《复合材料 Composites》课程教学资源（学习资料）第五章 陶瓷基复合材料_20fibrous molithic-25 Core/shell volume effect on the microstructure and mechanical properties of fibrous Al2O3–（m-ZrO2）/t-ZrO2 composite

MATERIALS CIENCE EIEERIG ELSEVIER Materials Science and Engineering A 432(2006)317-323 www.elsevier.com/locate/msea Core/shell volume effect on the microstructure and mechanical properties of fibrous Al2O3-(m-Zro2)/t-zrO2 composite Byong Taek Lee a,, Swapan Kumar Sarkar Asit Kumar Gain a Ho Yeon song School of Advanced Materials Engineering, Engineering College, Kongju National University, 182 Shinkwan-dong Kongju City, Chungnam 314-701, South Korea School of Medicine, Soonchunhyang University, 366-1 Ssangyoun-dong, Cheonan Ciry, Chungnam 330-090, South Korea Received 3 March 2006; accepted 10 June 2006 Al2O3-monoclinic-ZrO2 )/tetragonal-zrO2 composites(Al2O3-(m-ZrO2)t-zrO2)were fabricated by the multi-pass extrusion process to intro- duce continuously fibrous textures with different core/shell volume fractions of AlO3-(m-zrO2) and t-zrO2(60/40, 50/50, 40/60, 30/70 and 20/80). The values of the bending strength and fracture toughness increased as the t-ZrO2 shell volume fraction increased, with maximum values of 031 MPa and8. 1 MPam in the 20(core)/80(shell)composite, respectively. However, the Vickers hardness value decreased as the core volume fraction decreased due to the increase of t-ZrO, content. o 2006 Elsevier B. v. All rights reserved Keywords: Core/shell volume effect; Alumina-zirconia; Fibrous microstructure 1. Introduction crack bridging and fibrous pull-out mechanisms. However,to obtain the composites using them, many problems occurred in 1203-ZrO2 composites have been considered excellent the fabrication process and it was quite costly [9] materials for use in many industrial components requiring high Recently, an interesting approach, based on macroscale degrees of strength, wear resistance, and corrosion resistance, microstructure control, using the multi-pass extrusion process, as well as good oxidation and thermal stability at high temper- has been recognized as an effective way to improve the fracture atures, due to their excellent mechanical, thermal and chemical toughness of brittle ceramics. Extensive works and the in-depth properties [1]. The biocompatibility of the Al2O3-ZrO2 com- investigation have been reported by the work of Halloran and posites also makes them appropriate for use in bio-implants, others on Si3N4/BN fibrous monoliths [10-12]. Furthermore, such as total hip and joint prostheses and dental implants [2-4]. using this process, the fibrous microstructure, as well as the con- However, their inherently low fracture toughness limits their tinuously porous bodies, can be easily controlled [9, 13, 14]. In potential for biomedical applications. In order to enhance the previous work, well controlled, fine, homogeneous Al2O3-( fracture toughness, many researchers have focused on the fab- ZrO2)t-ZrO2 fibrous composite of 50/50 core/shell composition rication of composites by dispersion of the second phase, either with excellent mechanical properties was obtained [14]. The ceramic or metal [5]. The approaches have been based on the fracture toughness was remarkably increased(9.6 MPam) incorporation of fibers, whiskers or particles as reinforcement due to toughening mechanisms, such as crack bridging, microc- using ball milling [6], electroless deposition [7], sol-gel process racks and phase transformation. Although it was recognized that B], etc. It has been reported that the whisker and fiber reinforced the material properties of Al2O3/ZrO2 composites are strongly ceramic matrix composites have been focused on during the last dependent on the volume fraction of constituent phases, there decade because of their remarkable fracture toughening due to was no report on the core/shell volume effect on the microstru ture and mechanical properties In this work, the fibrous Al2O3(m-Zro2)/t-zrO2 compos- Corresponding author. Tel. +82 41 8508677: fax: +82 41 8582939 ites with different core/shell volume fractions were fabricated E-mail address: Ibt @kongju ac kr(B T. Lee). using the multi-pass extrusion process. The basic concept of the 0921-5093/S-see front matter O 2006 Elsevier B.V. All rights reserved

Materials Science and Engineering A 432 (2006) 317–323 Core/shell volume effect on the microstructure and mechanical properties of fibrous Al2O3–(m-ZrO2)/t-ZrO2 composite Byong Taek Lee a,∗, Swapan Kumar Sarkar a, Asit Kumar Gain a, Soo-Jae Yimb, Ho Yeon Song b a School of Advanced Materials Engineering, Engineering College, Kongju National University, 182 Shinkwan-dong, Kongju City, Chungnam 314-701, South Korea b School of Medicine, Soonchunhyang University, 366-1 Ssangyoun-dong, Cheonan City, Chungnam 330-090, South Korea Received 3 March 2006; accepted 10 June 2006 Abstract Al2O3–(monoclinic-ZrO2)/tetragonal-ZrO2 composites (Al2O3–(m-ZrO2)/t-ZrO2) were fabricated by the multi-pass extrusion process to intro￾duce continuously fibrous textures with different core/shell volume fractions of Al2O3–(m-ZrO2) and t-ZrO2 (60/40, 50/50, 40/60, 30/70 and 20/80). The values of the bending strength and fracture toughness increased as the t-ZrO2 shell volume fraction increased, with maximum values of 1031 MPa and 8.1 MPa m1/2 in the 20 (core)/80 (shell) composite, respectively. However, the Vickers hardness value decreased as the core volume fraction decreased due to the increase of t-ZrO2 content. © 2006 Elsevier B.V. All rights reserved. Keywords: Core/shell volume effect; Alumina–zirconia; Fibrous microstructure 1. Introduction Al2O3–ZrO2 composites have been considered excellent materials for use in many industrial components requiring high degrees of strength, wear resistance, and corrosion resistance, as well as good oxidation and