- Email: [email protected]
Preparation of nacre-like composites by reactive infiltration of a magnesium alloy into porous silicon carbide derived from ice template
Materials Letters 183 (2016) 299–302
Contents lists available at ScienceDirect
Materials Letters journal homepage: www.elsevier.com/locate/matlet
Preparation of nacre-like composites by reactive infiltration of a magnesium alloy into porous silicon carbide derived from ice template Heng Zhang, Ping Shen n, Alateng Shaga, Ruifen Guo, Qichuan Jiang Key Laboratory of Automobile Materials (Ministry of Education), Department of Materials Science and Engineering, Jilin University, No. 5988 Renmin Street, Changchun, 130025 PR China
art ic l e i nf o
a b s t r a c t
Article history: Received 11 April 2016 Received in revised form 18 July 2016 Accepted 26 July 2016 Available online 27 July 2016
Lightweight and high-strength AZ91/SiC composites with lamellar structures were successfully prepared by using freeze casting (ice-templating) and reactive infiltration techniques. The infiltration dynamics was measured and activation energy calculated to be 7.74 kJ/mol. The compressive and flexural strengths of the composites with 30 vol% initial solid load reached 7437 20 MPa and 599 744 MPa, about 1.8 and 1.5 times of those of the AZ91 alloy, respectively. & 2016 Elsevier B.V. All rights reserved.
Keywords: Metal matrix composites (MMC) Multilayer structure Biomimetic Liquid infiltration Mechanical properties
1. Introduction The distinct “brick-and-mortar” structure provides nacre with high strength and toughness [1]. Inspired by nacre's fantastic structure and fabulous property, novel techniques are being developed to fabricate nacre-like materials [2–4]. Recently, Deville et al. [2] have adopted freezing as a flexible path to build sophisticated porous and layer-hybrid materials. Following this technique, Launey et al. [3] prepared the composites by infiltrating an Al– 12.6 wt% Si alloy into a 36 vol% porous Al2O3 scaffold with lamellar structures at 1173 K using gas-pressure infiltration, which exhibited fracture (crack-growth) toughness of 40 MPa m1/2 and bending strength of 300 MPa. Up to date, the preparation of the nacre-like metal–ceramic composites is limited to the Al matrix and most of the studies employed high-cost pressure infiltration technique. For the Mg matrix, to our knowledge, there is no report. As known, Mg has lower density and higher specific strength and modulus than Al, receiving wide attention in recent years for potential structural applications. However, Mg has strong chemical reactivity and high vapor pressure and thus is easy to ignite and evaporate at elevated temperatures, making the preparation of Mg-matrix composites difficult. Nevertheless, our previous study revealed that Mg has very good wettability with silica [5], favoring the spontaneous n
Corresponding author. E-mail address: [email protected] (P. Shen).
http://dx.doi.org/10.1016/j.matlet.2016.07.126 0167-577X/& 2016 Elsevier B.V. All rights reserved.
infiltration into silica-coated ceramic scaffolds. Based on this knowledge, in this work we prepared for the first time the lamellar Mg/SiC composites using freeze casting and reactive infiltration techniques.
