MATERIALS HIENGE& ENGIEERING ELSEVIER Materials Science and Engineering A 487(2008)340-346 www.elsevier.com/locate/msea Plastic deformation of ductile ceramics in the Al2T1Os-MgTi2O5 system T. Shimazu,b,*,M. Miura, N Isua, T Ogawa K. Otad, H Maeda, E.H. Ishida b General Research Institute of Technology, INAX Corporation, Mi Tokoname, Aichi 479-8588, Japan b Graduate School of Environmental Studies, Tohoku Unive Sendai 980-8579, Japan nicho,Toki, 5202,Jpm Q INVERSE Incorporated, Mino, Osaka 562-0001, Japan Received 7 June 2007; received in revised form 5 October 2007; accepted 14 October 2007 Abstract Ceramics are typically rigid and brittle. If this brittleness could be overcome, their application can be widely expanded. Itacolumite is known a highly flexible unique rock in the sandstone group. The microstructure of itacolumite has many narrow gaps at the grain boundaries; these gaps low a slight displacement of the particles, and the rock can plastically bend when stressed. The authors tried to develop highly ductile ceramics by mimicking itacolumite, which resulted in the Al2T1Os-MgTi2Os solid solution ceramics. These ceramics have a high ductility and good stress relaxation properties against flexure stress because there are many thin gaps along the grain boundaries formed by discontinuous grain growth during the sintering process and anisotropic thermal expansion during the cooling process. The ductility is caused by the integration of a slight displacement of the grains. This phenomenon was observed as a high-intermal friction, reflecting the total friction of a high-grain boundary area C 2007 Elsevier B. V. All rights reserved. Keywords: Porous ceramics; Plastic deformation; Bending strain; Discontinuous grain growth; Microcrack; Internal friction; Grain boundary 1. Introduction but there is a mineral with a peculiar flexibility. The mineral is called itacolumite. Itacolumite is known as a highly ductile rock Ceramics are well known as useful materials in various engi- in the sandstone group, which consists of quartz and a slight neering fields because of their advantages such as stiffness, heat amount of muscovite [2]. The microstructure of itacolumite is resistance, abrasion resistance and corrosion resistance. They composed of large particles of over hundred micrometers and also have some disadvantages such as brittleness, low worka- interconnecting narrow pores of a bility(no ductility, low-cutting performance). These are their along the grain boundaries[3]. The groundwater-dissolved inter- original properties, but if these disadvantages could be over- stitial materials result in a narrow gap between the quartz grain ome, their application can be widely expanded. Many studies [2]. When an external force was applied to an itacolumite thin have tried to improve the ceramic brittleness, such as a method of plate, the integration of a slight displacement of the quartz increasing the fracture toughness by the addition of fine particles grains allows the plastic deformation [4]. These characteris of ZrO2 to the Al2O3 matrix that introduce many microcracks tics of itacolumite indicate the possibility of a new ceramic at the grain boundaries [1]. We tried to develop new ceramic function such as stress relaxation properties as a plastic microde- functions in order to spread their application. For example, if formation and high-damping capacity (internal friction). In other flexibility could be added to the ceramics, an increase in the types words, itacolumite can be called a natural functional ceramics of applications and processing will be expected. As we began the material development of this new function, we got a clue from a mineral We tried to develop ductile ceramics by introducing compl igidity and flexibility are competing properties in ceramics, shaped and interconnecting pores to the matrix thus mimicking he microstructure of itacolumite. In previous studies [5,6] thermal expansion during firing was applied and evaluated for Corresponding author at: General Research Institute of Technology, INAX introducing narrow pores into the matrix Ota et al. were able orporation, Minatomachi, Tokoname, Aichi 479-8588, Japan. to form many microcracks by using different compositions of Tel:+81569434866;fax:+81569436114 many elemental raw powder materials which have different ther E- mail address: mazu@ i2 inax co jp (T. Shimazu mal expansion coefficients and never mutually react. However, 0921-5093 2007 Elsevier B. V. All rights reserved. doi:10.101
Materials Science and Engineering A 487 (2008) 340–346 Plastic deformation of ductile ceramics in the Al2TiO5–MgTi2O5 system T. Shimazu a,b,∗, M. Miura a, N. Isu a, T. Ogawa c, K. Ota d, H. Maeda b, E.H. Ishida b a General Research Institute of Technology, INAX Corporation, Minatomachi, Tokoname, Aichi 479-8588, Japan b Graduate School of Environmental Studies, Tohoku University, Aoba, Sendai 980-8579, Japan c Acoh Ceramic Corporation, Oroshicho, Toki, Gifu 509-5202, Japan d Q INVERSE Incorporated, Mino, Osaka 562-0001, Japan Received 7 June 2007; received in revised form 5 October 2007; accepted 14 October 2007 Abstract Ceramics are typically rigid and brittle. If this brittleness could be overcome, their application can be widely expanded. Itacolumite is known as a highly flexible unique rock in the sandstone group. The microstructure of itacolumite has many narrow gaps at the grain boundaries; these gaps allow a slight displacement of the particles, and the rock can plastically bend when stressed. The authors tried to develop highly ductile ceramics by mimicking