J.Am.cerm.Soc.83|8]1885-88(200 urna Electrophoretic Deposition Forming of Nickel-Coated-Carbon-Fiber Reinforced Borosilicate- Glass-Matrix Composites Cengiz kaya Interdisciplinary Research Centre (IRC) for High Performance Applications and School of Metallurgy and Materials, The University of Birmingham, Edgbaston, Birmingham, B15 2TT, United Kingdom Aldor boccaccini Technische Universitat Ilmenau, Fachgebiet Werkstofftechnik, PF 100565, D-98684 Ilmenau, Germany Krishan K Chawla' Department of Materials and Mechanical Engineering, University of Alabama, Birmingham, Alabama 35294 The present paper introduces a novel processing technique electric field is applied, will move and deposit onto an oppositely hat involves in situ electrophoretic deposition(EPD), followed charged electrode. Theoretical models and experimental findings by pressureless sintering, to produce dense, defect-minimized, have shown that EPD consists of two main processes: electro- carbon-fiber-reinforced borosilicate-glass-matrix composites with a nickel interface. The process relies on the deposition of charged particles that are suspended in a stable sol/suspension, under the influence of an electric field that is applied across the sectionally aligned nickel-coated carbon fibers. The prepara- uspension. Particles with a net electrostatic charge move in a tion and characterization of a kinetically stable nanosized rection parallel to the applied electric field and deposit onto the borosilicate sol suitable for EPd are described. The most- important EPD processing parameters in the formation of oppositely charged electrode(the second of EPD), depending dense, fully infiltrated, green-body compacts are described, on the charge of the particles. The rate of deposition is very fast and issues that concern the infiltration of very tight carbon d can be controlled by controlling the applied potential. Thus, fiber preforms are discussed and effectively solved. Using the high-green-density products that contain no organic compounds crack-path-propagation test, the metallic nickel interface is (hence, a burn-out step is not necessary) are achievable in very determined to be very effective to improve the composite short times(as short as 2 min). In fiber-reinforced CMCs, the interface between the matrix and mechanical performance in terms of the nonbrittle fracture the fiber determines the composite behavior, in terms of mechan behavior. Catastrophic crack growth is prevented by such ical strength, fracture mode, and thermomechanical stability. The and fiber debonding and pullout. The proposed processing key issue in obtaining damage-tolerant fracture behavior in these technique has great potential to fabricate defect-minimized composites is to supply a weak interface, to optimize the coupling and damage-tolerant fiber-reinforced brittle-matrix compos- of the fiber to the matrix. The effect of the interfacial structure on ites with a ductile interface. Overall, this new approach offers the mechanical response of CMCs at room temperature and high a cost-effective and short-time processing route for the fabri- cation of continuous-fiber-reinforeed ceramic-matrix compos- ystems, such as carbon, zirconia(ZrO2), boron nitride(BN) alumina(Al,O,), mullite, and monazite. However, a very limited number of studies on metallic interfaces have been pub- lished in the literature, 20-2 thus, more experimental data are L. Introduction quired to assess their full potential for applications in the design of CMCs. The use of ductile interfaces is expected to provide a E techniques for fabricating ceramic-matrix composites degree of pseudoductility by such mechanisms as fiber debonding ( CMCs)that have been reinforced with fibers, whiskers, or and pullout, which are typical for CMCs, as well as a dissipation lenges, as emphasized recently in detail. Fecal stages and chal- of energy through constrained plastic deformation,thus prevent- on(EPD)has been proposed as a novel, relatively simple, ing catastrophic crack growth during the failure of the composite cost-effective and high-forming-rate technique for producing ce- The debonding process at a tough, ductile interface also will require relatively high plastic work, which results further and fiber-reinforced composites- EPD has been use sfully on different reinforcement and matrix material com contribution to the increase in toughness and noncatastrophic fracture behavior. 24 ns. This process relies on the presence of very small ged particles in liquid suspension, i.e., a sol, which, when ar The present study examines the feasibility of producing nickel- coated-carbon-fiber-reinforced borosilicate-glass-matrix compos- ites, using a simple technique that is based on EPD and pressure less sintering. Preliminary successful efforts on composite J. W. Halloran--contributing editor abrication are reported here, in addition to promising results regarding the damage-tolerant fracture behavior of the fabr composites. Indeed, carbon-fiber reinforcement of glass matrixes has been explored extensively in the past.25-27However, the m.18BMm由159u39.9 ost-intensive hot-pressing fabrication technique has been used TAlso affiliated with Metallurgical Engineering Department, Yildiz Technical invariably, and no special consideration has been given to optimi- University, Istanbul 8750, Turkey zation of the interface. It is anticipated that the use of a nickel
