

(17)

segments at a cooling rate of 5^oC/min: (a) filler content in terms of weight percent, wt%, and (b) filler content in terms of volume percent, vol%……….……177 Fig. 3.49 The TGA diagrams of the PEEK nanocomposites filled with 30 nm (a) SiO₂ and

(b) Al2O3 particles………...……….178

(18)

ABSTRACT

Magnesium alloys have attracted considerable attention owing to its low density of ~1.7 g/cm³. On the other hand, the carbon fiber (CF) reinforced polyether ether ketone (PEEK) polymer composites possess extraordinary specific strength and stiffness along the longitudinal (or fiber) direction. It follows that the combination of Mg/CF/PEEK would offer an alternative in forming a high specific strength and stiffness composite. In the first part of this study, the low density and high performance Mg-based laminated composites were fabricated by means of sandwiching the AZ31 Mg foils with the carbon-fiber/PEEK prepreg through hot pressing. Proper surface treatments of AZ31 sheet using CrO3 base etchants are necessary in order to achieve good interface bonding characteristics. The resulting Mg base laminated composite, with a low density of 1.7 g/cm³, exhibits high modulus of 75 GPa and tensile strength of 932 MPa along the longitudinal direction. The experimentally measured tensile modulus and strength data along both the longitudinal and transverse direction are within 90-100% of the theoretical predictions by rule of mixtures, suggesting that the bonding between layers and the load transfer efficiency are satisfactory. The flexural stress and modulus along the longitudinal direction are 960 MPa and 54.6 GPa, respectively, suggesting a sufficiently high resistance against bending deflection. The peel strengths are about 2.75 and 4.85 N/mm along the longitudinal and transverse directions, respectively, superior to that of the epoxy-resin-adhered and carbon-fiber-reinforced aluminum laminated composites.

Polymer nanocomposites have attracted considerable attention during the past decade due to their versatile and extra-ordinary performances. The polymer nanocomposites can be prepared by the well-known sol-gel method. It is well known that PEEK is a good solvent resistant polymer. Hence, it is impossible to fabricate the PEEK nanocomposite by means of sol-gel method. In the second part of this study, the PEEK nanocomposites filled with nano-sized silica or alumina measuring 15-30 nm to 2.5-10 weight percent were fabricated by

(19)

vacuum hot press molding at 400^oC. The resulting nanocomposites with 5-7.5 wt% SiO₂ or Al2O3 nanoparticles exhibit the optimum improvement of hardness, elastic modulus, and tensile strength by 20-50%, with the sacrifice of tensile ductility. With no surface modification for the inorganic nanoparticles, the spatial distribution of the nanopartilces appears to be reasonably uniform. There seems no apparent chemical reaction or new phase formation between the nanoparticle and matrix interface. The crystallinity degree and thermal stability of the PEEK resin with the addition of nanopartilces were examined by X-ray diffraction, differential scanning calorimetry, and thermogravity analyzer, and it is found that a slight decrease in crystallinity fraction and a higher degradation temperature would result in as compared with the prestine PEEK.

(20)

中文提要

鎂合金由於俱低密度 (1.7-1.8 g/cm³) 的特性，故可做為輕量化金屬結構材料方面之

應用，因此在這幾年廣獲青睞。眾所周知，碳纖維 (CF) 強化聚二醚酮 (PEEK) 高分子複合材料 (CF/PEEK) 在其縱向俱有超高之比強度與比剛性；故Mg/CF/PEEK複合材料將是製備一高比強度及比剛性複合材料之另一方式。本研究第一部分將以三明治堆疊方式利用AZ31 鎂薄板與CF/PEEK預浸布在真空熱壓機中壓製低密度及高性能鎂基夾層複

合材料。為獲致良好的界面接著性能，在熱壓前鎂板需利用CrO₃行表面處理。真空熱壓

製得之Mg/CF/PEEK鎂基夾層複合材料俱低密度 (1.7 g/cm³) 之特性，且在縱向之彈性

模數及最大抗拉強度分別高達 75 GPa及 932 MPa。而不論是縱向及橫向之彈性模數及最大抗拉強度更高達 90 至 100%的理論值，顯示其界面接著及負荷傳遞是非常有效且充分。在Mg/CF/PEEK鎂基夾層複合材料之撓曲及剝離性質方面，在縱向撓曲模數及應力也分別高達 54.6 GPa及 960 MPa，顯示此鎂基夾層複合材料俱有很高的抗彎曲特性。再者，在縱向及橫向之剝離強度也分別達 2.75 及 4.85 N/mm，優於環氧乙烷接著之鋁基碳纖維強化夾層複合材料。