itacolumite, which resulted in the Al2TiO5–MgTi2O5 solid solution ceramics. These ceramics have a high ductility and good stress relaxation properties against flexure stress because there are many thin gaps along the grain boundaries formed by discontinuous grain growth during the sintering process and anisotropic thermal expansion during the cooling process. The ductility is caused by the integration of a slight displacement of the grains. This phenomenon was observed as a high-internal friction, reflecting the total friction of a high-grain boundary area. © 2007 Elsevier B.V. All rights reserved. Keywords: Porous ceramics; Plastic deformation; Bending strain; Discontinuous grain growth; Microcrack; Internal friction; Grain boundary 1. Introduction Ceramics are well known as useful materials in various engineering fields because of their advantages such as stiffness, heat resistance, abrasion resistance and corrosion resistance. They also have some disadvantages such as brittleness, low workability (no ductility, low-cutting performance). These are their original properties, but if these disadvantages could be overcome, their application can be widely expanded. Many studies have tried to improve the ceramic brittleness, such as a method of increasing the fracture toughness by the addition of fine particles of ZrO2 to the Al2O3 matrix that introduce many microcracks at the grain boundaries [1]. We tried to develop new ceramic functions in order to spread their application. For example, if flexibility could be added to the ceramics, an increase in the types of applications and processing will be expected. As we began the development of this new function, we got a clue from a mineral. Rigidity and flexibility are competing properties in ceramics, ∗ Corresponding author at: General Research Institute of Technology, INAX Corporation, Minatomachi, Tokoname, Aichi 479-8588, Japan. Tel.: +81 569 43 4866; fax: +81 569 43 6114. E-mail address: mazuu@i2.inax.co.jp (T. Shimazu). but there is a mineral with a peculiar flexibility. The mineral is called itacolumite. Itacolumite is known as a highly ductile rock in the sandstone group, which consists of quartz and a slight amount of muscovite [2]. The microstructure of itacolumite is composed of large particles of over hundred micrometers and interconnecting narrow pores of a dozen micrometers in width along the grain boundaries[3]. The groundwater-dissolved interstitial materials result in a narrow gap between the quartz grains [2]. When an external force was applied to an itacolumite thin plate, the integration of a slight displacement of the quartz grains allows the plastic deformation [4]. These characteristics of itacolumite indicate the possibility of a new ceramic function such as stress relaxation properties as a plastic microdeformation and high-damping capacity (internal friction). In other words, itacolumite can be called a natural functional ceramics material. We tried to develop ductile ceramics by introducing complexshaped and interconnecting pores to the matrix thus mimicking the microstructure of itacolumite. In previous studies [5,6], thermal expansion during firing was applied and evaluated for introducing narrow pores into the matrix. Ota et al. were able to form many microcracks by using different compositions of many elemental raw powder materials which have different thermal expansion coefficients and never mutually react. However, 0921-5093/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2007.10.026
T Shimazu et al. Materials Science and Engineering A 487(2008)340-346 this method required a strict raw material selection(reactivity, Table I thermal expansion and particle size). Also, it is preferable that Molar ratio of Al2 TiOs and Mg Ti2Os in fired bodies(molar ratio(%) the composition of the matrix is homogeneous. Thus we tried a to find a much simpler method. In this study, we focused series, which has many narrow pores along the grain boundaries Mg Ti20s 100 8775 especially the Al2TiO5-MgT12O5 The ductility and damping capacity of the Al2TiOs-MgTi20 ceramics and the relation between these properties and the were mixed with ethanol in a plastic ball mill for 20h.After microstructure were measured and investigated drying the slurry, the obtained powders were formed by die pressing a50 MPa to a size of80mm×12mm×5mm.The 2. Experimental formed bodies were then fired at 1500C for 2 h 2. 1. Synthesis of the Al21iO5-MgTi205 ceramics 2.2. Evaluation In this study, we focused on the phenomenon that the microc The fired specimens were characterized as follows: the crys- racks were formed in the ceramic matrix by anisotropic thermal talline phase was confirmed by powder X-ray diffraction.The expansion of the particles in order to introduce complex and bulk density was measured by the Archimedean method.The continuous micropores in the matrix. Aluminum titanate oxide porosity was calculated from the bulk density and the absolute (Al2TiO5 )is one of such famous materials. The crystal structure specific gravity of the powdered state was measured by the pyc- of Al2TiOs is the pseudobrookite type(orthorhombic system), nometer method, and the pore size distribution was measured and the thermal expansion coefficients (B)of each axis are quite by mercury intrusion porosimetry. The microstructure of the differences [7](Ba=-29 x 10-K, Bb=10.3 x 10-K fractured cross-section was observed by SEM and Bc=20.1 x 10-6K-). Therefore, many microcracks were The maximum bending strain (o) was used as an indicator formed due to reducing the thermal stress related to the thermal of the