Electrophoretic Deposition Forming of Nickel-Coated-Carbon-FiberReinforced Borosilicate-Glass-Matrix Composites Cengiz Kaya† Interdisciplinary Research Centre (IRC) for High Performance Applications and School of Metallurgy and Materials, The University of Birmingham, Edgbaston, Birmingham, B15 2TT, United Kingdom Aldo R. Boccaccini* Technische Universitat Ilmenau, Fachgebiet Werkstofftechnik, PF 100565, D-98684 Ilmenau, Germany Krishan K. Chawla* Department of Materials and Mechanical Engineering, University of Alabama, Birmingham, Alabama 35294 The present paper introduces a novel processing technique that involves in situ electrophoretic deposition (EPD), followed by pressureless sintering, to produce dense, defect-minimized, carbon-fiber-reinforced borosilicate-glass-matrix composites with a nickel interface. The process relies on the deposition of submicrometer-sized, colloidal charged particles onto unidirectionally aligned nickel-coated carbon fibers. The preparation and characterization of a kinetically stable nanosized borosilicate sol suitable for EPD are described. The mostimportant EPD processing parameters in the formation of dense, fully infiltrated, green-body compacts are described, and issues that concern the infiltration of very tight carbon fiber preforms are discussed and effectively solved. Using the crack-path-propagation test, the metallic nickel interface is determined to be very effective to improve the composite mechanical performance, in terms of the nonbrittle fracture behavior. Catastrophic crack growth is prevented by such mechanisms as constrained plastic deformation of the interface and fiber debonding and pullout. The proposed processing technique has great potential to fabricate defect-minimized and damage-tolerant fiber-reinforced brittle-matrix composites with a ductile interface. Overall, this new approach offers a cost-effective and short-time processing route for the fabrication of continuous-fiber-reinforced ceramic-matrix composites. I. Introduction THE techniques for fabricating ceramic-matrix composites (CMCs) that have been reinforced with fibers, whiskers, or particulates/platelets involve some very critical stages and challenges, as emphasized recently in detail.1 Electrophoretic deposition (EPD) has been proposed as a novel, relatively simple, cost-effective and high-forming-rate technique for producing ceramic and fiber-reinforced composites.2 EPD has been used successfully on different reinforcement and matrix material combinations.2–14 This process relies on the presence of very small charged particles in liquid suspension, i.e., a sol, which, when an electric field is applied, will move and deposit onto an oppositely charged electrode. Theoretical models and experimental findings have shown that EPD consists of two main processes: electrophoresis and deposition. The first process is the movement of charged particles that are suspended in a stable sol/suspension, under the influence of an electric field that is applied across the suspension. Particles with a net electrostatic charge move in a direction parallel to the applied electric field and deposit onto the oppositely charged electrode (the second stage of EPD), depending on the charge of the particles. The rate of deposition is very fast and can be controlled by controlling the applied potential. Thus, high-green-density products that contain no organic compounds (hence, a burn-out step is not necessary) are achievable in very short times (as short as 2 min).9 In fiber-reinforced CMCs, the interface between the matrix and the fiber determines the composite behavior, in terms of mechanical strength, fracture mode, and thermomechanical stability.15 The key issue in obtaining damage-tolerant fracture behavior in these composites is to supply a weak interface, to optimize the coupling of the fiber to the matrix. The effect of the interfacial structure on the mechanical response of CMCs at room temperature and high temperature has been widely studied using different fiber-coating systems, such as carbon, zirconia (ZrO2), boron nitride (BN), alumina (Al2O3), mullite, and monazite.16–19 However, a very limited number of studies on metallic interfaces have been published in the literature;20–22 thus, more experimental data are required to assess their full potential for applications in the design of CMCs. The use of ductile interfaces is expected to provide a degree of pseudoductility by such mechanisms as fiber debonding and pullout, which are typical for CMCs, as well as a dissipation of energy through constrained plastic deformation,23 thus preventing catastrophic crack growth during the failure of the composite. The debonding process at a tough, ductile interface also will require relatively high plastic work, which results in a further contribution to the increase in toughness and noncatastrophic fracture behavior.24 The present study examines the feasibility of producing nickelcoated-carbon-fiber-reinforced borosilicate-glass-matrix composites, using a simple technique that is based on EPD and pressureless sintering. Preliminary successful efforts on composite fabrication are reported here, in addition to promising results regarding the damage-tolerant fracture behavior of the fabricated composites. Indeed, carbon-fiber reinforcement of glass matrixes has been explored extensively in the past.25–27 However, the cost-intensive hot-pressing fabrication technique has been used invariably, and no special consideration has been given to optimization of the interface. It is anticipated that the use of a nickel J. W. Halloran—contributing editor Manuscript No. 189500. Received March 15, 1999; approved July 29, 1999. *Member, American Ceramic Society. † Also affiliated with Metallurgical Engineering Department, Yildiz Technical University, Istanbul 8750, Turkey. J. Am. Ceram. Soc., 83 [8] 1885–88 (2000) 1885 journal