高分子奈米複合材料由於俱多樣化及超高性能之特性，在過去這十年來也吸引眾多注目的眼光；眾所周知，高分子奈米複合材料可利用溶膠-凝膠法製得。PEEK因俱耐溶劑特性，因此，無法利用溶膠-凝膠法製備PEEK奈米複合材料。本研究第二部分將利用

熱壓成型法在 400^oC真空熱壓機中製備PEEK奈米複合材料，並利用 15 及 30 奈米大小

的氧化矽及氧化鋁作為強化相，此強化相之重量分率在 2.5 至 10%之間。經實驗證實，

氧化矽及氧化鋁含量在 5 至 7.5 之重量百分率時，PEEK奈米複合材料之硬度、彈性模數，及最大抗拉強度可提高百分之 20 至 50，但其斷裂伸度則下降。在無任何的奈米粉體表面改質下，氧化矽及氧化鋁奈米粉體在PEEK基材中之分散還算均勻；且經X-ray繞射證實，氧化矽及氧化鋁奈米粉體與PEEK高分子間並無明顯的化學反應產生。氧化矽及氧化鋁奈米粉體的添加對PEEK高分子結晶性及熱穩定性的影響則利用示差掃瞄卡計

(21)

(DSC) 及熱重分析儀 (TGA) 檢測之，實驗證實經氧化矽及氧化鋁奈米粉體強化之

PEEK奈米複合材料的結晶度會稍微下降，而熱裂解溫度則會較諸純PEEK提高約 40^oC。

(22)

致謝

人的一生總是會經歷許許多多的抉擇，這些抉擇到最後就會印證當初決定的正確與否。當初，承蒙恩師黃志青教授的恩慈與啟示，使我由高分子組，毅然地跨到金屬組，

當時也不知這決定是否正確；不過，經歷了這三年半在恩師的教導、關心與鼓勵當中，

讓我能夠一直充滿信心，並順利取得博士學位。有了當時恩師的開導與啟示，使我在這博士學涯中才得以克服種種困難，我要再次說聲：老師，謝謝您，您辛苦了。

我也要由衷地感謝本所陳明與高伯威教授，由於您們的指導、鼓勵與幫助，使我實驗的進行才得以更加順利；還有機電所的任明華教授，由於您的協助，使我的實驗進行順利。

最令我難忘的是黃幫所有成員，在這個大家族中，大家互相幫助，相互討論，彼此鼓勵，這種體驗是很難得的。凱琳學姐在 TEM 上的幫忙，建超的鼓勵，佩如學姐、鉉凱、英博、敬仁、小明在實驗上的協助與金屬材料知識的開示，子翔、志溢、宇庭、政信、家豪、海明、世儒在實驗上的協助，炎暉在口試時的幫忙，謝謝您們。另外，我也要感謝庾忠義學長與吳玉娟同學在 Instron 及 SEM 上的教導與協助，機電所曾育鍾同學在實驗上的熱心協助。

最後，我要將這一份喜悅獻給我年邁的母親，由於您時時的鼓勵與支持，使我才有足夠的毅力與信心，完成博士學位。我也要向我三位可愛又乖巧的子女，鈺菁、筱娟、

哲昇說聲對不起，爸爸這幾年沒能全心全力地照顧你們。在天國的父親，願您也能感受到兒子的這一份小小的成就。

謝謝所有關心我的大家。

(23)

Chapter 1 Background and Research Motive

1.1 Light-weight magnesium based alloys

1.1.1 Characteristics of magnesium alloys

Magnesium alloys have attracted considerable attention and interest worldwide during the past five years, due to the improvement of casting and processing techniques. Magnesium is the 8th most abundant element; the earth crust and ocean consist of 1.93 and 0.13 mass percents of magnesium, respectively [1]. In 1808, almost two hundreds year ago, magnesium was first extracted into a pure form by Davy [2]. Through extensive basic studies over the years, the chemical and physical properties of magnesium are well established. It is noted that Mg is the lightest structural metals on earth; Li and Be are indeed even lighter but the former cannot be present in individual metal form and the latter is extremely toxic. Because of its low density of ~1.7 (similar to or only slightly above the densities of most polymers and polymer composites), as shown in Table 1-1 [3], this metal raises the possibility of weight saving in metallic structures, and particularly in aircraft, vehicles and transportation equipment. Moreover, magnesium alloys have been, or have potential to be, applied by their characteristic natures of high specific strength and stiffness, superior damping capacity, high thermal conductivity, high dimensional stability, and good machinability [4].

Magnesium can be alloyed with various solute elements, including aluminum, zinc, lithium, thorium, silver and several rare earth elements such as cerium, neodymium and yttrium [2,5,6]. The addition of aluminum can largely increase the alloy strength through solution and precipitation strengthening, while a small amount of Zn will improve the cast capability. The designation of magnesium alloys is based on the abbreviation of the including

(24)

solute elements and their contents in weight percent. For example, the AZ31 alloy is referred to the magnesium base alloy added with nominally 3 wt% of aluminum (A) and 1 wt% of zinc (Z).