ceramics ductility in this study, and it was calculated expansion anisotropy on the surfaces of the elongated grains using the following equation with the fracture strength mea- during the cooling process [8,9.Thus, Al2TiOs is known as a surement data (E: maximum displacement(mm), I: geometrical low-thermal expansion ceramic [8,9 moment of inertia=(WT)12, S: span(mm), Z: modulus of We applied these properties to form continuous and section=(WT2)6, W: width(mm), T: thickness(mm) complex-shaped narrow pores in the ceramics matrix. In the Al2O3-TiO2-MgO system, Al2TiO5 forms solid solutions [10] a(%) 12EI with MgTi2 Os that has the same type of crystal structure [11] In this study, we prepared seven different solid solution compo- The fracture strength was measured by the three-point bend sitions(Fig. 1). The molar ratios of the solid solution after firing ing test(cross-head speed: 0.5 mm/min, span: 30 mm, the are shown in Table 1 specimen number: 5) The raw powders of Al2O3(AL-160SG-4: Showa Denko The internal friction was measured by the free resonance KK ) TiO2(KA-10C: Titan Kogyo Co Ltd )and Mgo (#500: method(Nippon Techno-plus Co Ltd. JE-RT) Fig. 2 shows Tateho Chemical Industries Co Ltd )in appropriate amounts a schematic view of the free resonance method. The specimen was placed on stainless wires( 50 um) at the node of vibra tion A non-contact electrostatic transmitter produces a vibration in the specimen and the resonance frequency was determined by shifting the oscillating frequency. Since the ceramic spec 80 30 specimen △ △ Fig 1 Composition of the AlzTiOs-MgTi2Os system(mol%). Fig. 2. Free resonance method, Youngs modulus and the internal friction
T. Shimazu et al. / Materials Science and Engineering A 487 (2008) 340–346 341 this method required a strict raw material selection (reactivity, thermal expansion and particle size). Also, it is preferable that the composition of the matrix is homogeneous. Thus we tried to find a much simpler method. In this study, we focused on Al2O3–TiO2–MgO ceramics, especially the Al2TiO5–MgTi2O5 series, which has many narrow pores along the grain boundaries. The ductility and damping capacity of the Al2TiO5–MgTi2O5 ceramics and the relation between these properties and the microstructure were measured and investigated. 2. Experimental 2.1. Synthesis of the Al2TiO5–MgTi2O5 ceramics In this study, we focused on the phenomenon that the microcracks were formed in the ceramic matrix by anisotropic thermal expansion of the particles in order to introduce complex and continuous micropores in the matrix. Aluminum titanate oxide (Al2TiO5) is one of such famous materials. The crystal structure of Al2TiO5 is the pseudobrookite type (orthorhombic system), and the thermal expansion coefficients (β) of each axis are quite differences [7] (βa = −2.9 × 10−6 K−1, βb = 10.3 × 10−6 K−1 and βc = 20.1 × 10−6 K−1). Therefore, many microcracks were formed due to reducing the thermal stress related to the thermal expansion anisotropy on the surfaces of the elongated grains during the cooling process [8,9]. Thus, Al2TiO5 is known as a low-thermal expansion ceramic [8,9]. We applied these properties to form continuous and complex-shaped narrow pores in the ceramics matrix. In the Al2O3–TiO2–MgO system, Al2TiO5 forms solid solutions [10] with MgTi2O5 that has the same type of crystal structure [11]. In this study, we prepared seven different solid solution compositions (Fig. 1). The molar ratios of the solid solution after firing are shown in Table 1. The raw powders of Al2O3 (AL-160SG-4: Showa Denko K.K.), TiO2 (KA-10C: Titan Kogyo Co. Ltd.) and MgO (#500: Tateho Chemical Industries Co. Ltd.) in appropriate amounts Fig. 1. Composition of the Al2TiO5–MgTi2O5 system (mol%). Table 1 Molar ratio of Al2TiO5 and MgTi2O5 in fired bodies (molar ratio (%)) a bcdef g Al2TiO5 100 87 75 70 65 50 0 MgTi2O5 0 13 25 30 35 50 100 were mixed with ethanol in a plastic ball mill for 20 h. After drying the slurry, the obtained powders were formed by diepressing at 50 MPa to a size of 80 mm × 12 mm × 5 mm. The formed bodies were then fired at 1500 ◦C for 2 h. 2.2. Evaluation The fired specimens were characterized as follows: the crystalline phase was confirmed by powder X-ray diffraction. The bulk density was measured by the Archimedean method. The porosity was calculated from the bulk density and the absolute specific gravity of the powdered state was measured by the pycnometer method, and the pore size distribution was measured by mercury intrusion porosimetry. The microstructure of the fractured cross-section was observed by SEM. The maximum bending strain (σ) was used as an indicator of the ceramics ductility in this study, and it was calculated using the following equation with the fracture strength measurement data (E: maximum displacement (mm), I: geometrical moment of inertia = (WT3)/12, S: span (mm), Z: modulus of section = (WT2)/6, W: width (mm), T: thickness (mm)). σ (%) = 12EI S2Z × 100 The fracture strength was measured by the three-point bending test (cross-head speed: 0.5 mm/min, span: 30 mm, the specimen number: 5). The internal friction was measured by the free resonance method (Nippon Techno-plus Co. Ltd.: JE-RT). Fig. 2 shows a schematic view of the free resonance method. The specimen was placed on stainless wires (φ 50m) at the node of vibration. A non-contact electrostatic transmitter produces a vibration in the specimen and the resonance frequency was determined by shifting the oscillating frequency. Since the ceramic specFig. 2. Free resonance method, Young’s modulus and the internal friction.