2. Materials and methods The raw ceramic materials used in this work were commercially available SiC powders with an average particle size (d50) of 5 mm and a purity of 98.5 wt% together with 16.5 wt% Al2O3, 1.5 wt% MgO and 2 wt% Y2O3. They were added as sintering aids. Besides, 0.6 wt% CMC-Na was used as dispersant and deionized water as solvent. First, water-based SiC slurry involving sintering aids with a total solid load of 30 vol% was ball-milled and then de-aired. Subsequently, the slurry was poured into Teflon molds to shape into cylinders with 18 mm in diameter and 25 mm in height by unidirectional freezing at –20 °C. After demolding, the cold SiC bodies were freeze-dried at –50 °C in a 10 Pa vacuum for 24 h to remove the ice inside [6]. The resultant SiC bodies were then fired at 1100 °C in air for 2 h to form an oxide layer (silica) at the SiC surface and further sintered at 1500 °C in Ar atmosphere for 2 h. During the firing process, the heating and cooling rates were 5 °C/ min. In order to determine optimum infiltration parameters, in-situ observation of spontaneous infiltration for a commercial AZ91
300
H. Zhang et al. / Materials Letters 183 (2016) 299–302
(Mg–9Al–1Zn) alloy into porous SiC scaffolds was carried out. The AZ91 alloy was pre-placed on the SiC scaffold and the couple was heated in a high-purity (99.999%) Ar gas at 5 °C/min in a tube furnace, which was initially evacuated to a vacuum less than 5 Pa. During the heating and isothermal dwelling (at the temperature between 610 and 690 °C with an interval of 20 °C) process, the shape of the AZ91 alloy was monitored by a video camera through quartz windows to record its morphological change. The spontaneous infiltration dynamics was characterized by time variations in the alloy height and contact diameter. On the basis of the in-situ observation results, the infiltration of the AZ91 alloy into the SiC scaffolds was performed at 650 °C with a holding time of 5 min. Fully infiltrated samples were readily achieved. The microstructures of the composites were observed using an optical microscope (Axio Imager A2m, Carl Zeiss, Germany) and a scanning electron microscope (SEM, Evo18, Carl Zeiss, Germany), and phases were identified by X-ray diffraction (XRD, D/Max 2500PC Rigaku, Japan). The compressive and three-point flexural strengths of the composites were measured by a universal material testing machine (Instron 5689 Corp., USA).
3. Results and discussion Fig. 1 shows the microstructures of the SiC scaffold prepared by freeze casting. The sample displayed a well-developed lamellar structure in both longitudinal and transverse sections. Because SiC has a strong covalent bond and the atoms have low diffusion coefficients, it is difficult to achieve densification even sintering at a temperature of 2000 °C [7]. However, in the present work, the SiC surface was preoxidized at 1100 °C for 2 h, forming a SiO2 layer, which may further react with the sintering aids during subsequent high-temperature (1500 °C) firing and thus greatly promote sinterability and enhance the strength of the scaffold. Nevertheless, many cavities were still observed in the SiC lamellae; namely, the ceramic layers were still far from densification. By way of the preoxidation treatment, the spontaneous infiltration was readily achieved and the entire process can be divided into three stages (Fig. 2(a)): (i) alloy melting stage, (ii) inoculation stage and (iii) rapid infiltration stage. Because of the coverage of an oxide film at the Mg alloy surface, no obvious infiltration was observed in stages (i) and (ii). The decrease in the drop height and increase in the contact diameter in the end of stage (i) were virtually due to the collapse of the alloy after melting. During the inoculation stage (stage ii), the height of the
alloy and the contact diameter did not change. However, with the progressive evaporation of Mg, the oxide film was finally disrupted at some locations, thus paving the way for the spontaneous infiltration. The duration of the inoculation stage depends on temperature (Fig. 2(b)). Clearly, a higher temperature favors a larger Mg vapor pressure and thus faster disruption of the oxide film. As the clean liquid Mg alloy contacted the pre-oxidized SiC surface, rapid infiltration occurred, which is characterized by the sharp decrease in the height of the alloy drop together with a certain extent of liquid spreading (i.e., increase in the contact diameter as shown in Fig. 2(a). The infiltration rate (v) can be roughly evaluated from the time dependence of the drop height (dH/dt). Note that the infiltration rate reached 10.2 mm/min at 650 °C and 14.3 mm/min at 690 °C (Fig. 2(c)); thus the whole sample with a height of 25 mm could be fully infiltrated by the alloy in no more than 3 min providing that the amount of the alloy was sufficient. Furthermore, according to the Arrhenius equation:
lnv = − Ea/(RT ) + lnA,
(1)
where Ea is the infiltration activation energy, A is a constant, R is the gas constant, and T is the absolute temperature, we calculated Ea ¼7.74 kJ/mol. This small value suggests easy activation for the infiltration of the alloy into the porous SiC scaffolds. Fig. 3 shows the macro- and micro-structures of the infiltrating sample, particularly at the liquid infiltration front area. The sample could be characterized by three regions: (i) a fully infiltrated region, (ii) a partially infiltrated region and (iii) a non-infiltrated region. In the fully infiltrated region, the alloy not only distributed in the open channels but also in the cavities of the ceramic layers; whereas, due to the limited amount of liquid in the infiltration front area, the alloy was found only in the small cavities of the ceramic layers but not in the large open channels (Fig. 3(b)-2). The elemental mapping graphs (Fig. 3(c)) indicate that Mg had a selective distribution in the front area, whose position largely coincided with that of Si in the SiC layer, suggesting that the liquid alloy preferentially penetrated into the small cavities in the ceramic layer and then into the large open channels between the ceramic layers under the capillary force. This phenomenon can be explained by the Laplace equation:
P = 2σ cos θ /r ,
(2)
where P is the capillary force, s is the surface tension of the melt, r is the capillary radius and θ is the contact angle. In the case of θ { 90° for Mg on SiO2 [5], the smaller the capillary radius, the larger the capillary force, and as a consequence, the small cavities
Fig. 1. SEM micrographs showing the lamellar structure of the SiC scaffold: (a) perpendicular to the freezing direction; (b) parallel to the freezing direction.