Since magnesium has fairly low plastic formability and limited ductility because of its HCP (hexagonal close-packed) crystalline structure, the fabrication of magnesium products are usually proceeded by die-casting or thixomolding [7]. This is distinctly different from the case of aluminum alloys, for which the wrought-typed aluminum alloys are more frequently applied; and the wrought alloys generally exhibit higher fracture toughness than the cast alloys. For commercial wrought magnesium base alloys currently available, it is still difficult to manufacture structural components. Recently, the plastic forming of magnesium alloys can be greatly improved by means of (1) the reduction of impurities during extraction metallurgy and casting routine, and (2) the structure control through secondary thermomechanical processing treatments in order to refine mainly the grain size. As a result, numerous Mg alloys were processed to exhibit superplasticity at elevated temperatures of ~0.5-0.8 Tm, where Tm is the material melting point expressed in Kelvin. Thus, new processing means, such as superplastic forming, press forming, and injection molding, are gradually becoming more important techniques to fabricate a hard-to-form material into complex shapes [8-10].

In 1999, press forming of the AZ31 magnesium sheets was conducted under a more economical condition, i.e., proceeding at a speed and temperature faster and lower than the superplastic forming practice [11]. It is known that a small grain size can improve the superplasticity performance of alloy, the smaller the grain size will lead to better ductility and higher optimum strain rate for superplasticity. Therefore, hot extrusion and powder metallurgy methods have been utilized to produce superplastic microstructures [12,13]. The AZ91 and AZ31 alloys with grain sizes of ~5 µm, hot extruded at a reduction ratio of 100:1,

(25)

exhibited tensile elongations of 350% at 3×10^-4 s^-1 and 200ôC and 620% at 10^-4 s^-1 and 325ôC, respectively [12,13]. And the AZ91 and ZK61 alloys processed by powder metallurgy showed tensile elongation over 300% at high strain rates of 10^-2 s^-1 to 10^-1 s^-1 [13]. Equal channel angular extrusion (ECAE), recently, has been developed to produce ultra-fine grains of 0.7 µm in the AZ91 alloy [14], resulting in a maximum elongation of 660% obtained at 6×10^-5 s^-1 and a relatively low temperature of 200ôC.

With the improvement of Mg alloy processing, the properties of Mg alloys have gradually reached the requirements for high functionalities of mass products such as automobiles and electronic devices. Undoubtedly, Mg alloys are the extra light metals in the 21st Century.

In the past decade, electronics industry has made giant growth, especially in computer and communication areas. Due to the considerations of weight-saving, damping, electric and magnetic shielding, better heat dissipation, environmental stability, and recycling ability, the use of Mg alloys over polymers or polymer composites is under steady growth. It is apparent that, as a structural material, magnesium has numerous advantages over aluminum and engineering plastics. Therefore, Mg alloys have become more and more attractive for the design of new lines of video or photo graphic equipment, portable personal computers and notebooks, cellular and satellite cell phones, personal LCD projectors, and portable communication equipment. Another new field of application for Mg alloys is the medical uses, such as an implant material for surgery [15]. Magnesium offers a low dosage that is an essential element and does not harm the tissue; moreover, magnesium promotes the healing of the bone. Furthermore, the elastic modulus of magnesium is closely to that of the corticalis and the ultimate tensile strength (UTS) is higher than those of polymers. In addition, polymer based biodegradable implants would provoke a rejection by the body but magnesium would

(26)

not.

Magnesium matrix composites reinforced mostly by ceramic particulates may become structural materials for vehicles and aerospace applications because of their high specific mechanical properties. It was shown that magnesium matrix composites revealed increased hardness [16-18] and elastic modulus [16-19], and even low temperature superplasticity [20].

They can be fabricated by casting or powder metallurgy techniques and by deposition of matrix from semi-solid or vapor phase.

Nano-particle reinforced magnesium composites have been shown to enhance the mechanical properties [21,22]. Hwang and Nishimura [23] synthesized the Mg-TiC nanocomposite by mechanical milling. It was shown that the as-milled Mg-TiC nanocomposite contained magnesium matrix grain size ranging from 25 to 60 nm with a dispersion of ultra-fine nano-sized ceramic TiC particles (3-7 nm). It was also shown that Mg-TiC nanocomposite exhibited remarkably high ductility [23].

1.1.2 The properties of AZ31 magnesium alloy

It is well-known that Mg alloys show poor plastic formability due to the HCP structure.