T Shimazu et al. Materials Science and Engineering A 487(2008)340-346 10 Fig 4. Lattice constant of c-axis in the Al2TiOs-MgTi2Os system. 3. Results and discussion 3.1. Characterization of porous ceramics 20O) Fig3. X-ray diffraction pattens of the AlzTiOs-M The X-ray diffraction patterns of the fired body of each com- (b) Mgo. 13Al1.74Ti1130s, (c)Mgo.25Al1sTi12s 3AlL..30s. (e) position are shown in Fig. 3. The X-ray diffraction peaks from Mgo.35Al13T1135Os, (f) Mgo. AITi1sOs and (g)Mg Ti2Os (7) The peaks of Al2 TiOs were observed only on the Al end-member (a), and Al2 TiOs and(A)the peaks of MgTi20. that of Mg Ti2Os was observed only in the Mg end-member(g) The samples(b)(f have peaks related to the Al2 T1O5-MgT12O5 imen does not have an electrical conductivity, the surface of solid solution, and the peaks were shifted from the Al2 TiOs posi the specimen was coated with thin carbon paint that provided tion to the lower angle. The diffraction angle decreases with the the electrical conductivity. The internal friction(@-)was cal- increasing MgO composition. Fig 4 shows the lattice constant culated by the half-width method using the following equation of the c-axis; these composites were calculated by the computer fr: resonance frequency, fi and f2: frequency at of maximum program UNIT CELL[12]. The lattice constant of c-axis propor- vibration amplitude ): tionally increased with the MgTi2 Os rate and each composition osition as indicated in sect 了-1-0 The microstructures are shown in Fig. 5. The Al end-member f (a)is composed of isotropic grains of approximately 2-5 um 10 um 10 um Fig. 5. Cross-sectional view of microstructure of the Al2TiOs-MgTi2Os system. (a) Al2TiOs,( b) Mgo.13Al174Ti113Os,(c)Mgo.25 Al1sTi1.2sOs,(d) Mgo3All. 4Ti13Os,(e) Mgo. Al1.3Ti135Os and (g) MgTi20
342 T. Shimazu et al. / Materials Science and Engineering A 487 (2008) 340–346 Fig. 3. X-ray diffraction patterns of the Al2TiO5–MgTi2O5 system. (a) Al2TiO5, (b) Mg0.13Al1.74Ti1.13O5, (c) Mg0.25Al1.5Ti1.25O5, (d) Mg0.3Al1.4Ti1.3O5, (e) Mg0.35Al1.3Ti1.35O5, (f) Mg0.5AlTi1.5O5 and (g) MgTi2O5. () The peaks of Al2TiO5 and () the peaks of MgTi2O5. imen does not have an electrical conductivity, the surface of the specimen was coated with thin carbon paint that provided the electrical conductivity. The internal friction (Q−1) was calculated by the half-width method using the following equation (fr: resonance frequency, f1 and f2: frequency at of maximum vibration amplitude): Q−1 = f2 − f1 fr Fig. 4. Lattice constant of c-axis in the Al2TiO5–MgTi2O5 system. 3. Results and discussion 3.1. Characterization of porous ceramics The X-ray diffraction patterns of the fired body of each composition are shown in Fig. 3. The X-ray diffraction peaks from Al2TiO5 were observed only on the Al end-member (a), and that of MgTi2O5 was observed only in the Mg end-member (g). The samples (b)–(f) have peaks related to the Al2TiO5–MgTi2O5 solid solution, and the peaks were shifted from the Al2TiO5 position to the lower angle. The diffraction angle decreases with the increasing MgO composition. Fig. 4 shows the lattice constant of the c-axis; these composites were calculated by the computer program UNIT CELL [12]. The lattice constant of c-axis proportionally increased with the MgTi2O5 rate and each composition was the assumed composition as indicated in Section 2. The microstructures are shown in Fig. 5. The Al end-member (a) is composed of isotropic grains of approximately 2–5 m, Fig. 5. Cross-sectional view of microstructure of the Al2TiO5–MgTi2O5 system. (a) Al2TiO5, (b) Mg0.13Al1.74Ti1.13O5, (c) Mg0.25Al1.5Ti1.25O5, (d) Mg0.3Al1.4Ti1.3O5, (e) Mg0.35Al1.3Ti1.35O5 and (g) MgTi2O5.