thermal stability at high temper￾atures, due to their excellent mechanical, thermal and chemical properties [1]. The biocompatibility of the Al2O3–ZrO2 com￾posites also makes them appropriate for use in bio-implants, such as total hip and joint prostheses and dental implants [2–4]. However, their inherently low fracture toughness limits their potential for biomedical applications. In order to enhance the fracture toughness, many researchers have focused on the fab￾rication of composites by dispersion of the second phase, either ceramic or metal [5]. The approaches have been based on the incorporation of fibers, whiskers or particles as reinforcement using ball milling [6], electroless deposition [7], sol–gel process [8], etc. It has been reported that the whisker and fiber reinforced ceramic matrix composites have been focused on during the last decade because of their remarkable fracture toughening due to ∗ Corresponding author. Tel.: +82 41 8508677; fax: +82 41 8582939. E-mail address: lbt@kongju.ac.kr (B.T. Lee). crack bridging and fibrous pull-out mechanisms. However, to obtain the composites using them, many problems occurred in the fabrication process and it was quite costly [9]. Recently, an interesting approach, based on macroscaled microstructure control, using the multi-pass extrusion process, has been recognized as an effective way to improve the fracture toughness of brittle ceramics. Extensive works and the in-depth investigation have been reported by the work of Halloran and others on Si3N4/BN fibrous monoliths [10–12]. Furthermore, using this process, the fibrous microstructure, as well as the con￾tinuously porous bodies, can be easily controlled [9,13,14]. In previous work, well controlled, fine, homogeneous Al2O3–(m￾ZrO2)/t-ZrO2 fibrous composite of 50/50 core/shell composition with excellent mechanical properties was obtained [14]. The fracture toughness was remarkably increased (9.6 MPa m1/2) due to toughening mechanisms, such as crack bridging, microc￾racks and phase transformation. Although it was recognized that the material properties of Al2O3/ZrO2 composites are strongly dependent on the volume fraction of constituent phases, there was no report on the core/shell volume effect on the microstruc￾ture and mechanical properties. In this work, the fibrous Al2O3–(m-ZrO2)/t-ZrO2 compos￾ites with different core/shell volume fractions were fabricated using the multi-pass extrusion process. The basic concept of the 0921-5093/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2006.06.058

B T. Lee et al. Materials Science and Engineering A 432(2006)317-323 microstructure design was to introduce different volume ratios eter. The process was repeated until the fourth passed filaments of core/shell structure; i. e. (1)the core comprised with Al2O3 were obtained. The extrusion temperature and rate were 120C and 25 vol. m-zrO2 that enhance the microcracks and crack and 8 mm/min, respectively. The binder was removed from the deflection, (2)the shell of t-ZrO2 phase that can be introduced green body during the binder burning-out process, which was with the phase transformation toughening mechanism and (3)the carried out at 700C under a flowing nitrogen atmosphere. The fibrous microstructure control that leads to the crack deflection samples were then sintered at 1450C for 2 h in air. The den nd crack bridging phenomena. Furthermore, the relationship sity of the sintered bodies was measured using the Archimedes between microstructure and material properties was investi- Method. The hardness of the sintered bodies was measured gated, depending on the volume ratio of core/shell. using the Vickers Hardness Tester(Hv-1 12, Akashi, Yokohama, Japan)under a load of l kg for 10s. To investigate the trend of 2. Experimental procedure variation of the fracture toughness, Kic was measured by the indentation method using Evans Equation. The equation is The multi-pass extrusion process was carried out by mixing KIC=0.16H a/(c/a-3/2 the ceramic powder with an organic binder, making a feed roll of preferential geometry and repeatedly extruding the successive where H is the Vickers hardness, a the half of indentation diag- nic zanbA rouping them in a predetermined arrangement. onal and c is the half of crack length from indentation center filaments after In order to measure the bending strength, a four point bend- (TZ-3Y, Tosoh, Japan)had an average particle size of about 300, (Unitech", R&B, Korea). The bend bars were cut 30mm in 60 and 80 nm, respectively. To fabricate the core/ shell structure length and the diameter was 2.65 mm. No surface preparation of A12O3-(25%m-ZrO2)t-ZrO2 composites, the Al2O3-(25% was made because the sintered specimens were smooth surfaced m-ZrO2) powders were mixed by ball milling for 24 h. These and round shape(see Fig. 7). Microstructural variation of the mixtures of Al2O3-(m-zrO2)powders and tetragonal zirconia fibrous composites with the volume fraction was examined using (t-ZrO2) powder were mixed separately with a polymer binder a Field Emission Scanning Electron Microscope(FE-SEM,JSM sing a heated blender(Shina Platec Co., Suwon, South Korea) 6335F, JEOL, Tokyo, Japan)and Transmission Electron Micro- at a temperature of 120C. Ethylene vinyl acetate(EVA)(Elvax scope (TEM, JEM2010, JEOL, Tokyo, Japan) 250 and Elvax 210, Dupont, USA) polymers were used as binders of the ceramic powders. For lubrication during blend 3. Results and discussion ing, stearic acid( CH3(CH2)16COOH, Daejung Chemicals Metals Co. Ltd, Korea) was added. Mixtures of Al2O3-(m- Fig. 1 shows the cross-sectional SEM images of the ZrO2 )polymer(depicted as AP)was extruded as a cylindrical fourth passed Al203-(m-zrO2/1-ZrO2 composites, where the rod shape and ZrO/polymer(depicted as zp)was compacted core/shell was: (a)60/40,(b)40/60 