Except for this structural limitation, however, it is often difficult to fabricate large Mg products with high strength and high ductility by the casting process because of coarse grain size. In view of plastic formability and post-deformation mechanical properties, it has been reported that hot deformation process such as extrusion and ECAE can account for grain refinement [24-26] owing to the dynamic recrystallization in Mg alloys [27-29]. Higashi et.

al. [24] conducted their study on the effect of high-strain-rate forming process on the grain refinement of commercial AZ31 (Mg-3wt%Al-1wt%Zn-0.2 wt%Mn) alloy. There were

(27)

various processes being conducted including (a) extruding the cast alloy with a high ratio of 100:1, (b) severe plastic deformation through ECAE, (c) powder metallurgy (P/M) procession of machined chip [24].

For process (a), the grain size of AZ31 was refined from initially ~15 µm to an equiaxal grain structure with the average size of ~5 µm at an extrusion temperature of 350ôC. On the other hand, the grain size developed by process (b) at 160 to 220ôC was varied from 0.5 to 3 µm according to the extrusion temperature. In addition, process (c) developed grain size varying from 2 to 4 µm at a temperature ranging from 210 to 430ôC. In view of superplasticity, the grain size must be refined to 0.5~2 µm for a target strain-rate range of 10^-2~10⁰s^-1 for mass production. Accordingly, the sub-micron grained structures can be achieved by process (b) ECAE of the cast AZ31 alloy.

Mukai et al. [26] also conducted their study on the ductility enhancement in AZ31 Mg alloy by controlling its grain structure. Two different processed AZ31 alloys were inspected in this study, one was the as-ECAE processed alloy, initially of grain size ~1 µm, which was followed by annealing (AZ31-ECAE/annealed) process to coarsen the grains to a grain size

~15 µm, the other was the conventionally extruded AZ31 alloy having grain size ~15 µm in average. In terms of tensile mechanical properties, the yield stress of AZ31-ECAE/annealed exhibited a half value compared with that of the as-extruded alloy owing to the difference of texture in the two alloys. However, the ultimate tensile strength of AZ31-ECAE/annealed revealed almost the same value as the as-extruded alloy; exhibiting a remarkable strain hardening and a large uniform elongation as compared with the as-extruded alloy. It is well known that cast magnesium alloys exhibit higher specific strength than those of steels due to their lower densities [30,31]. On the other hand, the values of elongation-to-failure for the Mg alloys also exhibit remarkable low values compared with the structural steels. According

(28)

to the study conducted by Mukai et al. [26], AZ31-ECAE/annealed exhibited a similar value of specific strength compared with the cast-magnesium alloys and a large value of elongation-to-failure (up to 50 %), similar to steels. Accordingly, the AZ31-ECAE/annealed alloy has a high potential in structural uses.

As shown in Table 1.2 [32], for extruded alloy, AZ80 exhibits comparable tensile strength as Al alloy 6061, but with less ductility. However, Mg sheet metal, such as AZ31 alloy, provides slightly lower strength but a higher ductility than commonly used 5XXX series Al sheet alloys. Also shown in Table 1.2 are the physical properties of PC/ABS plastics. Mg alloys are slightly heavier than the plastics, but they are much stiffer due to elastic modulus of magnesium is almost 20 times over a plastic material such as PC/ABS.

1.2 Thermoplastic high temperature polymer PEEK

1.2.1 The properties of PEEK

The high performance polymer poly(ether-ether-ketone) (PEEK) was firstly prepared by Bonner in 1962 [33]. It is a derivative of poly(aryl-ether-ketones). The PEEK polymer was reported to be synthesized by a nucleophilic aromatic substitution reaction, using dephenyl sulfone as solvent, at temperatures approaching the melting point of the polymer [34].

PEEK is chemically recognized as a linear poly(aryletherketone) and is a melt processable aromatic polymer; the melting point Tm is between 330 and 385 ^oC, depending on the relative proportion of ether-ketone groups linking the phenylene rings [34]. It is highly crystalline. Dawson and Blundell [35] reported the values for the lattice constants of the annealed isotropic samples and other parameters of chain conformations, as shown in Table

(29)

1-3.

The bulk properties of PEEK, compared with those for ‘Victrex’ polyether sulphone, are shown in Table 1-4 [34]. The crystallinity of PEEK and its lower glass transition temperature Tg highlight the major differences between these high temperature performance thermoplastics. As shown in Table 1-4, PEEK has a lower heat distortion temperature, just above its Tg. However, it maintains useful long-term mechanical strength up to 200ôC, due to its high Tm. According to the DMA (Dynamic Mechanical Analyzer) measurement conducted on PEEK and polyethersulphone [34], the polytehersulphone loses all mechanical strength at 200ôC. However, PEEK remains some rigidity until significant melting of the crystallites occurs near 300ôC.

Also shown in Table 1-4 is the resistance to solvent stress cracking for PEEK as compared with the polyethersulphone. And it is also derived from the crystallinity of PEEK.