T. Shimazu et al. Materials Science and Engineering A 487(2008)340-346 .30(a) 007 0.10 MgTi? Os molar ratio(%) Fig. 6. Firing shrinkage vs composition in the AlzTiOs-MgTi2Os system 0.0 and many large cracks with a 10 um width are observed. The others have distinct microstructures, which consist of thick elon OLL gated columnar grains and narrow pores of several micrometers in width along the grain boundaries. The large grains sizes are 20-30 um in diameter and from 20 to over 100 umin length. The average firing shrinkage and porosity are shown in Figs. 6 and 7. The firing shrinkage of the Al end-member(a) is much lower 0.02 than the others. The shrinkage of the others slightly increased with the amount of Mg Ti2O5. The porosities of (bg)were not very different at around 10%0, while(a)was much greater than the others. No remarkable mutual relation was found between the porosity and the composition of the Al2T1O5-MgTi2Os system. NM According to Ohya et al. [13], the addition of the MgO com- ponent to the Al2TiOs promotes sintering, so the sintered state 0.10「mm千m and microstructure of(b)g)that included the MgO component were different from the Al end-member(a) Furthermore, according to Daimon [14], in the Al] T1O5- MgTi2O5 system, an increase in the MgTi2O5 rate decreased with the temperature of the solid solution. In this study, a dis- continuous grain growth was observed in the composition with Fig. 8. Pore size distributions in the Al2TiOs-MgTi2Os system(a) Al2TiOs (b)Mgo. 13Al1.74Ti113O5, (c) Mgo.25Al15T1125 Os, (d) Mgo.3Al14T113Os,(e) Mgo.35Al1.3T113sOs, (f) Mgo.sAITi1sOs and (g) MgTi2Os very large grains in length. The aspect ratio of the grains increases with the increasing MgTi2Os amount. The complex-shaped narrow pores along the grain boundaries are supposed to be formed during cooling by the anisotropic dif- ference in the expansion coefficient of the grains as in the case of the Al2TiO5. On the surfaces of the elongated pores of several micrometers in diameter were observed. It is postulated that these were closed pores formed during firing and ppeared on the grain surface because the microcracks formed The pore size distributions of the solid solutions are shown in Fig. 8. Each sample has nearly the same size pores, and the MgTi,Os molar ratio(%) pores size is around 0.5 um. These pore sizes correspond to the narrow pores in the composition, and the distribution is narrow Fig. 7. Porosity vs composition in the Al2TiOs-MgTi2Os system in the order of (b)g). These results indicate that the pores of the
T. Shimazu et al. / Materials Science and Engineering A 487 (2008) 340–346 343 Fig. 6. Firing shrinkage vs. composition in the Al2TiO5–MgTi2O5 system. and many large cracks with a 10 m width are observed. The others have distinct microstructures, which consist of thick elongated columnar grains and narrow pores of several micrometers in width along the grain boundaries. The large grains sizes are 20–30m in diameter and from 20 to over 100m in length. The average firing shrinkage and porosity are shown in Figs. 6 and 7. The firing shrinkage of the Al end-member (a) is much lower than the others. The shrinkage of the others slightly increased with the amount of MgTi2O5. The porosities of (b)–(g) were not very different at around 10%, while (a) was much greater than the others. No remarkable mutual relation was found between the porosity and the composition of the Al2TiO5–MgTi2O5 system. According to Ohya et al. [13], the addition of the MgO component to the Al2TiO5 promotes sintering, so the sintered state and microstructure of (b)–(g) that included the MgO component were different from the Al end-member (a). Furthermore, according to Daimon [14], in the Al2TiO5– MgTi2O5 system, an increase in the MgTi2O5 rate decreased with the temperature of the solid solution. In this study, a discontinuous grain growth was observed in the composition with Fig. 7. Porosity vs. composition in the Al2TiO5–MgTi2O5 system. Fig. 8. Pore size distributions in the Al2TiO5–MgTi2O5 system. (a) Al2TiO5, (b) Mg0.13Al1.74Ti1.13O5, (c) Mg0.25Al1.5Ti1.25O5, (d) Mg0.3Al1.4Ti1.3O5, (e) Mg0.35Al1.3Ti1.35O5, (f) Mg0.5AlTi1.5O5 and (g) MgTi2O5. MgO component. Especially, the solid solutions (b)–(e) have very large grains of over 100 m in length. The aspect ratio of the grains increases with the increasing MgTi2O5 amount. The complex-shaped narrow pores along the grain boundaries are supposed to be formed during cooling by the anisotropic difference in the expansion coefficient of the grains as in the case of the Al2TiO5. On the surfaces of the elongated grains, many pores of several micrometers in diameter were observed. It is postulated that these were closed pores formed during firing and appeared on the grain surface because the microcracks formed. The pore size distributions of the solid solutions are shown in Fig. 8. Each sample has nearly the same size pores, and the pores size is around 0.5 m. These pore sizes correspond to the narrow pores in the composition, and the distribution is narrow in the order of (b)–(g). These results indicate that the pores of the