and(c)20/80. The black into the shell shape by warm pressing. After combining them and white contrasts are Al2 03-(m-ZrO2)core and t-ZrO2 shell together, they were made into a feed roll 30 mm in diameter. Dif- respectively. Thus, it is confirmed that a t-zrO2 phase was seen ferent volume fractions in the core/shell were made by changing with a network-type microstructure. The white contrast parti the core diameter and shell thickness. The feed roll was mounted cles(m-zrO2)in the black cores were homogenously dispersed on a heated extrusion die and then extruded with about a 73: 1 in the Al2O3 matrix. A hexagonal, cell-like microstructure extrusion ratio. Compositions of the five fibrous composites are indicated with dotted lines, was observed in all composites, given in Table 1. The shell thickness and core diameter of the independent of volume fraction of core/shell. These cells were individual composition were made with standard die so that the comprised with 61 core/shell structures and they were the indi- overall volume fraction of each phase matches with those shown vidual second passed filaments. Although the second passed in Table 1. The volume fractions of the composites described in filaments were circular-shaped in the cross section, during the this paper is thus stands for those of the green bodies. repeated extrusions, they changed to a hexagon-like shape, to The feed roll for each composition was extruded to obtain the form a stable frame structure. The Al203-(m-ZrO2)core diam- first passed filaments, which were 3.5 mm in diameter. The first eter decreased as the core volume fraction decreased from 60 to passed filaments were arranged in the same die and extruded 20 vol g as shown in Fig. 1(a-c). However, the shell thickness again to obtain the second passed filaments with the same diam- increased. The microstructure of the composites gives clear evi dence of the microstructural variation retaining the homogeneity Table I However, the morphology of a few Al2O3-(m-ZrO2)cores devi- Composition of the fibrous Al2O3-(m-ZrOz)t-ZrO2 composites ated at the boundary regions which were previously mentioned as individual second passed cells. Similar change in geometry Al2O3-(m-ZrO2)(vol%) t-ZxOz(vol%) in the green body with extrusion pass was reported in the work Hallora more one pass extrusion and the geometrical symmetry was still 30/70 retained in some extent. In general, during the extrusion pro- cess to obtain the second passed filaments, the circumferential first passed filaments received severe friction near the extru-

318 B.T. Lee et al. / Materials Science and Engineering A 432 (2006) 317–323 microstructure design was to introduce different volume ratios of core/shell structure; i.e. (1) the core comprised with Al2O3 and 25 vol.% m-ZrO2 that enhance the microcracks and crack deflection, (2) the shell of t-ZrO2 phase that can be introduced with the phase transformation toughening mechanism and (3) the fibrous microstructure control that leads to the crack deflection and crack bridging phenomena. Furthermore, the relationship between microstructure and material properties was investi￾gated, depending on the volume ratio of core/shell. 2. Experimental procedure The multi-pass extrusion process was carried out by mixing the ceramic powder with an organic binder, making a feed roll of preferential geometry and repeatedly extruding the successive filaments after grouping them in a predetermined arrangement. The starting Al2O3 powder (AKP-50, Sumimoto, Japan), mono￾clinic ZrO2 (TZ-0Y, Tosoh, Japan) and tetragonal ZrO2 powders (TZ-3Y, Tosoh, Japan) had an average particle size of about 300, 80 and 80 nm, respectively. To fabricate the core/shell structure of Al2O3–(25% m-ZrO2)/t-ZrO2 composites, the Al2O3–(25% m-ZrO2) powders were mixed by ball milling for 24 h. These mixtures of Al2O3–(m-ZrO2) powders and tetragonal zirconia (t-ZrO2) powder were mixed separately with a polymer binder using a heated blender (Shina Platec. Co., Suwon, South Korea) at a temperature of 120 ◦C. Ethylene vinyl acetate (EVA) (Elvax 250 and Elvax 210, Dupont, USA) polymers were used as binders of the ceramic powders. For lubrication during blend￾ing, stearic acid (CH3(CH2)16COOH, Daejung Chemicals & Metals Co. Ltd., Korea) was added. Mixtures of Al2O3–(m￾ZrO2)/polymer (depicted as AP) was extruded as a cylindrical rod shape and ZrO2/polymer (depicted as ZP) was compacted into the shell shape by warm pressing. After combining them together, they were made into a feed roll 30 mm in diameter. Dif￾ferent volume fractions in the core/shell were made by changing the core diameter and shell thickness. The feed roll was mounted on a heated extrusion die and then extruded with about a 73:1 extrusion ratio. Compositions of the five fibrous composites are given in Table 1. The shell thickness and core diameter of the individual composition were made with standard die so that the overall volume fraction of each phase matches with those shown in Table 1. The volume fractions of the composites described in this paper is thus stands for those of the green bodies. The feed roll for each composition was extruded to obtain the first passed filaments, which were 3.5 mm in diameter. The first passed filaments were arranged in the same die and extruded again to obtain the second passed filaments with the same diam￾Table 1 Composition of the fibrous Al2O3–(m-ZrO2)/t-ZrO2 composites Sample ID Al2O3–(m-ZrO2) (vol%) t-ZrO2 (vol%) 60/40 60 40 50/50 50 50 40/60 40 60 30/70 30 70 20/80 20 SO eter. The process was repeated until the fourth passed filaments were obtained. The extrusion temperature and rate were 120 ◦C and 8 mm/min, respectively. The binder was removed from the green body during the binder burning-out process, which was carried out at 700 ◦C under a