PEEK has good resistance to many organic solvents. Nevertheless, PEEK can be dissolved in concentrated H2SO4 and CH3SO3H, as shown in Table 1-5 [36]. It is believed that the protonation of PEEK, when dissolved, gives rise to repulsive electronstatic forces which can overcome the strong attractive forces in this highly crystalline polymer. Accordingly, the dissolution of PEEK in various sulfonic acids, followed by recovery of the polymer, provides a route to a new type of ionomer [37].

Morphologically, neat PEEK resin, similar to other semicrystalline polymers, possesses a spherulite structure as cooled from its melt. The degree of crystallinity of the polymer highly depends on its thermal histories and on the processing conditions, such as the cooling rate and annealing treatments. And in turn the degree of crystallinity of the processed PEEK imparts a very significant effect on the material properties and mechanical behaviors of the

(30)

resulting composites. Gao and Kim [38] found that the interface bond strength decreased with increasing cooling rate; the tensile strength and elastic modulus of PEEK resin decreased, while the ductility increased, with increasing cooling rate through its dominate effect on crystallinity and spherullite size. Accordingly, a slower cooling rate will result in polymers more brittle in nature than those fabricated by faster cooling from its melt state.

Thermoplastic polymers have been considered as substitutes for thermosetting matrices for high-performance composite materials. They offer advantages such as higher processability, easier repair and bonding operations, and reprocessabilies. Aromatic Polymer Composites (APC) based on continuous carbon fibers embedded in PEEK matrix represent one of the most developed high-performance thermoplastic composites. As well-known, some of the principal limitations of thermosets are their relative brittleness and water sensitivity. However, PEEK and APC have been turned out to be a good impact behavior and very low water absorption compared with high-performance epoxy systems [39-42].

1.2.2 Applications of PEEK

PEEK is a semicrystalline polymer and capable of providing many of the unique properties in terms of temperature and solvent resistances. High-performance microfiltration membranes from PEEK were prepared [43] for the application in large internal diameter hollow fiber (tubule) form, resulting in the highest cross-flow efficiency. The resulted membranes are a polymer blends and their pore sizes vary from approximately 0.12 µm at 16% PS (polysulphone) to approximately 0.3 µm at 25% PS for a given PEEK content, in which PS provides sufficient melt strength to the extruding blend. It was also shown that PEEK membranes provide superior performance to a PS membrane upon exposure to a warm surfactant/oil stream. However, the porosity of PEEK/PS membranes is about 5%, due to

(31)

membrane shrinkage during the leach step.

Furthermore, extrusion of film-microfiltration membranes of PEEK was also conducted [44], and the maximum pore size of the resulting membranes was less than 0.05 µm. In view of porosity of the membranes, however, it was reported that the ultrafiltration film membranes from PEEK, a 14% PEEK and 18% PS blend, yields a porosity of 79%.

PEEK has been shown to be environmental resistance and passive biocompatibility, i.e., absence of toxicity and biological inertness [45,46]. The native PEEK film was found to be a very poor substrate for cell cultivation, extremely reluctant to allow cellular adhesion [47].

Increasing of the surface hydrophilicity by introduction of polar groups has been investigated to improve the bioadhesion [48,49]. Surface carboxylated PEEK films were prepared from PEEK-OH films which are common key-intermediate, and these films revealed surface functionalities such as grafting of bioactive molecules like proteins and peptides [50].

Surface amination of the PEEK film [51], on the other hand, also proceeded by the grafting of glutamine, and this film displayed α-amino acid motifs fixed on the polymer backbone via a short spacer-arm. The surface fluorination of PEEK film has been successfully prepared [52]. It was realized to be a blood compatible material. Sulphonated PEEK films [53] have been shown to be capable of ion-exchange, and, as a result, exhibited a high permeability for copper ions due to the presence of fixed negative charges and to their swelling capacity in an aqueous phase.

PEEK polymers reinforced with nanoparticles have been reported [54-56]. The PEEK fine powders, ~100 µm, were fully mixed with Si3N4 nanoparticles and subsequently formed by compression molding [54]. As a result, the incorporation of Si3N4 nanoparticles into PEEK caused a significant improvement in the tribological characteristics, resulting

(32)

considerably in decreased frictional coefficient and wear rate. Furthermore, it was proposed that a thin and uniform transferred film could be formed on this composite during the friction process. Due to the increase of adhesion strength of the transferred film through chemical reaction between Si3N4 nanoparticle and the steel substrate [54], consequently, sliding occurred between the composite and the transferred film, resulting in a lower wear rate. In addition, PEEK polymers reinforced with other nanoparticles such as ZnO2 [55] and SiC [56]

were also examined, and similar results were reported.

Through the treatment with a dilute chlorosulfonic acid solution, the inner walls of the PEEK capillaries reveal an increase in surface area, which is suitable for the application of electrosmotically driven open-tubular liquid chromatography (LC) [57].