T Shimazu et al. Materials Science and Engineering A 487(2008)340-346 20 80100 MgTi,2Os molar ratio(%) displacement(mm) Fig. 10. Maximum bending strain vs composition in the Al2TiOs-MgTi20s Fig9. Stress-displacement curves in the Al2TiOs-MgTi2Os system. (O) Sy Al2TOs,(×)Mg03Al1.4Ti1 Os and(▲)MgTi2O5 pulled out by the stress, but the large grains prevented fracture solid solutions were mainly composed of microcracks formed of the specimens after being moved by rubbing each other. by the anisotropic thermal expansion of the grains. The results Figs. 5 and 8 shows that(b)g) are remarkably constant 3.3. Evaluation of intemal friction of the porosity, pore size and shape with the composition and the gra A2TiOs-MgTi2O5 system The behavior of the stress-strain curve of the Al TiO5- 3.2. Evaluation of ductility of the Al2TiO5-Mgli2Os system Mg Ti2O5 system was different from the common brittle ceram- The ductility of the ceramics was calculated from the ics and showed ductility like plastic deformation. Their ductility was supposed to contribute by integration of the slight displace stress-strain curve measured by the three-point bending test. ment of the grains depending on the grain boundary slip.Thus Fig9 shows the stress-strain curves. The Al21105-Mgli2 05 moving the grains causes many contact points with the neighbo system has a higher work of fracture than the usual brittle ing grains and mechanical friction would occur on each contact After reaching the maximum flexural strength, the surface. We expected that these phenomena would be picked up stress slowly decreased. Especially, on the solid solution (d), as an internal energy loss factor (internal friction) The internal friction of the Al2TiO5-MgTi2O5 system was tained the maximum strain and did not fracture after the test. measured by the free resonance method. These results are shown These results indicate that the solid solutions have a stress relax in Fig. 13. The internal friction changes versus the mgTi?o ation mechanism Fig. 10 shows the maximum bending strain The strains of all the specimens were large as visually confirmed, and the solid solutions(b)(e)showed very high- The large distortion was supposed to be caused by integration of the slight movement of the grains as in the case of itacolumite These unique properties are derived from the many narrow pores formed along the grain boundaries. These pores decrease 6 the bonding strength of the grain boundaries and gave the grains interspaces to allow movement, and caused a very low flexural trength of less than 10 MPa. The flexural strength of the system is shown in Fig. 11. The solid solutions(che) have the lowest flexural strength, which correspond with the high-bending strain 2 Fig. 12 shows the microstructure of the fracture surface of the Al2T1O5-MgTi2Os system Solid solution(d), which shows a very high-bending strain, and many elongated columnar grains due to steric crowding at the surface as if being pulled out, can be observed. The grains of solid solutions(b)-(e) were easily Fig. 11. Flexural strength vs composition in the Al2 TiOs-MgTi2Os system
344 T. Shimazu et al. / Materials Science and Engineering A 487 (2008) 340–346 Fig. 9. Stress–displacement curves in the Al2TiO5–MgTi2O5 system. () Al2TiO5, (×) Mg0.3Al1.4Ti1.3O5 and () MgTi2O5. solid solutions were mainly composed of microcracks formed by the anisotropic thermal expansion of the grains. The results in Figs. 5 and 8 shows that (b)–(g) are remarkably constant in porosity, pore size and shape with the composition and the grain growth. 3.2. Evaluation of ductility of the Al2TiO5–MgTi2O5 system The ductility of the ceramics was calculated from the stress–strain curve measured by the three-point bending test. Fig. 9 shows the stress–strain curves. The Al2TiO5–MgTi2O5 system has a higher work of fracture than the usual brittle ceramic. After reaching the maximum flexural strength, the stress slowly decreased. Especially, on the solid solution (d), the stress was very gradually changed, and the specimens maintained the maximum strain and did not fracture after the test. These results indicate that the solid solutions have a stress relaxation mechanism. Fig. 10 shows the maximum bending strain. The strains of all the specimens were large as visually confirmed, and the solid solutions (b)–(e) showed very high-plastic strains among them. The large distortion was supposed to be caused by integration of the slight movement of the grains as in the case of itacolumite. These unique properties are derived from the many narrow pores formed along the grain boundaries. These pores decrease the bonding strength of the grain boundaries and gave the grains interspaces to allow movement, and caused a very low flexural strength of less than 10 MPa. The flexural strength of the system is shown in Fig. 11. The solid solutions (c)–(e) have the lowest flexural strength, which correspond with the high-bending strain area. Fig. 12 shows the microstructure of the fracture surface of the Al2TiO5–MgTi2O5 system. Solid solution (d), which shows a very high-bending strain, and many elongated columnar grains due to steric crowding at the surface as if being pulled out, can be observed. The grains of solid solutions (b)–(e) were easily Fig. 10. Maximum bending strain vs. composition in the Al2TiO5–MgTi2O5 system. pulled out by the stress, but the large grains prevented fracture of the specimens after being moved by rubbing each other. 3.3. Evaluation of internal friction of the Al2TiO5–MgTi2O5 system The behavior of the stress–strain curve of the Al2TiO5– MgTi2O5 system was different from the common brittle ceramics and showed ductility like plastic deformation. Their ductility was supposed to contribute by integration of the slight displacement of the grains depending on the grain boundary slip. Thus, moving the grains causes many contact points with the neighboring grains and mechanical friction would occur on each contact surface. We expected that these phenomena would be picked up as an internal energy loss factor (internal friction). The internal friction of the Al2TiO5–MgTi2O5 system was measured by the free resonance method. These results are shown in Fig. 13. The internal friction changes versus the MgTi2O5 Fig. 11. Flexural strength vs. composition in the Al2TiO5–MgTi2O5 system.