Journal of the American Ceramic Society-Kaya et al interface will not compromise the temperature stability of carbon- fiber-reinforced glass-matrix composites, which is dictated by the softening temperature of the viscous matrix(550C for borosi cate glass)or the oxidation resistance of the carbon fiber IL. Experimental Work (1) Materials Digital balance Nanosized borosilicate powders suitable for use in EPD exper- iments were synthesized via the hydrolysis and condensation of tetraethyl orthosilicate (Si(oC2Os)4) and triisopropylborate B(OC3 H,)3). The average particle size of the powders produced was 80 nm, and they had a spherical morphology, as determined via transmission electron microscopy(TEM) and particle-size analysis. The latter procedure was performed using a X-ray disk ntrifuge system(Model Bl-XDC, Holland) that could measure ultrafine particles that were dispersed in diluted (1-5 wt%) suspensions very accurately. This powder can be sintered to full density at 800C, as shown elsewhere. To prepare a kinetically stable borosilicate colloidal sol, distilled water was stirred vige ously while the nanosized powders were added at a rate of 0.5 g/min. The low rate of addition prevented the formation of large, heteroflocculated clusters The solids loading was adjusted within the range of 15-20 wt% of the dispersion liquid, with simultaneous ultrasonic agitation to enhance powder dispersion. Homogeneous well-dispersed, and agglomerate-free stable suspensions were obtained at a pH value of 8.5. The surface-charge properties of the powders, in terms of their electrophoretic mobility and net surface charge, were measured as a function of the sol pH, using a urface-charge analyzer. Nickel-coated carbon fibers (INCOFIB Electrode ER, Product 12 K50, Inco Specialty Powdered Products, Lor colloidal sc Ni coated fibers Electrode don, U. K)were used as reinforcement. These fibers were supplied in the form of continuous tows of nickel-coated single carbon Frame fibers: 12 000 tows were coated with 50 wt% nickel. which was deposited using gas plating technology. The fiber diameter and nicke ckelsoated earbon fibers as the de in situ EPD cell that incorporates ition(anode)electrode. nickel coating were very homogeneous and had average 15 and 0.5 um, respectively. The nickel coating provided conductivity, which is essential for EPD, as well of fiber handling and adequate wettability I h in a nitrogen atmosphere, To prepare green and sintered fiber-reinforced CMCs for cross-sectional scanning electron mi (2 In Situ Electrophoretic Deposition roscopy (SEM) analysis, the samples were placed in a vacuum An in situ EPd cell was designed to infiltrate the nickel-coated chamber and vacuum-impregnated with resin(Epofix, Struers carbon-fiber tows, which were unidirectional aligned Tech A/S, Denmark). Impregnated green and sintered CMCs grooved Perspex frame with the borosilicate glass sol. The EPD samples were allowed to harden overnight and then cut into slices ell that was used is shown schematically in Fig. 1. The distance a diamond saw. High-resolution SEM(Model $-4000 between each tow was chosen to be 0.2 mm. Nickel-coated carbon field-emission-gun SEM microscope, Hitachi, Tokyo, Japan)was fiber held in the frame was used as the deposition electrode used to examine the microstructural details. such as ductile nterface, deposit thickness, and infiltration of the matrix into the (anode). Two stainless-steel plates on either side of the anode fiber architecture on both green and sintered samples. To charac- was placed in the sol, the system was vacuum-degassed to remove terize the interfacial behavior of the composite, the crack-path any entrapped air, then, the cell electrodes were connected to a bservation technique was used 0-60 V dc power supply. EPD was performed subsequently, under constant-voltage conditions (6,8, 10, and 12 V), for various II Results and discussion deposition times (100, 200, 280, and 380 s)with a constant electrode separation distance of 15 mm. Under the applied electric The graph in Fig. 2 shows the results of the electrophoretic ield, the very fine borosilicate glass particles( which possess a net deposit weight, thickness, and deposit formation rate, as a function negative surface charge, as determined from the electrophoretic of deposition time under an optimized applied voltage (6 V) and mobility data) migrated toward the positive electrode, i.e., the electrode distance (15 mm). The increase in deposit weight and nickel-coated carbon-fiber tows. The particles infiltrated the fiber thickness is almost linear over the time period that was studied tows and deposited until a sufficient matrix thickness, which EPD experiments were performed for durations of up to 280 S, veloped the fiber tows, was achieved. The fiber preform that because this time period gave a deposit thickness of 650 ur ted as an electrode was connected to a balance that was linked to which was sufficient to produce a composite with an acceptable computer. The apparatus was able to record the weight gain per green density. The deposit thickness increased as the deposition millisecond during the deposition process, ie, in real time. The time increased, because the amount of particles deposited onto the dimensions of the anode(25 mm X 25 mm) were half of those of fiber preform increased. Real-time measurements of the deposit the cathode (50 mm X 50 mm), to eliminate the"edge effect, weight provided very reliable data to determine the deposition rate which may give an inhomogeneous thickness from the center to accurately. The deposition rate given here was normalized for the the edges of the anode. The EPD-prepared green-body specimens weight of nickel-coated carbon fibers that were aligned in the that had a fiber loading of -35 vol% were dried in a humidity perspex frame. Thus, the deposition rate has been given in terms of controlled atmosphere for I day and left in ambient air for another grams of electrophoretic deposit formed per gram of fiber per day before being subjected to pressureless sintering at 800C fo second. For the first 40 S, the deposition rate was very high but