Among the so many applications of PEEK, the continuous fibers reinforced composites have become the most high-performance and advanced materials over the past decades.

Undoubtedly, owing to the potential advantages of high fracture toughness, high temperature resistance, repairability, biocompatibility, and ease of manufacture, fiber reinforced composites of PEEK can extend their influences to many areas including aerospace materials, structural materials, and biomedical materials. Carbon fiber reinforced PEEK composite (CF/PEEK) is a shinny star among the PEEK composites, and is being considered as the candidates to replace the conventional epoxy-based composites for aerospace applications.

And we will further discuss in more details in Section 1.3.2.

1.3 Introduction to polymer matrix composites (PMC)

1.3.1 Polymer matrix composites

(33)

It is well known that, in the past several decades, polymer matrix composites (PMC) have become advanced materials, and can be applied as engineering structural materials for aircrafts or vehicles, as well as biomedical materials for medical uses [58]. Polymer matrix composites are conventionally classified into two groups: thermoset matrix composites (TSC) and thermoplastic matrix composites (TPC). As shown in Table 1-6, thermoset composites have slightly different properties from the thermoplastic counterparts; the former ones usually exhibit much lower strains to failure.

Composite technology is based on taking advantages of the stiffness and strength of high-performance fibers by dispersing them in a matrix, which acts as a binder and transfers the acting load to the fibers across the fiber-matrix interface. To understand how the properties of a composite originate, it is necessary to know the properties of constituents form a composite system. The mechanical properties of a composite are determined by a number of factors, including the moduli and strengths of the fiber and matrix; aspect ratio, length distribution, volume fraction, uniformity and orientation of the fibers, as well as the integrity of the fiber-matrix interface and the interfacial bond strength [59].

The first generation of composite materials based on the more brittle thermoset matrix offers fracture toughness as low as 100 J/m². The development of toughened thermosets and a wide range of high performance thermoplastics have increased this value up to 2000 J/m² [60]. Advanced thermoset epoxy composites are now the most often used in high performance applications due to their unique performance-to-cost ratio. They generally possess excellent properties and are suitable for a large number of processing techniques.

However, thermoset epoxy composites have been found that the properties of toughness and dimensional stability will decrease as the glass transition temperature Tg of the resin used increases.

(34)

In addition, a change in temperature and moisture content could result in moisture-induced stress as well as dimensional change in composite body [61-64].

Furthermore, the recursive changes of internal stresses due to water absorption-desorption processes may induce fatigue damage, and in turn influence long-term durability and performance of composite [65].

Thermoplastic matrix composites present a number of advantages over thermoset composites, including increased fracture toughness, lower moisture absorption, potential for reduced life-cycle cost, good welding property, and recyclability [66,67].

1.3.2 High performance carbon-fiber/PEEK (CF/PEEK) composite

Due to the high fracture toughness, high temperature resistance, repairability and ease of manufacture, thermoplastic matrix composites have been studied extensively [68-73]. Among these, the carbon-fiber/PEEK(CF/PEEK) composite is one of candidates to replace conventional epoxy-based composites for aerospace applications. Because of the short processing time needed, the CF/PEEK composite provides flexibility in adapting various manufacturing technologies to improve the production efficiency. However, the recommended processing condition for the CF/PEEK composite requires a forming temperature of 400^oC and a pressure of 1.4 MPa for 15 min [38], which are much higher than those for the epoxy-based composites. The higher requirement of processing conditions might therefore limit the potential to make use of cost-effective manufacturing technologies for fabricating components from the CF/PPEK composite.

During the past decades, many researches have been conducted to study the processing conditions in order to search for the opportunities of broadening the processing window for

(35)

the CF/PPEK composites. An inevitable variation of the processing condition is the cooling rate. Gao and Kim [38] found that the cooling rate controlled the degree of crystallinity which in turn was correlated to the interface adhesion, the crystalline morphology, and the bulk mechanical properties of neat PEEK resin. As a result, the interface bond strength, as well as the tensile strength and elastic modulus, decreased with increasing cooling rate.

However, the ductility increased with increasing cooling rate due to its effect on crystallinity and spherullite size. In addition, the interface failure was recognized as brittle debonding in slow-cooled composites. In contrast, the amorphous PEEK-rich interface introduced in fast cooled specimens failed in a ductile manner with extensive plastic yielding.

Morphologically, it was shown that the presence of carbon fibers within the matrix would induce nucleation and growth of crystallites perpendicular to the fiber surface, i.e., transcrystallization, which might impose considerable influence on the fiber/matrix interfacial interaction and the failure behavior in both the matrix and the interface region [70].