T. Shimazu et al. Materials Science and Engineering A 487(2008)340-346 (a) 80100 Fig. 13. Internal friction vs. composition in the Al2 TiOs-MgTi2Os system. internal friction was higher for compositions(bH(e), and this almost corresponded to the higher maximum bending strain hown in Fig. 10. These results show the deformation behav ior associated with the internal friction, and indicate that a large distortion of the specimens progressed with the friction at many The Al end component(a) shows a high-internal friction but the mechanism of generating the internal friction is not the same as the others. As previously mentioned, the porosity and the microstructure of (a) is quite different from the oth- ers On the other hand, the Mg members(b)(g) do not have any remarkable difference in porosity, pore size and shape. The microstructure must be a very important factor regarding the deformation behavior and internal friction. The analysis of the g) relation between the composition of the Al2T1O5-MgTi2Os sys- tem and the microstructure was not sufficient. Development of more plastic ceramics can be expected by optimizing the microstructure, such as grain size, aspect ratio and bonding conditions of the grain boundaries In the Al2TiO5-MgTi2O5 system, a complex-shaped microstructure was formed during firing by introducing many narrow pores along the large elongated columnar grains due to discontinuous grain growth and the anisotropic thermal expaN sion coefficient of the grain axis The Al2T1O5-MgTi2O5 system showed a very large distor tion like plasticity with flexure stress due to the weak of bonding Fig. 12. Fracture cross-sectional surface of the Al2TiOs-MgTi20s system(a) of the grain boundaries resulting from many microcracks. A Al2TiOs, (d) Mgo.3Al1 4Ti13Os and(g) MgTi2O large distortion occurs due to the slight sliding of the grains. This phenomenon content.The internal friction of all the specimens was greater the total friction of the grain surf than 1.0 x 10-and the solid solution(c)had the highest value of 3.7x 10-, which is much higher than the other common Acknowledgements ceramics [15] such as alumina(porosity 30%0, internal fric- tion 2.0x 10-)and mullite(porosity 25%, internal friction This work was supported by Ministry of Economy, Trade and 1.5 x 10), and on the same order as the polymer grade. The Industry for the Regional Consortium Research Development
T. Shimazu et al. / Materials Science and Engineering A 487 (2008) 340–346 345 Fig. 12. Fracture cross-sectional surface of the Al2TiO5–MgTi2O5 system. (a) Al2TiO5, (d) Mg0.3Al1.4Ti1.3O5 and (g) MgTi2O5. content. The internal friction of all the specimens was greater than 1.0 × 10−2 and the solid solution (c) had the highest value of 3.7 × 10−2, which is much higher than the other common ceramics [15] such as alumina (porosity 30%, internal friction 2.0 × 10−4) and mullite (porosity 25%, internal friction 1.5 × 10−3), and on the same order as the polymer grade. The Fig. 13. Internal friction vs. composition in the Al2TiO5–MgTi2O5 system. internal friction was higher for compositions (b)–(e), and this almost corresponded to the higher maximum bending strain shown in Fig. 10. These results show the deformation behavior associated with the internal friction, and indicate that a large distortion of the specimens progressed with the friction at many points of the grain boundaries. The Al end component (a) shows a high-internal friction, but the mechanism of generating the internal friction is not the same as the others. As previously mentioned, the porosity and the microstructure of (a) is quite different from the others. On the other hand, the Mg members (b)–(g) do not have any remarkable difference in porosity, pore size and shape. The microstructure must be a very important factor regarding the deformation behavior and internal friction. The analysis of the relation between the composition of the Al2TiO5–MgTi2O5 system and the microstructure was not sufficient. Development of more plastic ceramics can be expected by optimizing the microstructure, such as grain size, aspect ratio and bonding conditions of the grain boundaries. 