flowing nitrogen atmosphere. The samples were then sintered at 1450 ◦C for 2 h in air. The den￾sity of the sintered bodies was measured using the Archimedes Method. The hardness of the sintered bodies was measured using the Vickers Hardness Tester (HV-112, Akashi, Yokohama, Japan) under a load of 1 kg for 10 s. To investigate the trend of variation of the fracture toughness, KIC was measured by the indentation method using Evan’s Equation. The equation is, KIC = 0.16 H a1/2(c/a) −3/2 where H is the Vickers hardness, a the half of indentation diag￾onal and c is the half of crack length from indentation center. In order to measure the bending strength, a four point bend￾ing test was carried out using a Universal Testing Machine (UnitechTM, R&B, Korea). The bend bars were cut 30 mm in length and the diameter was 2.65 mm. No surface preparation was made because the sintered specimens were smooth surfaced and round shape (see Fig. 7). Microstructural variation of the fibrous composites with the volume fraction was examined using a Field Emission Scanning Electron Microscope (FE-SEM, JSM 6335F, JEOL, Tokyo, Japan) and Transmission Electron Micro￾scope (TEM, JEM2010, JEOL, Tokyo, Japan). 3. Results and discussion Fig. 1 shows the cross-sectional SEM images of the fourth passed Al2O3–(m-ZrO2)/t-ZrO2 composites, where the core/shell was: (a) 60/40, (b) 40/60 and (c) 20/80. The black and white contrasts are Al2O3-(m-ZrO2) core and t-ZrO2 shell, respectively. Thus, it is confirmed that a t-ZrO2 phase was seen with a network-type microstructure. The white contrast parti￾cles (m-ZrO2) in the black cores were homogenously dispersed in the Al2O3 matrix. A hexagonal, cell-like microstructure, as indicated with dotted lines, was observed in all composites, independent of volume fraction of core/shell. These cells were comprised with 61 core/shell structures and they were the indi￾vidual second passed filaments. Although the second passed filaments were circular-shaped in the cross section, during the repeated extrusions, they changed to a hexagon-like shape, to form a stable frame structure. The Al2O3-(m-ZrO2) core diam￾eter decreased as the core volume fraction decreased from 60 to 20 vol% as shown in Fig. 1(a–c). However, the shell thickness increased. The microstructure of the composites gives clear evi￾dence of the microstructural variation retaining the homogeneity. However, the morphology of a few Al2O3–(m-ZrO2) cores devi￾ated at the boundary regions which were previously mentioned as individual second passed cells. Similar change in geometry in the green body with extrusion pass was reported in the work Halloran and his group [10,15]. However, in this work we did more one pass extrusion and the geometrical symmetry was still retained in some extent. In general, during the extrusion pro￾cess to obtain the second passed filaments, the circumferential first passed filaments received severe friction near the extru-

B T Lee et al. Materials Science and Engineering A 432(2006)317-323 Fig. 1. Cross-sectional SEM images of fourth passed Al2O3-m-zrO2 Mt-ZrO2 composites depending on volume fractions of core/shell:(a)60/40, (b)40/60 and(c) 20/80. (a (c) (d) 2um Longitudinal sectional SEM images of fourth passed Al2O3-(m-ZrO2)t-ZrO2 composites depending on volume fractions of core/shell: (a and b)60/40, and

B.T. Lee et al. / Materials Science and Engineering A 432 (2006) 317–323 319 Fig. 1. Cross-sectional SEM images of fourth passed Al2O3–(m-ZrO2)/t-ZrO2 composites depending on volume fractions of core/shell: (a) 60/40, (b) 40/60 and (c) 20/80. Fig. 2. Longitudinal sectional SEM images of fourth passed Al2O3–(m-ZrO2)/t-ZrO2 composites depending on volume fractions of core/shell: (a and b) 60/40, and (c and d) 20/80

B T. Lee et al. Materials Science and Engineering A 432(2006)317-323 ion die wall. Thus, an irregular flow of the mixtures(AP and Table 2 ZP)occurred and the original circular shape was deformed. The Core diameter and shell thickness of the Al203-(m-ZrO2)/O2 composites geometric shape of the cores was nearly circular in cross sec- Sample ID Al2O3-(m-ZrO2)core t-ZrO, shell tion and was uniformly dispersed within the t-ZrO2 Phase with diameter (um) thickness (um) quite similar thickness all over the composite. However, a small deviation of geometry compared with the as received feed roll 3.6 1.3 was observed. This was because the rheological property of core 40/60 3.3 1.8 (AP)and shell(ZP)was not exactly the same and the slight mis- 3070 2.5 match makes a difference in the linear flow behavior of the core 2.2 3.4 AP)and shell(zP)materials. Moreover, to obtain the successive passed filaments, the cylindrical filaments were pre-compacted observed without any processing defects such as large cracks in the extrusion die before carrying out every extrusion. This or shrinkage cavities. In the shell region(b), taken from the caused a lateral flow of the green materials to cover the gaps in P mark in Fig. 3(a), it is confirmed that the average grain size between the individual filaments, thereby contributing a slight was about 250 nm in diameter. However, in the Al2O3(m-zrO2 deformation. But the deviation was nominal and homogeneous core region(c)taken from Q mark in Fig 3(a), the dark and bright as well as very fine scaled microstructure was obtained contrasts were Al2O3 and m-zrO2 phases and their average grain ig. 2 shows the longitudinal section SEM images of the sizes were about 300 and 200 nm, respectively Al2O3-m-Zro2)/t-zrO2 composites which are comprised with ig. 4 shows the TEM images of the 60/40 composites. In (a)60/40 and(c)20/80 composites. In the low magnification the low magnification image(a), core/shell microstructure with images(a, c), continuous fibers were clearly observed. The bright and dark contrast was clearly observed