In view of the effect of residence time in the molten state of the PEEK reinforced with carbon fibers (APC-2 prepreg by ICI/Fiberite Company, USA) on the number of spherulites present in the bulk matrix, it was found that increasing the residence time would result in a decrease in the number of spherulties, and a well-defined transcrystalline region was subsequently developed on the carbon fiber surface [74]. Consequently, the unidirectional CF/PEEK composite containing a transcrystalline phase showed a higher transverse tensile strength than that of the matrix, owing to a strong interfacial bond between the carbon fiber and the PEEK matrix.

Gao and Kim [75] also conducted a study on the effect of cooling rate on interlaminar fracture toughness of unidirectional CF/PEEK matrix composites. It was shown that the PEEK resin displayed a remarkable 230% improvement in fracture toughness when the

(36)

cooling rate was changed from 1 to 80 ^oC/min. Furthermore, they also conducted the study on the effect of cooling rate on impact damage performance of CF/PEEK laminates, and compared with CF/epoxy laminates [76]. They concluded that the ability to resist damage initiation upon impact was higher in the order of fast-cooled CF/PEEK, slow-cooled CF/PEEK, and CF/epoxy laminates. Meanwhile, they showed that the threshold impact energy was higher and the compression-after-impact (CAI) strength reduction rate was lower for the fast-cooled laminates than the slow-cooled counterparts, strongly indicating the higher impact tolerance of the former system.

The CF/PEEK composites possess extraordinary strength-to-weight and stiffness-to-weight ratios along the longitudinal (or fiber reinforced) direction, as compared with steel, Al or Ti alloys in Table 1-6 [77]. For this very reason, the CF/PEEK composites can be applied on high-requirement rigid aerospace or aircraft turbomachinery components, such as centrifugal impellers.

In terms of the biomedical applications, the CF/epoxy composite materials can be applied on the external fixation for bone fracture repair because of their lightweight and sufficient strength and stiffness [78]. On the other hand, the CF/PEEK composite materials have been applied on the internal fixation for bone fracture repair by different ways using implants such as wires, pins, screws, plates, and intramedullary nails [78]. Among various materials studied, CF/PEEK composite materials are reported to be biocompatible [79] and have good resistance to hydrolysis and radiation degradation. Except for their high strength and fatigue resistance, the CF/PEEK composite materials have been shown to be biological inertness with no mutagenicity or carcinogenicity. Moreover, the tissue response to CF/PEEK has been described as minimal. In view of the effect of exposure to saline solution (0.9% NaCl) on the flexural and fracture toughness properties of short carbon fiber reinforced PS (polysulfone),

(37)

PBT (polybutylene terephalate) and PEEK composites, CF/PS and CF/PBT composites showed significant degradation of mechanical properties following exposure to saline solution [80]. But there was no such reduction for the CF/PEEK composites, due to good bonding between the carbon fibers and PEEK matrix [81]. Animal studies showed that the CF/PEEK composite elicits minimal response from muscular tissue. Both the in vivo and in vitro aging studies confirmed the mechanical stability of CF/PEEK up to 6 months.

1.4 Particulate filled polymer composites

1.4.1 Characteristics of particulate filled composites

It is well known that the environment can significantly influence the mechanical performance of polymer matrix composites, especially for epoxy-based composites. As a result, the combined influence of moisture and thermal history can cause microcracking to develop along with plasticization; reducing the Tg of resin and increasing the dimension and tolerance of the materials [82]. Srivastava and Hogg [83] conducted their studies on the particle filled polymer composites to investigate moisture absorption behavior in 10 µm Al(OH)3 particle and 40 nm PE particle filled GFRP (glass-fiber reinforced epoxy-vinylester resin). It was found that increasing the filler content in GFRP composites resulted in an increase in the equilibrium water uptake and in turn an increase in the effective water diffusivity coefficient. Moreover, the filled Al(OH)3 GFRP composites revealed a higher content of moisture uptake and diffusivity coefficient than those of the PE-filled and unfilled GFRP composites. In terms of toughness of all composites, it was shown that the mode-I delamination toughness increased with increasing moisture content but changed little under mode-II testing. Furthermore, the Al(OH)3 filled GFRP composites exhibited higher values of mode-I and mode-II fracture toughness than those of the PE-filled and unfilled GFRP

(38)

composites.

In view of damage performance of particle filled GFRP, with no concern about water uptake, it was shown that the interlaminar toughness (GIC and GIIC), absorbed energy, and residual compressive strength values of the GFRP composites increased with increasing particle content [84], due to stress-concentration induced plastic deformation and crack bridging. Based on this postulate, the PE-filled composites revealed higher values of mode-I, mode-II and impact toughness than those of the Al(OH)3 filled composites.