H. Zhang et al. / Materials Letters 183 (2016) 299–302
301
Fig. 2. (a) Variations in the alloy height and contact diameter with time during heating to and dwelling at 650 °C. Due to the coverage of an oxide film, the alloy was in an irregular shape after melting, and thus the height was characterized by the average of the highest and lowest points, as schematically shown in the inserted image. (b) Variation in the alloy height with time during isothermal dwelling at different temperatures; (c) relationship between infiltration rate (v and lnv) and temperature (T and 1/T).
Fig. 3. (a) Macrograph of the infiltrating sample for dynamic analysis after holding at 650 °C; (b) optical micrographs showing the microstructures at partially infiltrated region; (c) SEM microstructure and elemental distributions at the liquid infiltration front.
in the ceramic layer were preferentially filled by the liquid alloy. The larger open pores would be certainly filled by the molten alloy as well providing that the amount of the alloy was sufficient. Besides, the rapid infiltration could be associated with the heat released from the reaction between Mg and SiO2 or silicate phase, which increased temperature and improved fluidity of the alloy at the infiltration front. As a consequence, the infiltration was
accelerated. Fig. 4(a) gives the optical micrograph of the composite structure. The bright layers corresponded to the alloy and the dark ones to the SiC ceramic. They were alternately arranged. Some alloys were also discretely distributed in the ceramic layers while a few grey phases were observed in the metal layers. EDS analysis demonstrated that they were Mg2Si phase, whose amount,
302
H. Zhang et al. / Materials Letters 183 (2016) 299–302
Fig. 4. (a) Optical micrograph showing the microstructure of the AZ91/SiC composite prepared by reactive infiltration at 650 °C for 5 min; (b) flexural strengths of the AZ91/ SiC composite and the AZ91 alloy. The inset figure shows the fracture surface of the composite; (c) engineering compressive strengths of the AZ91/SiC composite and the AZ91 alloy. The inset figure shows the strength of the SiC scaffold.
morphology and distribution are presumed to have a significant influence on the mechanical property of the composite. Fig. 4(b)– (c) shows the flexural and compressive strength curves of the AZ91/SiC composites compared with the AZ91 alloy. The flexural strength reached 599 7 44 MPa, about 1.5 times of that of the alloy; while the compressive strength reached 7437 20 MPa, about 1.8 times of that of the alloy and 18.6 times of that of the SiC scaffold. As seen from the flexural fracture morphology given in the inset of Fig. 4(b), the cracks appeared in the ceramic layers and at the metal–ceramic interface. We conjecture that the cracks initiated in the brittle ceramic layer and then propagated to the interface. Because of the impedance of the tough metal layer, the cracks deflected and then propagated along the metal–ceramic interface. This led to the extension of the crack propagation path and increase in the energy for the crack growth, thus enhancing the toughness of the composite. In addition, the filling of the alloy in the cavities of the ceramic layers also increased the strength and toughness of the composite, contributing to superior combining properties. Nevertheless, some structural defects such as residual pores and over-grown Mg2Si phase could have detrimental effects on the mechanical properties, especially for the flexural strength and toughness, and thus should be eliminated or carefully controlled.