interface will not compromise the temperature stability of carbonfiber-reinforced glass-matrix composites, which is dictated by the softening temperature of the viscous matrix (;550°C for borosilicate glass) or the oxidation resistance of the carbon fiber. II. Experimental Work (1) Materials Nanosized borosilicate powders suitable for use in EPD experiments were synthesized via the hydrolysis and condensation of tetraethyl orthosilicate (Si(OC2O5)4) and triisopropylborate (B(OC3H7)3).28 The average particle size of the powders produced was 80 nm, and they had a spherical morphology, as determined via transmission electron microscopy (TEM) and particle-size analysis. The latter procedure was performed using a X-ray disk centrifuge system (Model BI-XDC, Holland) that could measure ultrafine particles that were dispersed in diluted (1–5 wt%) suspensions very accurately. This powder can be sintered to full density at 800°C, as shown elsewhere.29 To prepare a kinetically stable borosilicate colloidal sol, distilled water was stirred vigorously while the nanosized powders were added at a rate of 0.5 g/min. The low rate of addition prevented the formation of large, heteroflocculated clusters. The solids loading was adjusted within the range of 15–20 wt% of the dispersion liquid, with simultaneous ultrasonic agitation to enhance powder dispersion. Homogeneous, well-dispersed, and agglomerate-free stable suspensions were obtained at a pH value of 8.5. The surface-charge properties of the powders, in terms of their electrophoretic mobility and net surface charge, were measured as a function of the sol pH, using a surface-charge analyzer. Nickel-coated carbon fibers (INCOFIBERTM, Product 12 K50, Inco Specialty Powdered Products, London, U.K.) were used as reinforcement. These fibers were supplied in the form of continuous tows of nickel-coated single carbon fibers; 12 000 tows were coated with 50 wt% nickel, which was deposited using gas plating technology. The fiber diameter and nickel coating were very homogeneous and had average values of 15 and 0.5 mm, respectively. The nickel coating provided excellent conductivity, which is essential for EPD, as well as ease of fiber handling and adequate wettability. (2) In Situ Electrophoretic Deposition An in situ EPD cell was designed, to infiltrate the nickel-coated carbon-fiber tows, which were unidirectionaly aligned in a grooved Perspex frame with the borosilicate glass sol. The EPD cell that was used is shown schematically in Fig. 1. The distance between each tow was chosen to be 0.2 mm. Nickel-coated carbon fiber held in the frame was used as the deposition electrode (anode). Two stainless-steel plates on either side of the anode served as the negative (cathode) electrodes. After the fiber preform was placed in the sol, the system was vacuum-degassed to remove any entrapped air; then, the cell electrodes were connected to a 0–60 V dc power supply. EPD was performed subsequently, under constant-voltage conditions (6, 8, 10, and 12 V), for various deposition times (100, 200, 280, and 380 s) with a constant electrode separation distance of 15 mm. Under the applied electric field, the very fine borosilicate glass particles (which possess a net negative surface charge, as determined from the electrophoretic mobility data) migrated toward the positive electrode, i.e., the nickel-coated carbon-fiber tows. The particles infiltrated the fiber tows and deposited until a sufficient matrix thickness, which enveloped the fiber tows, was achieved. The fiber preform that acted as an electrode was connected to a balance that was linked to a computer. The apparatus was able to record the weight gain per millisecond during the deposition process, i.e., in real time. The dimensions of the anode (25 mm 3 25 mm) were half of those of the cathode (50 mm 3 50 mm), to eliminate the “edge effect,” which may give an inhomogeneous thickness from the center to the edges of the anode. The EPD-prepared green-body specimens that had a fiber loading of ;35 vol% were dried in a humiditycontrolled atmosphere for 1 day and left in ambient air for another day before being subjected to pressureless sintering at 800°C for 1 h in a nitrogen atmosphere. To prepare green and sintered fiber-reinforced CMCs for cross-sectional scanning electron microscopy (SEM) analysis, the samples were placed in a vacuum chamber and vacuum-impregnated with resin (Epofix, Struers Tech A/S, Denmark). Impregnated green and sintered CMCs samples were allowed to harden overnight and then cut into slices using a diamond saw. High-resolution SEM (Model S-4000 field-emission-gun SEM microscope, Hitachi, Tokyo, Japan) was used to examine the microstructural details, such as ductile interface, deposit thickness, and infiltration of the matrix into the fiber architecture on both green and sintered samples. To characterize the interfacial behavior of the composite, the crack-path observation technique was used.30 III. Results and Discussion The graph in Fig. 2 shows the results of the electrophoretic deposit weight, thickness, and deposit formation rate, as a function of deposition time under