4. Conclusions In the Al2TiO5–MgTi2O5 system, a complex-shaped microstructure was formed during firing by introducing many narrow pores along the large elongated columnar grains due to discontinuous grain growth and the anisotropic thermal expansion coefficient of the grain axis. The Al2TiO5–MgTi2O5 system showed a very large distortion like plasticity with flexure stress due to the weak of bonding of the grain boundaries resulting from many microcracks. A large distortion occurs due to the slight sliding of the grains. This phenomenon was observed as a high-internal friction reflecting the total friction of the grain surface. Acknowledgements This work was supported by Ministry of Economy, Trade and Industry for the Regional Consortium Research Development
T Shimazu et al. Materials Science and Engineering A 487(2008)340-346 Work 2003-2005, and supported by the grant-in-aid for Scien- [6] TOta, K Tamaki, N. Adachi, 1. Sato, Jpn. Ceram Soc(2005)313, Fall tific Research for Japan Society for the Promotion of Science meeting abstract. (no.18201014). [7 G. Bayer, J Less-Common Met. 24(1971)129-138. [8 W.R. Buessem, N.R. Thielke, R V. Sarakauskas, Ceram. Age 60(1952) 38-40. References [9]K Hamano, J. Technol. Assoc. Ref. Jpn. 27(1975)520-527 [10] A.S. Berezhnoi, N V Gul'ko, Ukrain. Khim. Zhur. 21(2)(1955)158- [1 N. Claussen, J Steeb, R F Pabst, Am. Ceram Soc. Bull. 56(1977)559-562 [2] H. Suzuki, D Shimizu, J. Geo. Soc. Jpn. 5(1993)391-401. [11] J.J. Cleveland, R C. Br 61(1978)478-481 [3] T Shimazu, M. Miura, N. Isu, T Ogawa, A Ichikawa, E.H. Ishida, Pro- [12] H. Toraya. J Appl Cryst. ceedings of Third Water Dynamics, Sendai, Japan, 2006, pp 69-71 [13] Y. Ohya, K. Hamano, Z. yokai.-Sh94(1986)665- [4]K. Yamaguchi, Y Matsufuji, T. Koyama, Proceedings of the Second Inter national Workshop on Sustainable Habitat Systems, Fukuoka, Japan, 2005, [14 K Daimon, J Ceram Soc. Jpn. 98(1990)365-369. [15] T. Shimazu, M. Miura, H. Kuno, N Isu, K Ota, E.H. Ishida, Key Eng [5] M. Mizutani, J Sakanoue, Y Ichikawa, T Ota, K. Daimon, Y Hikichi, Jpn. Mater.319(2006)173-180 Ceram Soc (2003)238. Annual meeting abstract
346 T. Shimazu et al. / Materials Science and Engineering A 487 (2008) 340–346 Work 2003–2005, and supported by the grant-in-aid for Scientific Research for Japan Society for the Promotion of Science (no. 18201014). References [1] N. Claussen, J. Steeb, R.F. Pabst, Am. Ceram. Soc. Bull. 56 (1977) 559–562. [2] H. Suzuki, D. Shimizu, J. Geo. Soc. Jpn. 5 (1993) 391–401. [3] T. Shimazu, M. Miura, N. Isu, T. Ogawa, A. Ichikawa, E.H. Ishida, Proceedings of Third Water Dynamics, Sendai, Japan, 2006, pp. 69–71. [4] K. Yamaguchi, Y. Matsufuji, T. Koyama, Proceedings of the Second International Workshop on Sustainable Habitat Systems, Fukuoka, Japan, 2005, pp. 87–96. [5] M. Mizutani, J. Sakanoue, Y. Ichikawa, T. Ota, K. Daimon, Y. Hikichi, Jpn. Ceram. Soc (2003) 238, Annual meeting abstract. [6] T. Ota, K. Tamaki, N. Adachi, I. Sato, Jpn. Ceram. Soc (2005) 313, Fall meeting abstract. [7] G. Bayer, J. Less-Common Met. 24 (1971) 129–138. [8] W.R. Buessem, N.R. Thielke, R.V. Sarakauskas, Ceram. Age 60 (1952) 38–40. [9] K. Hamano, J. Technol. Assoc. Ref. Jpn. 27 (1975) 520–527. [10] A.S. Berezhnoi, N.V. Gul’ko, Ukrain. Khim. Zhur. 21 (2) (1955) 158– 166. [11] J.J. Cleveland, R.C. Bradt, J. Am. Ceram. Soc. 61 (1978) 478–481. [12] H. Toraya, J. Appl. Cryst. 26 (1993) 583–590. [13] Y. Ohya, K. Hamano, Z. Nakagawa, Yogyo-Kyokai-Shi 94 (1986) 665– 670. [14] K. Daimon, J. Ceram. Soc. Jpn. 98 (1990) 365–369. [15] T. Shimazu, M. Miura, H. Kuno, N. Isu, K. Ota, E.H. Ishida, Key Eng. Mater. 319 (2006) 173–180