4. Conclusions We prepared the nacre-like lamellar AZ91/SiC composites by using a two-step processing route involving freeze casting (ice templating) and reactive infiltration. The preoxidation of the SiC scaffolds during sintering greatly promoted the wettability, leading to rapid spontaneous infiltration of the Mg alloy into their
bodies. The infiltration activation energy was estimated to be 7.74 kJ/mol and the infiltration rate reached 10.2 mm/min at 650 °C. The resultant composites exhibited much higher compressive and flexural strengths than the matrix AZ91 alloy.
Acknowledgment This work is supported by National Natural Science Foundation of China (No. 51571099) and National Basic Research Program of China (973 Program) (No. 2012CB619600).
References [1] M.R. Begley, N.R. Philips, B.G. Compton, D.V. Wilbrink, R.O. Ritchie, M. Utz, Micromechanical models to guide the development of synthetic ‘brick and mortar’ composites, J. Mech. Phys. Solids 60 (2012) 1545–1560. [2] S. Deville, E. Saiz, R.K. Nalla, A.P. Tomsia, Freezing as a path to build complex composites, Science 311 (2006) 515–518. [3] M.E. Launey, E. Munch, D.H. Alsem, E. Saiz, A.P. Tomsia, R.O. Ritchie, A novel biomimetic approach to the design of high-performance ceramic–metal composites, J. R. Soc. Interface 7 (2010) 741–753. [4] S. Roy, A. Wanner, Metal/ceramic composites from freeze-cast ceramic preforms: domain structure and elastic properties, Compos. Sci. Technol. 68 (2008) 1136–1143. [5] L.X. Shi, P. Shen, D. Zhang, E. Dong, Q.C. Jiang, Reactive wetting in liquid magnesium/silica and magnesium/silicon systems, Appl. Surf. Sci. 274 (2013) 124–130. [6] A. Shaga, P. Shen, R.F. Guo, Q.C. Jiang, Effects of oxide addition on the microstructure and mechanical properties of lamellar SiC scaffolds and Al–Si–Mg/SiC composites prepared by freeze casting and pressureless infiltration, Ceram. Inter. 42 (2016) 9653–9659. [7] V. Naglieri, H.A. Bale, B. Gludovatz, A.P. Tomsia, R.O. Ritchie, On the development of ice-templated silicon carbide scaffolds for nature-inspired structural materials, Acta Mater. 61 (2013) 6948–6957.
Contents lists available at ScienceDirect
Materials Letters journal homepage: www.elsevier.com/locate/matlet
Preparation of nacre-like composites by reactive infiltration of a magnesium alloy into porous silicon carbide derived from ice template Heng Zhang, Ping Shen n, Alateng Shaga, Ruifen Guo, Qichuan Jiang Key Laboratory of Automobile Materials (Ministry of Education), Department of Materials Science and Engineering, Jilin University, No. 5988 Renmin Street, Changchun, 130025 PR China
art ic l e i nf o
a b s t r a c t
Article history: Received 11 April 2016 Received in revised form 18 July 2016 Accepted 26 July 2016 Available online 27 July 2016
Lightweight and high-strength AZ91/SiC composites with lamellar structures were successfully prepared by using freeze casting (ice-templating) and reactive infiltration techniques. The infiltration dynamics was measured and activation energy calculated to be 7.74 kJ/mol. The compressive and flexural strengths of the composites with 30 vol% initial solid load reached 7437 20 MPa and 599 744 MPa, about 1.8 and 1.5 times of those of the AZ91 alloy, respectively. & 2016 Elsevier B.V. All rights reserved.
Keywords: Metal matrix composites (MMC) Multilayer structure Biomimetic Liquid infiltration Mechanical properties
1. Introduction The distinct “brick-and-mortar” structure provides nacre with high strength and toughness [1]. Inspired by nacre's fantastic structure and fabulous property, novel techniques are being developed to fabricate nacre-like materials [2–4]. Recently, Deville et al. [2] have adopted freezing as a flexible path to build sophisticated porous and layer-hybrid materials. Following this technique, Launey et al. [3] prepared the composites by infiltrating an Al– 12.6 wt% Si alloy into a 36 vol% porous Al2O3 scaffold with lamellar structures at 1173 K using gas-pressure infiltration, which exhibited fracture (crack-growth) toughness of 40 MPa m1/2 and bending strength of 300 MPa. Up to date, the preparation of the nacre-like metal–ceramic composites is limited to the Al matrix and most of the studies employed high-cost pressure infiltration technique. For the Mg matrix, to our knowledge, there is no report. As known, Mg has lower density and higher specific strength and modulus than Al, receiving wide attention in recent years for potential structural applications. However, Mg has strong chemical reactivity and high vapor pressure and thus is easy to ignite and evaporate at elevated temperatures, making the preparation of Mg-matrix composites difficult. Nevertheless, our previous study revealed that Mg has very good wettability with silica [5], favoring the spontaneous n
Corresponding author. E-mail address: [email protected] (P. Shen).