an optimized applied voltage (6 V) and electrode distance (15 mm). The increase in deposit weight and thickness is almost linear over the time period that was studied. EPD experiments were performed for durations of up to 280 s, because this time period gave a deposit thickness of 650 mm, which was sufficient to produce a composite with an acceptable green density. The deposit thickness increased as the deposition time increased, because the amount of particles deposited onto the fiber preform increased. Real-time measurements of the deposit weight provided very reliable data to determine the deposition rate accurately. The deposition rate given here was normalized for the weight of nickel-coated carbon fibers that were aligned in the perspex frame. Thus, the deposition rate has been given in terms of grams of electrophoretic deposit formed per gram of fiber per second. For the first 40 s, the deposition rate was very high but Fig. 1. Schematic diagram of an in situ EPD cell that incorporates nickel-coated carbon fibers as the deposition (anode) electrode. 1886 Journal of the American Ceramic Society—Kaya et al. Vol. 83, No. 8
August 2000 ng of Ni-Coaled-C 200 diamond saw during the very-slow-speed cutting operation. Dense a composite samples were obtaine 8000C for I h Sintered relative densities of -98% were measured using the Archimedes technique. This result confirms that the present processing approach is very convenient, in terms of obtaining dense CMCs samples at relatively low sintering tempe procedure, which has been the common practice in previous studies on carbon-fiber-reinforced composites. A crack-path propagation test20, 30 was performed on sintered samples to characterize the behavior of the nickel interface, in terms of the plastic deformation, crack de Figure 4 shows the interactions between an indenter-induced crack that has been created within the borosilicate matrix and the ductile nickel interface. The crack first interacts with the nickel coating and then is deflected at the interface. The development of limited ectrophoretic Deposition Time(s) plastic deformation within the nickel interface also is visible, as of the weight/thickness of the electrophoretic deposit and indicated by the arrows. One can conclude from the crack path in ation rate, as a function of EPD time, under the optimized Fig 4 that a weak bonding between the nickel coating and the glass V)and electrode distance(15 mm), as obtained from the natrix exists. Thus, no significant chemical reaction between the design nickel interface and the borosilicate matrix at the sintering tem nickel coating evidently provides a potentially weak interface between the carbon fiber and the glass matrix, which results in decrease in deposition rate was attributed to the increase in the crack deflection and debonding. These mechanisms should result electrical resistance of the deposit; the current diminishes because of the increase in the deposit thickness and the removal of charged Because of plastic deformation of the interface, a further contri- borosilicate particles from the suspension. As a result of the bution to a desired noncatastrophic behavior with an enhanced decrease in the potential decrease across the suspension, the work of fracture is expected. elocity and deposition rate of the charged colloidal borosilicate articles decreased also, which was consistent with previous Figure 3 shows an SEM micrograph of an EPD-formed green body with a fiber loading of 35 vol%. High green densities(73% The electrophoretic deposition(EPD)technique that has been of the theoretical density) were achieved at 6 v for 280 s. The processing technique for the fabrication of nickel-coated-carbon- green density was measured by dividing the mass of the sample by fiber-reinforced borosilicate-glass-matrix composites. Using an ts geometrically determined volume. Figure 3 also shows that the applied voltage of 6 V for 280 s, high-quality green bodies were nickel-coated-carbon-fiber preform has been fully fabricated via the EPD infiltration of the fiber preforms by the borosilicate sol. Even regions where the nickel-coated fibers nanosized borosilicate-glass particles. These green bodies were were almost touching each other were fully impregnated by the densified by pressureless sintering at relatively low temperatures anosized borosilicate powders in a very short time (ie, 280 s). (800%C). Application of a nickel coating onto the fibers supplied a which leads to high-quality green bodies. Note that this green weak interfacial bonding, which resulted in crack deflection and sample was not polished, so that the ductile nickel interface would debonding mechanisms, crack-path-propagation tests have con- not be damaged. Thus. some cutting effects are visible on the firmed this observation. This interface also exhibited local plastic carbon fibers and the matrix(Fig. 3), which resulted from the deformation, which should contribute further to a desired non- catastrophic behavior with high strain to fracture. The use of a o um Fig 3. SEM micrograph of the EPD-infiltrated unsintered nickel-coated- 1.67μm arbon-fiber-reinforced borosilicate-glass-matrix composite with a fiber Fig. 4. SEM micrograph of EPD-infiltrated nickel-coated-carbon-fiber ding of 35 vol% before sintering, the fiber preform was infiltrated using reinforced borosilicate-glass-matrix composites with a fiber loading of 35 an applied voltage of 6 V for 280 s. Note the fully infiltrated microstruc vol%(sample sintered at 800C for I h under a nitrogen atmosphere) ure Damage observed on the fiber and matrix, as well as partial damage showing the interaction of an indenter-induced crack with the nickel on the nickel interface, is due to the cutting operation, because the green interface. Note the plastic deformation of the ductile nickel interface and sample was not polished. the occurrence of both crack-deflection and debonding mechanisms