In the enlarged fibrous monoliths of the core and shell phases were prolonged image of the core region(b), a strain field was observed due to the through the entire length. The uni-directional alignment of the phase transformation of t-m Zro2 during the sintering process fibrous monoliths was well-maintained, making a homogeneous [16, 17]. Some dislocations and twin defects were observed in microstructure In the enlarged images(b, d), it was clear that the the ZrO2 and Al2O3 grains shell thickness remarkably increased as the volume fraction of The dependence of relative density and Vickers hardness on the shell increased. The average values of t-ZrO2 shell thickness the core/shell volume fraction of the Al2O3-(m-ZrO2)/t-ZrO2 and the Al2O3(m-ZrO2)core diameter for all the composites composites is shown in Fig. 5. To calculate the relative density, are shown in Table 2 the initial volume fractions of the constituent phases tailored Fig 3(a) shows the enlarged cross-sectional SEM images of in the green ceramic was used in case of sintered ceramic too 60/40 composite Sound core/shell microstructure was clearly As the volume fraction of AlO(m-ZrO2)core decreased, the rums 200nm Fig 3. SEM cross-sectional images of (a) fourth passed Al2O3-(m-ZrO2)t-ZrO2 composites(60/40).(b and c) Enlarged images of shell and core region

320 B.T. Lee et al. / Materials Science and Engineering A 432 (2006) 317–323 sion die wall. Thus, an irregular flow of the mixtures (AP and ZP) occurred and the original circular shape was deformed. The geometric shape of the cores was nearly circular in cross sec￾tion and was uniformly dispersed within the t-ZrO2 phase with quite similar thickness all over the composite. However, a small deviation of geometry compared with the as received feed roll was observed. This was because the rheological property of core (AP) and shell (ZP) was not exactly the same and the slight mis￾match makes a difference in the linear flow behavior of the core (AP) and shell (ZP) materials. Moreover, to obtain the successive passed filaments, the cylindrical filaments were pre-compacted in the extrusion die before carrying out every extrusion. This caused a lateral flow of the green materials to cover the gaps in between the individual filaments, thereby contributing a slight deformation. But the deviation was nominal and homogeneous as well as very fine scaled microstructure was obtained. Fig. 2 shows the longitudinal section SEM images of the Al2O3–(m-ZrO2)/t-ZrO2 composites which are comprised with (a) 60/40 and (c) 20/80 composites. In the low magnification images (a, c), continuous fibers were clearly observed. The fibrous monoliths of the core and shell phases were prolonged through the entire length. The uni-directional alignment of the fibrous monoliths was well-maintained, making a homogeneous microstructure. In the enlarged images (b, d), it was clear that the shell thickness remarkably increased as the volume fraction of the shell increased. The average values of t-ZrO2 shell thickness and the Al2O3–(m-ZrO2) core diameter for all the composites are shown in Table 2. Fig. 3(a) shows the enlarged cross-sectional SEM images of 60/40 composite. Sound core/shell microstructure was clearly Table 2 Core diameter and shell thickness of the Al2O3–(m-ZrO2)/t-ZrO2 composites Sample ID Al2O3–(m-ZrO2) core diameter (
m) t-ZrO2 shell thickness (
m) 60/40 3.8 1.0 50/50 3.6 1.3 40/60 3.3 1.8 30/70 2.9 2.5 20/80 2.2 3.4 observed without any processing defects such as large cracks or shrinkage cavities. In the shell region (b), taken from the P mark in Fig. 3(a), it is confirmed that the average grain size was about 250 nm in diameter. However, in the Al2O3–(m-ZrO2) core region (c) taken from Q mark in Fig. 3(a), the dark and bright contrasts were Al2O3 and m-ZrO2 phases and their average grain sizes were about 300 and 200 nm, respectively. Fig. 4 shows the TEM images of the 60/40 composites. In the low magnification image (a), core/shell microstructure with bright and dark contrast was clearly observed. In the enlarged image of the core region (b), a strain field was observed due to the phase transformation of t–m ZrO2 during the sintering process [16,17]. Some dislocations and twin defects were observed in the ZrO2 and Al2O3 grains. The dependence of relative density and Vickers hardness on the core/shell volume fraction of the Al2O3–(m-ZrO2)/t-ZrO2 composites is shown in Fig. 5. To calculate the relative density, the initial volume fractions of the constituent phases tailored in the green ceramic was used in case of sintered ceramic too. As the volume fraction of Al2O3–(m-ZrO2) core decreased, the Fig. 3. SEM cross-sectional images of (a) fourth passed Al2O3–(m-ZrO2)/t-ZrO2 composites (60/40). (b and c) Enlarged images of shell and core regions

B T Lee et al. Materials Science and Engineering A 432(2006)317-323 (b) Strain field m-2 rOOm Fig 4. TEM images of fourth passed Al2O3-(m-ZrO2)t-ZrO2 composite (a)60/40 Composite and(b)enlarged image of Al2O3-( m-ZrO2)core relative density slightly decreased and their values were around and intact. The higher bending strength was achieved probably 98%0. However, the vickers hardness significantly decreased as because of the unique processing route and lack of large flaws the shell thickness increased. Their values in 60/40 and 20/80 and processing defects. The role of residual stress might have composites were about 1443 and 1333(Hv), respectively. The an effect as the thermal expansion coefficient( CTE)of Al2O3 reason for showing a decrease of hardness as the core volume and ZrO2 phases are different. However, further investigation fraction decreased may be due to the increase in ZrO2 content, is needed for the insight for this particular core/shell system of which has lower hardness than Al2O3 Al2O3-(m-zrO2)/t-ZrO2. The fracture toughness also improved Fig 6 shows the bending strength and fracture toughness of as the t-zrO2 content of the composites increased. For the 60/40 the fourth passed Al2O3-(m-zrO2)/t-zrO2 composites depend- composite, the value of fracture toughness was 5.1 MPam ing on the volume fraction of shell and core. Bending strength whereas for the 20/80 composites, the value increased remark increased almost linearly from 847 to 1031 MPa with increased ably to 8. 