http://dx.doi.org/10.1016/j.matlet.2016.07.126 0167-577X/& 2016 Elsevier B.V. All rights reserved.
infiltration into silica-coated ceramic scaffolds. Based on this knowledge, in this work we prepared for the first time the lamellar Mg/SiC composites using freeze casting and reactive infiltration techniques.
2. Materials and methods The raw ceramic materials used in this work were commercially available SiC powders with an average particle size (d50) of 5 mm and a purity of 98.5 wt% together with 16.5 wt% Al2O3, 1.5 wt% MgO and 2 wt% Y2O3. They were added as sintering aids. Besides, 0.6 wt% CMC-Na was used as dispersant and deionized water as solvent. First, water-based SiC slurry involving sintering aids with a total solid load of 30 vol% was ball-milled and then de-aired. Subsequently, the slurry was poured into Teflon molds to shape into cylinders with 18 mm in diameter and 25 mm in height by unidirectional freezing at –20 °C. After demolding, the cold SiC bodies were freeze-dried at –50 °C in a 10 Pa vacuum for 24 h to remove the ice inside [6]. The resultant SiC bodies were then fired at 1100 °C in air for 2 h to form an oxide layer (silica) at the SiC surface and further sintered at 1500 °C in Ar atmosphere for 2 h. During the firing process, the heating and cooling rates were 5 °C/ min. In order to determine optimum infiltration parameters, in-situ observation of spontaneous infiltration for a commercial AZ91
300
H. Zhang et al. / Materials Letters 183 (2016) 299–302
(Mg–9Al–1Zn) alloy into porous SiC scaffolds was carried out. The AZ91 alloy was pre-placed on the SiC scaffold and the couple was heated in a high-purity (99.999%) Ar gas at 5 °C/min in a tube furnace, which was initially evacuated to a vacuum less than 5 Pa. During the heating and isothermal dwelling (at the temperature between 610 and 690 °C with an interval of 20 °C) process, the shape of the AZ91 alloy was monitored by a video camera through quartz windows to record its morphological change. The spontaneous infiltration dynamics was characterized by time variations in the alloy height and contact diameter. On the basis of the in-situ observation results, the infiltration of the AZ91 alloy into the SiC scaffolds was performed at 650 °C with a holding time of 5 min. Fully infiltrated samples were readily achieved. The microstructures of the composites were observed using an optical microscope (Axio Imager A2m, Carl Zeiss, Germany) and a scanning electron microscope (SEM, Evo18, Carl Zeiss, Germany), and phases were identified by X-ray diffraction (XRD, D/Max 2500PC Rigaku, Japan). The compressive and three-point flexural strengths of the composites were measured by a universal material testing machine (Instron 5689 Corp., USA).