then began to decrease as the deposition time increased. The decrease in deposition rate was attributed to the increase in the electrical resistance of the deposit; the current diminishes because of the increase in the deposit thickness and the removal of charged borosilicate particles from the suspension. As a result of the decrease in the potential decrease across the suspension, the velocity and deposition rate of the charged colloidal borosilicate particles decreased also, which was consistent with previous findings.9 Figure 3 shows an SEM micrograph of an EPD-formed green body with a fiber loading of 35 vol%. High green densities (;73% of the theoretical density) were achieved at 6 V for 280 s. The green density was measured by dividing the mass of the sample by its geometrically determined volume. Figure 3 also shows that the nickel-coated-carbon-fiber preform has been fully infiltrated with the borosilicate sol. Even regions where the nickel-coated fibers were almost touching each other were fully impregnated by the nanosized borosilicate powders in a very short time (i.e., 280 s), which leads to high-quality green bodies. Note that this green sample was not polished, so that the ductile nickel interface would not be damaged. Thus, some cutting effects are visible on the carbon fibers and the matrix (Fig. 3), which resulted from the diamond saw during the very-slow-speed cutting operation. Dense composite samples were obtained via pressureless sintering at 800°C for 1 h. Sintered relative densities of ;98% were measured using the Archimedes technique. This result confirms that the present processing approach is very convenient, in terms of obtaining dense CMCs samples at relatively low sintering temperatures and avoiding the cost-intensive, hot-pressing fabrication procedure, which has been the common practice in previous studies on carbon-fiber-reinforced composites.25–27 A crack-path propagation test20,30 was performed on sintered samples to characterize the behavior of the nickel interface, in terms of the plastic deformation, crack deflection, and debonding. Figure 4 shows the interactions between an indenter-induced crack that has been created within the borosilicate matrix and the ductile nickel interface. The crack first interacts with the nickel coating and then is deflected at the interface. The development of limited plastic deformation within the nickel interface also is visible, as indicated by the arrows. One can conclude from the crack path in Fig. 4 that a weak bonding between the nickel coating and the glass matrix exists. Thus, no significant chemical reaction between the nickel interface and the borosilicate matrix at the sintering temperature (i.e., 800°C) has occurred. This figure also shows that the nickel coating evidently provides a potentially weak interface between the carbon fiber and the glass matrix, which results in crack deflection and debonding. These mechanisms should result in an overall “pseudo-ductile” fracture mode of the composite. Because of plastic deformation of the interface, a further contribution to a desired noncatastrophic behavior with an enhanced work of fracture is expected. IV. Conclusions The electrophoretic deposition (EPD) technique that has been presented in this paper is a cost-effective and high-forming-rate processing technique for the fabrication of nickel-coated-carbonfiber-reinforced borosilicate-glass-matrix composites. Using an applied voltage of 6 V for 280 s, high-quality green bodies were fabricated via the EPD infiltration of the fiber preforms by nanosized borosilicate-glass particles. These green bodies were densified by pressureless sintering at relatively low temperatures (800°C). Application of a nickel coating onto the fibers supplied a weak interfacial bonding, which resulted in crack deflection and debonding mechanisms; crack-path-propagation tests have confirmed this observation. This interface also exhibited local plastic deformation, which should contribute further to a desired noncatastrophic behavior with high strain to fracture. The use of a Fig. 2. Graph of the weight/thickness of the electrophoretic deposit and the deposit formation rate, as a function of EPD time, under the optimized applied voltage (6 V) and electrode distance (15 mm), as obtained from the in situ EPD cell design. Fig. 3. SEM micrograph of the EPD-infiltrated unsintered nickel-coatedcarbon-fiber-reinforced borosilicate-glass-matrix composite with a fiber loading of 35 vol% before sintering; the fiber preform was infiltrated using an applied voltage of 6 V for 280 s. Note the fully infiltrated microstructure. Damage observed on the fiber and matrix, as well as partial damage on the nickel interface, is due to the cutting operation, because the green sample was not polished. Fig. 4. SEM micrograph of EPD-infiltrated nickel-coated-carbon-fiberreinforced borosilicate-glass-matrix composites with a fiber loading of 35 vol% (sample sintered at 800°C for 1 h under a nitrogen atmosphere), showing the interaction of an indenter-induced crack with the nickel interface. Note the plastic deformation of the ductile nickel interface and the occurrence of both crack-deflection and debonding mechanisms. August 2000 Electrophoretic Deposition Forming of Ni-Coated-C-Fiber-Reinforced Borosilicate Glass-Matrix Composites 1887