1 MPam".The higher fracture toughness value is due t-ZrO2 phase. However, the bending strength of the fibrous to the higher volume fraction of t-ZrO2, which undergoes a stress monolithic composites is higher compared to the monolithic induced t-m phase transformation and absorbs the crack prop- t-ZrO2(830 MPa)or Al2O3-(m-ZrO2)(622 MPa)made by the agation energy. However the value is intermediate compared to same processing rout using single pass extrusion. The values that reported elsewhere [9, 13, 18] are higher compared to those made by analogous process- During the bending strength measurement, all samples did ing route [18]. The result indicates that the core/shell fibrous not show the flat fracture surface which was frequently observed microstructure also improves the bending strength. However, in the ceramic sintered bodies. Fig. 7 shows the fracture mode the observation shows opposite trend compared to the Si3N4- of the fourth passed(a)60/40 and(b)20/80 sintered bodies BN fibrous monoliths [ 19]. In this system, the weak inter-phase In the 20/80 composite, the crack path was larger than that in layer of the bn cause the degradation of the bending strength. the 60/40 composite. Although the bending strength for the for But in our system, the two phases were very strongly bonded mer one increased remarkably, still the eflection. The trend was observed in all the samples we have 350 1300 60 60/405050 30702080 Volume fraction(core/shell) Volume fraction(core/shell) zro2 composites depending on volume fractions of core/shell ZrO2 )/t-ZrO2 composites depending on volume fraction

B.T. Lee et al. / Materials Science and Engineering A 432 (2006) 317–323 321 Fig. 4. TEM images of fourth passed Al2O3–(m-ZrO2)/t-ZrO2 composite. (a) 60/40 Composite and (b) enlarged image of Al2O3–(m-ZrO2) core. relative density slightly decreased and their values were around 98%. However, the Vickers hardness significantly decreased as the shell thickness increased. Their values in 60/40 and 20/80 composites were about 1443 and 1333 (Hv), respectively. The reason for showing a decrease of hardness as the core volume fraction decreased may be due to the increase in ZrO2 content, which has lower hardness than Al2O3. Fig. 6 shows the bending strength and fracture toughness of the fourth passed Al2O3–(m-ZrO2)/t-ZrO2 composites depend￾ing on the volume fraction of shell and core. Bending strength increased almost linearly from 847 to 1031 MPa with increased t-ZrO2 phase. However, the bending strength of the fibrous monolithic composites is higher compared to the monolithic t-ZrO2 (830 MPa) or Al2O3–(m-ZrO2) (622 MPa) made by the same processing rout using single pass extrusion. The values are higher compared to those made by analogous process￾ing route [18]. The result indicates that the core/shell fibrous microstructure also improves the bending strength. However, the observation shows opposite trend compared to the Si3N4- BN fibrous monoliths [19]. In this system, the weak inter-phase layer of the BN cause the degradation of the bending strength. But in our system, the two phases were very strongly bonded Fig. 5. Relative density and Vickers hardness of fourth passed Al2O3–(m￾ZrO2)/t-ZrO2 composites depending on volume fractions of core/shell. and intact. The higher bending strength was achieved probably because of the unique processing route and lack of large flaws and processing defects. The role of residual stress might have an effect as the thermal expansion coefficient (CTE) of Al2O3 and ZrO2 phases are different. However, further investigation is needed for the insight for this particular core/shell system of Al2O3–(m-ZrO2)/t-ZrO2. The fracture toughness also improved as the t-ZrO2 content of the composites increased. For the 60/40 composite, the value of fracture toughness was 5.1 MPa m1/2, whereas for the 20/80 composites, the value increased remark￾ably to 8.1 MPa m1/2. The higher fracture toughness value is due to the higher volume fraction of t-ZrO2, which undergoes a stress induced t–m phase transformation and absorbs the crack prop￾agation energy. However the value is intermediate compared to that reported elsewhere [9,13,18]. During the bending strength measurement, all samples did not show the flat fracture surface which was frequently observed in the ceramic sintered bodies. Fig. 7 shows the fracture mode of the fourth passed (a) 60/40 and (b) 20/80 sintered bodies. In the 20/80 composite, the crack path was larger than that in the 60/40 composite. Although the bending strength for the for￾mer one increased remarkably, still the crack has considerable deflection. The trend was observed in all the samples we have Fig. 6. Bending strength and fracture toughness of fourth passed of Al2O3–(m￾ZrO2)/t-ZrO2 composites depending on volume fractions of core/shell