3. Results and discussion Fig. 1 shows the microstructures of the SiC scaffold prepared by freeze casting. The sample displayed a well-developed lamellar structure in both longitudinal and transverse sections. Because SiC has a strong covalent bond and the atoms have low diffusion coefficients, it is difficult to achieve densification even sintering at a temperature of 2000 °C [7]. However, in the present work, the SiC surface was preoxidized at 1100 °C for 2 h, forming a SiO2 layer, which may further react with the sintering aids during subsequent high-temperature (1500 °C) firing and thus greatly promote sinterability and enhance the strength of the scaffold. Nevertheless, many cavities were still observed in the SiC lamellae; namely, the ceramic layers were still far from densification. By way of the preoxidation treatment, the spontaneous infiltration was readily achieved and the entire process can be divided into three stages (Fig. 2(a)): (i) alloy melting stage, (ii) inoculation stage and (iii) rapid infiltration stage. Because of the coverage of an oxide film at the Mg alloy surface, no obvious infiltration was observed in stages (i) and (ii). The decrease in the drop height and increase in the contact diameter in the end of stage (i) were virtually due to the collapse of the alloy after melting. During the inoculation stage (stage ii), the height of the
alloy and the contact diameter did not change. However, with the progressive evaporation of Mg, the oxide film was finally disrupted at some locations, thus paving the way for the spontaneous infiltration. The duration of the inoculation stage depends on temperature (Fig. 2(b)). Clearly, a higher temperature favors a larger Mg vapor pressure and thus faster disruption of the oxide film. As the clean liquid Mg alloy contacted the pre-oxidized SiC surface, rapid infiltration occurred, which is characterized by the sharp decrease in the height of the alloy drop together with a certain extent of liquid spreading (i.e., increase in the contact diameter as shown in Fig. 2(a). The infiltration rate (v) can be roughly evaluated from the time dependence of the drop height (dH/dt). Note that the infiltration rate reached 10.2 mm/min at 650 °C and 14.3 mm/min at 690 °C (Fig. 2(c)); thus the whole sample with a height of 25 mm could be fully infiltrated by the alloy in no more than 3 min providing that the amount of the alloy was sufficient. Furthermore, according to the Arrhenius equation:
lnv = − Ea/(RT ) + lnA,
(1)
where Ea is the infiltration activation energy, A is a constant, R is the gas constant, and T is the absolute temperature, we calculated Ea ¼7.74 kJ/mol. This small value suggests easy activation for the infiltration of the alloy into the porous SiC scaffolds. Fig. 3 shows the macro- and micro-structures of the infiltrating sample, particularly at the liquid infiltration front area. The sample could be characterized by three regions: (i) a fully infiltrated region, (ii) a partially infiltrated region and (iii) a non-infiltrated region. In the fully infiltrated region, the alloy not only distributed in the open channels but also in the cavities of the ceramic layers; whereas, due to the limited amount of liquid in the infiltration front area, the alloy was found only in the small cavities of the ceramic layers but not in the large open channels (Fig. 3(b)-2). The elemental mapping graphs (Fig. 3(c)) indicate that Mg had a selective distribution in the front area, whose position largely coincided with that of Si in the SiC layer, suggesting that the liquid alloy preferentially penetrated into the small cavities in the ceramic layer and then into the large open channels between the ceramic layers under the capillary force. This phenomenon can be explained by the Laplace equation:
P = 2σ cos θ /r ,
(2)
where P is the capillary force, s is the surface tension of the melt, r is the capillary radius and θ is the contact angle. In the case of θ { 90° for Mg on SiO2 [5], the smaller the capillary radius, the larger the capillary force, and as a consequence, the small cavities
Fig. 1. SEM micrographs showing the lamellar structure of the SiC scaffold: (a) perpendicular to the freezing direction; (b) parallel to the freezing direction.
H. Zhang et al. / Materials Letters 183 (2016) 299–302
301
Fig. 2. (a) Variations in the alloy height and contact diameter with time during heating to and dwelling at 650 °C. Due to the coverage of an oxide film, the alloy was in an irregular shape after melting, and thus the height was characterized by the average of the highest and lowest points, as schematically shown in the inserted image. (b) Variation in the alloy height with time during isothermal dwelling at different temperatures; (c) relationship between infiltration rate (v and lnv) and temperature (T and 1/T).
Fig. 3. (a) Macrograph of the infiltrating sample for dynamic analysis after holding at 650 °C; (b) optical micrographs showing the microstructures at partially infiltrated region; (c) SEM microstructure and elemental distributions at the liquid infiltration front.