Journal of the American Ceramic Society-Kaya er metallic interface, which has been explored in this work, should be Hirata, A. Nishimoto, and Y, Ishihara, "Forming of Alumina Powder by applicable to other oxide- fiber-oxide-matrix systems trophoretic Deposition, " Nippon Seramikkusu Kyokai Gakiyutusu Ronbunshi, 2108-13(1991) F. Harbach and H Nienburg, "Homogeneous Functional Ceramic Components oretic Deposition from Stable Colloidal Suspensions-. Basic Acknowledgments Concepts and Application to Zirconia, " J. Eur. Ceram. Soc., 18, 675-83(1998). A. R. Boccaccini and P. A. Trusty,""Electrophoretic Deposition Infiltration of Metallic Fabrics with a Boehmite Sol for the Preparation of Ductile-Phase-Toughened Professors M. H. Loretto and L. R. Harris are acknowledged for the provision of laboratory facilities at the IRC and School of Metallurgy and Materials and at the University of Birmingham, respectively. The authors also thank Dr, Less Renny(Inco P Sarkar and P. S. Nicholson, "Electrophoretic Deposition(EPD): Mechanisms, Speciality Powder Products)for supplying the fibers Kinetics, and Applications to Ceramics, J. Am. Ceram. Soc., 79[8] 1987-2002 K. K Chawla, Ceramic Matrix Composites. Chapman and Hall, London, U.K, References 16B. W. Sheldon, E. Y. Sun, S. R. Nutt, and JJ. Brennan,"Oxidation of BN-Coated SiC Fibers in Ceramic Matrix Composites, J. Am. Ceram. Soc., 79 [21 R. L Lehman, S. K. EI-Rahaiby, and J. B. Wachtman Jr. (eds ) Handbook on opposites. American Cera B. Davis, J. P. A. Lofvander, A G. Evans, and E. Bischoff, and M. L. Emiliani Fiber Coating Concepts for Brittle-Matrix Composites, J Am. Ceram Soc., 76[5] S Nicholson, P. Sarkar, and X Huang, "Electrophoretic Deposition and Its Use 1249-5701993) to Synthesize ZrO2Al2O3 Micro Laminate Ceramic/Ceramic Composites, ".Mater. J -S. Ha, K. K. Chawla, and E. E. Engdahl, "Effect of Processing and Fiber Coating on Fibre Matrix Interactio SP. S. Nicholson, P. Sarkar, and X. Huang lly Strong and Tough ZrO2 Mater. Sci Eng, A, A161, 303-308(1993 Based Ceramics Composites >1300C by Electrophoretic i9X. Gu, P A Trusty, E. G. Butler, and C B. Ponton, "Deposition of Zirconia roceedings of Science and Technology of Zirconia V Edited by S.P. S. Badw on Woven Fibre Preforms Using a Dip-Coating Technique, "J. Eur. Ceram. Soc., 20, M. J. Bannister and R. H. j. Hannink. Technomic. Lancaster PA. 1993. 4P. Sarkar, X Huang, and P S. Nicholson, "Structural Ceramic Microlaminates by J. Wendorff, R Janssen, and N. Claussen, ""Platinum as a Weak Interphase for Electrophoretic Deposition, "J. Am. Ceram Soc., 75[10] 2907-909(1992) Fiber-Reinforced Oxide-Matrix Composites,J. Am. Ceram Soc., 81[1012738-40 C. Kaya, P. A. Trusty, and C. B. Ponton, "Preparation of Heterocoagulated Colloidal Suspensions for Fabrication of Ceramic Matrix Composites by Electro- K. Niva and K. Hashimoto, "Interfaces in Ceramic Substrates, "Scr:. Metall. phoretic Filtration Deposition, Br. Ceram. Trans., 81 [10]48-54(1998). arer,318100712(199) Carpenter, J. W. Bohlen, and w. S. Steffier, "Method of Forming a Ductile P. A. Trusty, A.R. BoccaccinL, E. G. Butler, and C B. Ponton, "Novel Techniques Fiber Coating for Toughening Non-oxide Ceramic Matrix Composites,"U.S.Pat. No 5162271,199 23F. F. Lange. B. V. Velamakanni, and A. G. Evans, "Method for Processing FD) of S affi Alumina F tbre -Reinforced M ullite Mopltilaver Nano -cera Metal-Reinforced Ceramic Composites, J Am Ceram Soc., 73 [2]388-93(1990 omposites,Pp. 399-406 in Ceramic Transactions, Vol. 83, Ceramic Wires,Annual Report, University Research Initiative Contract No. No0O14-86-K ence. Edited by G. L. Messing, F. F. Lange, and S.-l. Hirano. Ameri 0753 in The Processing an Society. Westerville, OH, 1998 Performance Composites, Book 5. Edited by A G. Evans an tion of ni R. A.J. Sambell, D. H. Bowen, and D. C. Philips, "Carbon Fiber Composites Colloidal Suspensions for the Preparation of Multilayer Alumina Fiber-Reinforced with Ceramic and Glass Matrices, Part II: Continuous Fibers,J. Mater. Sci, 7 Mullite CMCs Using Elect tic Filtration Deposition(EFD), pp 93-102 in 676-81(1972) British Ceramic Proceedings, Vol. 56, Better Ceramics through Processing. Edited by J. Binner and J. Yeomans. Institute of Materials. Cirencester. U.K.1998 li-Visconti and G. A Cooper, "Mechanical Properties of a New Carbon erial, "Nature(London), 221, 754-55(1969) 9C. Kaya, A.R. Boccaccini, and P. A. Trusty, "Processing and Characterisation of K.M. Prew,“ACo High Failure Strain, Fiber-Reinforced Glass-Matrn Woven Metal Fibre-reinforced Multilayer Silica Matrix Composites Using Composite, J. Mater. Sci., 17, 3549-63(1982 rophoretic Deposition and Pressure Filtration, "J. Eur. Ceram. Soc., 19, thesis of Spherical Sub-micron Borosilicate Powders, J. Mater.Sai,14,971-74(1995) "Z. Zhang, Y. Huang, and Z. Jiang, "Electrophoretic Deposition Forming of 2C. Kaya, A R. Boccaccini, and K. K. Chawla, unpublished work. SiC-TZP Composites in a Nonaqueous Media, " J. Am. Cera Soc., 77[7] 1946-49 P. E. D. Morgan and D. B. Marshall, "Ceramic Composites of Monazite and (1994) Alumina, "J Am Ceram Soc., 78 6]1553-63(1995)