B T. Lee et al. Materials Science and Engineering A 432(2006)317-323 I ig. 7. Fracture mode showing crack deflection in fibrous AlzO3-(m-ZrO2 Vt-ZrO2 composites(a)60/40 and(b)20/80 composites. tested compared to the monolithic t-ZrO2 and Al2O3-(m-zrO2), 20/80 core/shell sample(44.72+ 14 um). The shortening of the where typical flat fracture surface was observed. This infers that crack length is due to the higher content of t-ZrO2 phase due the fibrous microstructure has an effect on the crack deflection. to its stress induced martensitic t-m phase transformation. The To clearly understand the fracture behavior of the fourth enlarged images(b, d)were taken from the crack tip regions passed filament, cracks were made by Vickers indention. Fig 8 marked with arrows in Fig 8(a and c), respectively. From the shows crack propagation of(a)60/40 and(c) 20/80 compos- enlarged images(b, d), it was clear that the crack propagation ites. In the low magnification images(a, c), typical median was an intra and intergranular type cracks were propagated from the corner of an indention site ig. 9 shows SEM fracture surfaces of the 4th passed(a) with deflection. However, the average crack length in the 60/40 60/40 and(b)20/80 composites. The typical mixed fracture with core/shell sample (78.47+8.6 um) is longer than that of the intergranular and transgranular modes, which are rough and flat (d) 104m Fig 8. SEM crack propagation of fourth passed (a)60/40 and(c)20/80 core/shell composites (b and d) Enlarged images of their crack tips respectively

322 B.T. Lee et al. / Materials Science and Engineering A 432 (2006) 317–323 Fig. 7. Fracture mode showing crack deflection in fibrous Al2O3–(m-ZrO2)/t-ZrO2 composites. (a) 60/40 and (b) 20/80 composites. tested compared to the monolithic t-ZrO2 and Al2O3–(m-ZrO2), where typical flat fracture surface was observed. This infers that the fibrous microstructure has an effect on the crack deflection. To clearly understand the fracture behavior of the fourth passed filament, cracks were made by Vickers indention. Fig. 8 shows crack propagation of (a) 60/40 and (c) 20/80 compos￾ites. In the low magnification images (a, c), typical median cracks were propagated from the corner of an indention site with deflection. However, the average crack length in the 60/40 core/shell sample (78.47 ± 8.6
m) is longer than that of the 20/80 core/shell sample (44.72 ± 14
m). The shortening of the crack length is due to the higher content of t-ZrO2 phase due to its stress induced martensitic t–m phase transformation. The enlarged images (b, d) were taken from the crack tip regions marked with arrows in Fig. 8(a and c), respectively. From the enlarged images (b, d), it was clear that the crack propagation was an intra and intergranular type. Fig. 9 shows SEM fracture surfaces of the 4th passed (a) 60/40 and (b) 20/80 composites. The typical mixed fracture with intergranular and transgranular modes, which are rough and flat Fig. 8. SEM crack propagation of fourth passed (a) 60/40 and (c) 20/80 core/shell composites. (b and d) Enlarged images of their crack tips respectively

B T. Lee et al. Materials Science and Engineering A 432(2006)317-323 Fig 9. SEM fracture surfaces of 4th passed(a)60/40 and(b)20/80 core/shell Al2O3-(m-zrO2)/t-ZrO2 composites. surfaces respectively, was observed in both samples. In the t- the fourth passed filaments. The sizes of Al2O3-(m-ZrO2)core ZrO2 shell region, the main fracture mode was the intergranular for 60/40, 50/50, 40/60, 30/70 and 20/80 composites were about type and very rough compared with the Al2O3-(m-ZrO2)core 3.8, 3.6, 3.3, 2.9 and 2.2 um in diameter, respectively. However, region, as shown in Fig 9(b). However, in the 20/80 composite, the shell thicknesses were about 1, 1.3, 1.8, 2.5 and 3. 4 um, the Al2O3-(m-ZrO2)core region, as indicated by dotted lines, respectively. Vickers hardness decreased with the increase clearly appeared on the fracture surface, but the pullout phe- shell volume fraction of the ZrO2 phase. The values of bending nomenon of fibers was not found strength and fracture toughness increased as shell volume frac- In this work, the fibrous monolithic Al2O3-(m-zrO2)t-zroz tion increased and showed a maximum value of 1031 MPa and ceramics with core/shell microstructure were developed with 8.1 MPam, respectively, in the 20/80 composite different core and shell volume fractions. The material prop erties of the fibrous Al203-(m-zrO2)t-zrO2 composites were Acknowledgement remarkably improved due to the formation of fibrous microstruc- ture. The high fracture toughness value was achieved due to the The work was supported by the nrl project of the Korean fine microstructure control with fibrous type and incorporation Ministry of Science and Technology of a few fracture toughening mechanisms due to the mismatch of thermal expansion and lattice parameters of constituent phases. References A high value of bending strength(1031 MPa) was obtained in the 20/80 composite. The reason for the high bending strengt [1] B.T. Lee, A. Nishiyama, K. Hiraga, Mater. Trans., JIM 34(1993) is basically due to the formation of uni-directionally aligned 682-686. fibrous microstructure with a unique processing. With the micro- [2] P.L. B.Tan, J.T.Dunne, J Prosthet Dent. 91(2004)215-218 [3 J. Chevalier, Biomaterials 27(2006)535-543 scaled fibrous microstructure control, it was expected that a pull [4) H.J. Fruh, G Willmann, H.G. Pfaff, Biomaterials 18(1997)873-876 out phenomenon might occur, but there was no evidence of [5] M.H. Bocanegra-Bernal, D D.L. Torre, J. Mater. Sci. 37(2002) this on the fracture surface of the composites. The 4947-4971 son was that the fibrous interfaces of Al2O3-(m-ZrO2)core an [6] B K Jang, Mater. Chem. Phys. 93 (2005)337-341 ZrO2 shell were strongly bonded. Thus, cracks were prop- [7 L.H. Oh, J.Y. Lee, J K. Han, H.. Lee, B T. Lee, Surf. Coat. Technol 192(2005)39-42. agated into the sintered bodies without a strong relation to [8)BT. Lee,IKHan, F. Saito, Mater. Lett. 59(2005)355-360 the fibrous morphology. However, as the shell volume fraction [9] B.T. Lee, K.H. Kim, J.K. Han, J. Mater. Res. 19(2004)3234-3241 increased, the fracture toughness and strength increased due to [10] C V. Hoy, A. Barda, M. Griffith, J.W.Halloran, J. Am. Ceram Soc. 81 the large volume percent of ZrO2 phase. To introduce the pull (1998)152-158. out phenomenon of fibrous microstructure as well as severe [1l] R W. Trice, J w. Halloran, J. Am. Ceram. Soc. 82(1999)2563-2565 crack deflection, the composites should be made with weak [12]C Wang, Y Huang, Q Zan, H Guo, S Cai, Mater. SciEng,C11 (2000)912 interfaces [13] B.T. Lee, D H Jang, I.C. Kang, C.W. Lee, J. Am. Ceram Soc. 88(2005) 2874-2878 4. Conclusion [14] B.T. Lee, I.C. Kang, S.H. Cho, H.Y. Song, J Am Ceram Soc. 88(2005) [15] D. Kovar, B.H. King, R.w. Trice, J W. Halloran, J. Am. Ceram. Soc. Fibrous Al2O3-(m-ZrO2)/t-ZrO2 composites with different 80(1997)2471-2487 core/shell volume fractions (60/40, 50/50, 40/60, 30/70 and [16] B.T. Lee, K. Hiraga, J Mater Res. 9(1994)1199-1207. 