in the ceramic layer were preferentially filled by the liquid alloy. The larger open pores would be certainly filled by the molten alloy as well providing that the amount of the alloy was sufficient. Besides, the rapid infiltration could be associated with the heat released from the reaction between Mg and SiO2 or silicate phase, which increased temperature and improved fluidity of the alloy at the infiltration front. As a consequence, the infiltration was
accelerated. Fig. 4(a) gives the optical micrograph of the composite structure. The bright layers corresponded to the alloy and the dark ones to the SiC ceramic. They were alternately arranged. Some alloys were also discretely distributed in the ceramic layers while a few grey phases were observed in the metal layers. EDS analysis demonstrated that they were Mg2Si phase, whose amount,
302
H. Zhang et al. / Materials Letters 183 (2016) 299–302
Fig. 4. (a) Optical micrograph showing the microstructure of the AZ91/SiC composite prepared by reactive infiltration at 650 °C for 5 min; (b) flexural strengths of the AZ91/ SiC composite and the AZ91 alloy. The inset figure shows the fracture surface of the composite; (c) engineering compressive strengths of the AZ91/SiC composite and the AZ91 alloy. The inset figure shows the strength of the SiC scaffold.
morphology and distribution are presumed to have a significant influence on the mechanical property of the composite. Fig. 4(b)– (c) shows the flexural and compressive strength curves of the AZ91/SiC composites compared with the AZ91 alloy. The flexural strength reached 599 7 44 MPa, about 1.5 times of that of the alloy; while the compressive strength reached 7437 20 MPa, about 1.8 times of that of the alloy and 18.6 times of that of the SiC scaffold. As seen from the flexural fracture morphology given in the inset of Fig. 4(b), the cracks appeared in the ceramic layers and at the metal–ceramic interface. We conjecture that the cracks initiated in the brittle ceramic layer and then propagated to the interface. Because of the impedance of the tough metal layer, the cracks deflected and then propagated along the metal–ceramic interface. This led to the extension of the crack propagation path and increase in the energy for the crack growth, thus enhancing the toughness of the composite. In addition, the filling of the alloy in the cavities of the ceramic layers also increased the strength and toughness of the composite, contributing to superior combining properties. Nevertheless, some structural defects such as residual pores and over-grown Mg2Si phase could have detrimental effects on the mechanical properties, especially for the flexural strength and toughness, and thus should be eliminated or carefully controlled.
4. Conclusions We prepared the nacre-like lamellar AZ91/SiC composites by using a two-step processing route involving freeze casting (ice templating) and reactive infiltration. The preoxidation of the SiC scaffolds during sintering greatly promoted the wettability, leading to rapid spontaneous infiltration of the Mg alloy into their
bodies. The infiltration activation energy was estimated to be 7.74 kJ/mol and the infiltration rate reached 10.2 mm/min at 650 °C. The resultant composites exhibited much higher compressive and flexural strengths than the matrix AZ91 alloy.
Acknowledgment This work is supported by National Natural Science Foundation of China (No. 51571099) and National Basic Research Program of China (973 Program) (No. 2012CB619600).
References [1] M.R. Begley, N.R. Philips, B.G. Compton, D.V. Wilbrink, R.O. Ritchie, M. Utz, Micromechanical models to guide the development of synthetic ‘brick and mortar’ composites, J. Mech. Phys. Solids 60 (2012) 1545–1560. [2] S. Deville, E. Saiz, R.K. Nalla, A.P. Tomsia, Freezing as a path to build complex composites, Science 311 (2006) 515–518. [3] M.E. Launey, E. Munch, D.H. Alsem, E. Saiz, A.P. Tomsia, R.O. Ritchie, A novel biomimetic approach to the design of high-performance ceramic–metal composites, J. R. Soc. Interface 7 (2010) 741–753. [4] S. Roy, A. Wanner, Metal/ceramic composites from freeze-cast ceramic preforms: domain structure and elastic properties, Compos. Sci. Technol. 68 (2008) 1136–1143. [5] L.X. Shi, P. Shen, D. Zhang, E. Dong, Q.C. Jiang, Reactive wetting in liquid magnesium/silica and magnesium/silicon systems, Appl. Surf. Sci. 274 (2013) 124–130. [6] A. Shaga, P. Shen, R.F. Guo, Q.C. Jiang, Effects of oxide addition on the microstructure and mechanical properties of lamellar SiC scaffolds and Al–Si–Mg/SiC composites prepared by freeze casting and pressureless infiltration, Ceram. Inter. 42 (2016) 9653–9659. [7] V. Naglieri, H.A. Bale, B. Gludovatz, A.P. Tomsia, R.O. Ritchie, On the development of ice-templated silicon carbide scaffolds for nature-inspired structural materials, Acta Mater. 61 (2013) 6948–6957.