metallic interface, which has been explored in this work, should be applicable to other oxide-fiber–oxide-matrix systems. Acknowledgments Professors M. H. Loretto and I. R. Harris are acknowledged for the provision of laboratory facilities at the IRC and School of Metallurgy and Materials and at the University of Birmingham, respectively. The authors also thank Dr. Less Renny (Inco Speciality Powder Products) for supplying the fibers. References 1 R. L. Lehman, S. K. El-Rahaiby, and J. B. Wachtman Jr. (eds.), Handbook on Continuous Fiber-Reinforced Ceramic Matrix Composites. American Ceramic Society, Westerville, OH, 1995. 2 P. S. Nicholson, P. Sarkar, and X. Huang, “Electrophoretic Deposition and Its Use to Synthesize ZrO2/Al2O3 Micro Laminate Ceramic/Ceramic Composites,” J. Mater. Sci., 28, 6274–78 (1993). 3 P. S. Nicholson, P. Sarkar, and X. Huang, “Potentially Strong and Tough ZrO2 Based Ceramics Composites .1300°C by Electrophoretic Deposition”; p. 503 in Proceedings of Science and Technology of Zirconia V. Edited by S. P. S. Badwal, M. J. Bannister, and R. H. J. Hannink. Technomic, Lancaster, PA, 1993. 4 P. Sarkar, X. Huang, and P. S. Nicholson, “Structural Ceramic Microlaminates by Electrophoretic Deposition,” J. Am. Ceram. Soc., 75 [10] 2907–909 (1992). 5 C. Kaya, P. A. Trusty, and C. B. Ponton, “Preparation of Heterocoagulated Colloidal Suspensions for Fabrication of Ceramic Matrix Composites by Electrophoretic Filtration Deposition,” Br. Ceram. Trans., 81 [10] 48–54 (1998). 6 P. A. Trusty, A. R. Boccaccini, E. G. Butler, and C. B. Ponton, “Novel Techniques for Manufacturing Woven Fiber Reinforced Ceramic Matrix Composites, I. Preform Fabrication,” Mater. Manuf. Processes, 10, 1215–26 (1995). 7 C. Kaya, P. A. Trusty, and C. B. Ponton, “Electrophoretic Filtration Deposition (EFD) of Saffil Alumina Fibre-Reinforced Mullite Multilayer Nano-ceramic Matrix Composites”; pp. 399–406 in Ceramic Transactions, Vol. 83, Ceramic Processing Science. Edited by G. L. Messing, F. F. Lange, and S.-I. Hirano. American Ceramic Society, Westerville, OH, 1998. 8 C. Kaya, P. A. Trusty, and C. B. Ponton, “Characterisation of Nano-sized Colloidal Suspensions for the Preparation of Multilayer Alumina Fiber-Reinforced Mullite CMCs Using Electrophoretic Filtration Deposition (EFD)”; pp. 93–102 in British Ceramic Proceedings, Vol. 56, Better Ceramics through Processing. Edited by J. Binner and J. Yeomans. Institute of Materials, Cirencester, U.K., 1998. 9 C. Kaya, A. R. Boccaccini, and P. A. Trusty, “Processing and Characterisation of 2-D Woven Metal Fibre-reinforced Multilayer Silica Matrix Composites Using Electrophoretic Deposition and Pressure Filtration,” J. Eur. Ceram. Soc., 19, 2859–66 (1999). 10Z. Zhang, Y. Huang, and Z. Jiang, “Electrophoretic Deposition Forming of SiC-TZP Composites in a Nonaqueous Media,” J. Am. Ceram. Soc., 77 [7] 1946–49 (1994). 11Y. Hirata, A. Nishimoto, and Y. Ishihara, “Forming of Alumina Powder by Electrophoretic Deposition,” Nippon Seramikkusu Kyokai Gakujutusu Ronbunshi, 99 [2] 108–13 (1991). 12F. Harbach and H. Nienburg, “Homogeneous Functional Ceramic Components through Electrophoretic Deposition from Stable Colloidal Suspensions—I. Basic Concepts and Application to Zirconia,” J. Eur. Ceram. Soc., 18, 675–83 (1998). 13A. R. Boccaccini and P. A. Trusty, “Electrophoretic Deposition Infiltration of Metallic Fabrics with a Boehmite Sol for the Preparation of Ductile-Phase-Toughened Ceramic Composites,” J. Mater. Sci., 33, 933–38 (1998). 14P. Sarkar and P. S. Nicholson, “Electrophoretic Deposition (EPD): Mechanisms, Kinetics, and Applications to Ceramics,” J. Am. Ceram. Soc., 79 [8] 1987–2002 (1996). 15K. K. Chawla, Ceramic Matrix Composites. Chapman and Hall, London, U.K., 1993. 16B. W. Sheldon, E. Y. Sun, S. R. Nutt, and J. J. Brennan, “Oxidation of BN-Coated SiC Fibers in Ceramic Matrix Composites,” J. Am. Ceram. Soc., 79 [2] 539–543 (1996). 17J. B. Davis, J. P. A. Lo¨fvander, A. G. Evans, and E. Bischoff, and M. L. Emiliani, “Fiber Coating Concepts for Brittle-Matrix Composites,” J. Am. Ceram. Soc., 76 [5] 1249–57 (1993). 18J.-S. Ha, K. K. Chawla, and E. E. Engdahl, “Effect of Processing and Fiber Coating on Fibre Matrix Interaction in Mullite Fibre–Mullite Matrix Composites,” Mater. Sci Eng., A, A161, 303–308 (1993). 19X. Gu, P. A. Trusty, E. G. Butler, and C. B. Ponton, “Deposition of Zirconia Sols on Woven Fibre Preforms Using a Dip-Coating Technique,” J. Eur. Ceram. Soc., 20, 675–84 (2000). 20J. Wendorff, R. Janssen, and N. Claussen, “Platinum as a Weak Interphase for Fiber-Reinforced Oxide-Matrix Composites,” J. Am. Ceram. Soc., 81 [10] 2738–40 (1998). 21K. Niva and K. Hashimoto, “Interfaces in Ceramic Substrates,” Scr. Metall. Mater., 31 [8] 1007–12 (1994). 22H. W. Carpenter, J. W. Bohlen, and W. S. Steffier, “Method of Forming a Ductile Fiber Coating for Toughening Non-oxide Ceramic Matrix Composites,” U.S. Pat. No. 5 162 271, 1992. 23F. F. Lange, B. V. Velamakanni, and A. G. Evans, “Method for Processing Metal-Reinforced Ceramic Composites,” J. Am. Ceram. Soc., 73 [2] 388–93 (1990). 24J. Eberhardt and M. F. Ashby, “Flow Characteristics of Highly Constrained Metal Wires,” Annual Report, University Research Initiative Contract No. N00014-86-K- 0753 in The Processing and Mechanical Properties of High-Temperature/HighPerformance Composites, Book 5. Edited by A. G. Evans and R. Mehrabian, 1988. 25R. A. J. Sambell, D. H. Bowen, and D. C. Philips, “Carbon Fiber Composites with Ceramic and Glass Matrices, Part II: Continuous Fibers,” J. Mater. Sci., 7, 676–81 (1972). 26I. Crivelli-Visconti and G. A Cooper, “Mechanical Properties of a New Carbon Fiber Material,” Nature (London), 221, 754–55 (1969). 27K. M. Prewo, “A Compliant, High Failure Strain, Fiber-Reinforced Glass-Matrix Composite,” J. Mater. Sci., 17, 3549–63 (1982). 28M. A. Harmer, “Synthesis of Spherical Sub-micron Borosilicate Powders,” J. Mater. Sci., 14, 971–74 (1995). 29C. Kaya, A. R. Boccaccini, and K. K. Chawla; unpublished work. 30P. E. D. Morgan and D. B. Marshall, “Ceramic Composites of Monazite and Alumina,” J. Am. Ceram. Soc., 78 [6] 1553–63 (1995). M 1888 Journal of the American Ceramic Society—Kaya et al. Vol. 83, No. 8