20/80)were fabricated using the multi-pass extrusion process, [17 B T. Lee, K. Hiraga, D. Shindo, A. Nishiyama, J Mater. Sci. 29(1994) and the dependence of the material properties on the volume [18 A.P. Quintin, M.H. Berger, A.R. Bunsell, C E. G. Butler. A. Woot. fractions was investigated. The microstructure was successfull controlled both in transverse and longitudinal directions up to [19]RW. Trice, J.W. Halloran, J. Am. Ceram Soc. 83(2000)311-316

B.T. Lee et al. / Materials Science and Engineering A 432 (2006) 317–323 323 Fig. 9. SEM fracture surfaces of 4th passed (a) 60/40 and (b) 20/80 core/shell Al2O3–(m-ZrO2)/t-ZrO2 composites. surfaces respectively, was observed in both samples. In the t￾ZrO2 shell region, the main fracture mode was the intergranular type and very rough compared with the Al2O3–(m-ZrO2) core region, as shown in Fig. 9(b). However, in the 20/80 composite, the Al2O3-(m-ZrO2) core region, as indicated by dotted lines, clearly appeared on the fracture surface, but the pullout phe￾nomenon of fibers was not found. In this work, the fibrous monolithic Al2O3–(m-ZrO2)/t-ZrO2 ceramics with core/shell microstructure were developed with different core and shell volume fractions. The material prop￾erties of the fibrous Al2O3–(m-ZrO2)/t-ZrO2 composites were remarkably improved due to the formation of fibrous microstruc￾ture. The high fracture toughness value was achieved due to the fine microstructure control with fibrous type and incorporation of a few fracture toughening mechanisms due to the mismatch of thermal expansion and lattice parameters of constituent phases. A high value of bending strength (1031 MPa) was obtained in the 20/80 composite. The reason for the high bending strength is basically due to the formation of uni-directionally aligned fibrous microstructure with a unique processing. With the micro￾scaled fibrous microstructure control, it was expected that a pull out phenomenon might occur, but there was no evidence of this on the fracture surface of the composites. The main rea￾son was that the fibrous interfaces of Al2O3–(m-ZrO2) core and t-ZrO2 shell were strongly bonded. Thus, cracks were prop￾agated into the sintered bodies without a strong relation to the fibrous morphology. However, as the shell volume fraction increased, the fracture toughness and strength increased due to the large volume percent of ZrO2 phase. To introduce the pull out phenomenon of fibrous microstructure as well as severe crack deflection, the composites should be made with weak interfaces. 4. Conclusion Fibrous Al2O3–(m-ZrO2)/t-ZrO2 composites with different core/shell volume fractions (60/40, 50/50, 40/60, 30/70 and 20/80) were fabricated using the multi-pass extrusion process, and the dependence of the material properties on the volume fractions was investigated. The microstructure was successfully controlled both in transverse and longitudinal directions up to the fourth passed filaments. The sizes of Al2O3–(m-ZrO2) core for 60/40, 50/50, 40/60, 30/70 and 20/80 composites were about 3.8, 3.6, 3.3, 2.9 and 2.2
m in diameter, respectively. However, the shell thicknesses were about 1, 1.3, 1.8, 2.5 and 3.4
m, respectively. Vickers hardness decreased with the increase of shell volume fraction of the ZrO2 phase. The values of bending strength and fracture toughness increased as shell volume frac￾tion increased and showed a maximum value of 1031 MPa and 8.1 MPa m1/2, respectively, in the 20/80 composite. Acknowledgement The work was supported by the NRL project of the Korean Ministry of Science and Technology. References [1] B.T. Lee, A. Nishiyama, K. Hiraga, Mater. Trans., JIM 34 (1993) 682–686. [2] P.L.B. Tan, J.T. Dunne, J. Prosthet. Dent. 91 (2004) 215–218. [3] J. Chevalier, Biomaterials 27 (2006) 535–543. [4] H.J. Fruh, G. Willmann, H.G. Pfaff, Biomaterials 18 (1997) 873–876. [5] M.H. Bocanegra-Bernal, D.D.L. Torre, J. Mater. Sci. 37 (2002) 4947–4971. [6] B.K. Jang, Mater. Chem. Phys. 93 (2005) 337–341. [7] I.H. Oh, J.Y. Lee, J.K. Han, H.J. Lee, B.T. Lee, Surf. Coat. Technol. 192 (2005) 39–42. [8] B.T. Lee, J.K. Han, F. Saito, Mater. Lett. 59 (2005) 355–360. [9] B.T. Lee, K.H. Kim, J.K. Han, J. Mater. Res. 19 (2004) 3234–3241. [10] C.V. Hoy, A. Barda, M. Griffith, J.W. Halloran, J. Am. Ceram. Soc. 81 (1998) 152–158. [11] R.W. Trice, J.W. Halloran, J. Am. Ceram. Soc. 82 (1999) 2563–2565. [12] C. Wang, Y. Huang, Q. Zan, H. Guo, S. Cai, Mater. Sci. Eng., C 11 (2000) 9–12. [13] B.T. Lee, D.H. Jang, I.C. Kang, C.W. Lee, J. Am. Ceram. Soc. 88 (2005) 2874–2878. [14] B.T. Lee, I.C. Kang, S.H. Cho, H.Y. Song, J. Am. Ceram. Soc. 88 (2005) 2262–2266. [15] D. Kovar, B.H. King, R.W. Trice, J.W. Halloran, J. Am. Ceram. Soc. 80 (1997) 2471–2487. [16] B.T. Lee, K. Hiraga, J. Mater. Res. 9 (1994) 1199–1207. [17] B.T. Lee, K. Hiraga, D. Shindo, A. Nishiyama, J. Mater. Sci. 29 (1994) 959–964. [18] A.P. Quintin, M.H. Berger, A.R. Bunsell, C. Kaya, E.G. Butler, A. Woot￾ton, M. H. Lewis J. Eur. Ceram. Soc. 24 (2004) 101–110. [19] R.W. Trice, J.W. Halloran, J. Am. Ceram. Soc. 83 (2000) 311–316