It is believed that polymers with linear molecules of smooth profile will reveal low shear strength, with “running” films on their own surfaces and transfer films on the counterfaces along the sliding direction in rubbing contacts. This, as a result, offers low sliding friction, but suffers high wearing rates. Inclusion of hard fillers, including metals, ceramics, glass, and special polymers such as polytetrafluoroethylene (PTFE, Teflon) and high density polyethylene can reduce the wear rate by up to three orders of magnitude [85,86]. But the negative effect of such hard fillers is an increase in friction and, more importantly, abrasion of the mating counterface. On the other hand, polymers having bulk side groups, brancher or crosslinks reveal better wear resistance than polymers with linear molecules. Inclusion of hard fillers, on the contrary, can provide friction reduction. Burroughs and Kim [87] found that inclusion of boric oxide particles (150 µm) in PTFE and epoxy composite materials can provide PTFE with a two-order reduction in wear rate against stainless steel surface, and under similar environments, can reduce the friction coefficient of epoxy from µ>0.7 to as low as µ=0.07. Through the application of acid-base interactions on absorption and dispersion of particles in polymer matrix, walllastonite (mineral filler) was successfully incorporated into PMMA (polymethyl methacrylate) polymer matrix [88]. It was shown that the tensile modulus of the composite containing wollastonite and surface-coated wollastonite increased

(39)

by 66% and 78%, respectively, as compared with the unfilled PMMA matrix.

The porosity of fillers also imparted an effect on the abrasion resistance of nanoporous silica gel/polymer composites [89]. An organic monomer solution consisted of triethylene glycoldimethacrylate and various initiators was introduced into the silica gel powders which were of different porous structures and their media diameters were 13 µm to form a paste, and subsequently were polymerized inside a glass mold. As a result, it was found that the wear rate of the composites decreased with increasing filler porosity. Through scanning electron microscope (SEM) examinations, it was also suggested that the better wear resistance was associated with fine-scale plastic deformation of the wear surface and the absence of filler particle pullout.

Since the rapid growth of the electronic industry, demand for better packaging materials has become more and more important, especially for those having specific physical properties in combination with electrical insulation. For instance, to improve the life time of organic- light-emitting-diode (OLED) devices, the packaging materials must meet a number of functions, such as heat dissipation, moisture resistance, and electrical insulation. The polymer matrix composites can achieve these conflicting properties. However, the inclusion of ceramic powders into polymer composites was shown to enhance the intrinsically low thermal conductance of the polymers [90-92]. These composites show very different moisture resistance from that of the unfilled polymers [93], due to the presence of polymer filler interfaces. As a consequence, the electrical insulation of the filled polymers will be altered by the presence of the filler phases. To study these issues, Al2O3 (20 or 100 µm)/PU, carbon fibers (Φ8×30 or Φ8×100 µm)/PU, and boron nitride (5-11 µm)/silicone polymer composites were prepared [94], and the effects of moisture uptake on thermal conductance and dielectric relaxation were studied. It was found that water molecules were absorbed not

(40)

only by the polymer matrix, but also by the interfaces introduced by the fillers, and, in turn, the absorbed water molecules induced the phenomenon of dielectric relaxation for all materials. Among these three composite materials, the boron nitride/silicone composite absorbed the least amount of moistures, and accounted for the highest thermal conductivity value as compared with the other two. It was concluded that the boron nitride/silicone could be the candidate for the packaging materials used in electronic devices that require heat dissipation and moisture resistance, in addition to electrical insulation.

Electric field induced particle alignment has been reported for many electrorheological fluid systems [95-99]. The characterizations of field-induced aligned structures have been limited to optical observations with thin layers being confined between glass plates. On the other hand, field-induced particulate alignments in polymer composites were proposed [100].

Prior to polymerization, particles of different shapes, sizes, and dielectric constants could be aligned in a photopolymerized fluid by an electric field. Urethanedimethacrylate (UDMA) mixed with 1.6-hexanediol dimethacrylate (HDDMA) in a 90/10 ratio, which gave a viscosity allowing particles to align. The inclusion was silica-zirconia in two forms: P50, having a particle size of 0.7 µm in average, and Z100, also having a particle size of 0.7 µm.

Applying the photosensitive initiator and accelerator, the particle-aligned UDMA/HDDMA (90/10) resin underwent in situ polymerization under a blue light gun. It was found that the rate of alignment depends on both of the dielectric constants of resin and particle, and on the particle size.

It is known that the electrical conductivity of polymers can be significantly improved by the introduction of metals or carbon black. Except for the change of electrical conductivity of polymers, the promising materials made from piezo ceramic/polymer composites can serve as ultransonic transducers for naval sonar devices, medical diagnostic systems, and

Fabrication and Characterization on High Performance Mg/Carbon-Fiber/PEEK Laminates and Nanoparticle/PEEK

國立中山大學材料科學研究所 博士論文

高性能鎂/碳纖/聚二醚酮夾層及奈米粉體強化聚二醚酮複材之 製備與特性分析