ournal JAm. Ceram.Soc,86[2]35-32(2003) Effectiveness of Monazite Coatings in Oxide/Oxide Composites after Long-Term Exposure at High Temperature ristin A. Keller, * f Tai-lI Mah, *, f Triplicane A. Parthasarathy, f Emmanuel E. Boakye,*, T Pavel Mogilevsky * f and Michael K. Cinibulk* Air Force Research Laboratory, Materials and Manufacturing Directorate, AFRL/MLLN Wright-Patterson Air Force Base, Ohio 45433 The effectiveness of monazite (LaPO4) in providing an crack deflection, even though crack deflection would have oc xidation-resistant weak fiber/matrix interface was evaluated curred had the fibers been sound. Related work on the develop- in a fiber roving/thin coating/ceramic-matrix composite witl >20% fiber volume fraction. Nextelmo 610/ monazite/alumina be severely degraded by processing if not properly controlled composites were fabricated and tensile tested after isothermal The second complicating point is related to matrix density. It is exposures of up to 1000 h. Some strength loss was seen after well-established that crack deflection around fibers occurs in short-term exposures(11000-1200oC/5-250 h); however, no uitably porous-matrix composites without the benefit or need of a further loss was observed after 1000 h at 1200C. Conversely fiber coating. 5-27 This introduces the possibility of a false positive control samples containing uncoated fiber displayed >70% result if there is significant matrix porosity by confusing the source strength losses after only 5 h at 1200oC. Fiber pullout was seen of the crack deflection: the coating or the matrix porosity in monazite-containing samples even after 1000 h at 1200oC. Obtaining a dense crystalline ceramic-matrix(90%)composite Debonding was predominantly in the coating or at either the however, has been problematic due to the constraints imposed by fiber/coating or coating/matrix interface Push-out testing con the fibers(e. g, see Sudre and Lange"). Additionally, the tempe firmed the weakness of the monazite coating interface. ature limitations of small diameter oxide fiber tows(e. g, Nextelmg M Corp, Minneapolis, MN) limit processing temperatures to below those necessary to attain high-density crystalline ceram- ics. Unfortunately, glassy matrixes that might at least provide a L. Introduction model material wet the monazite grain boundaries, invade the T Is well-known that ceramic-matrix composites require crack coating, and disrupt coating continuity deflectio or before reaching) the fiber/matrix interface to Over-temperature degradation of coatingless porous-matrix prevent brittle failure. Current technology composites rely or carbon, "boron nitride, , or porous matrixes' for crack deflec- tion CMC lives are limited in the two former cases by oxidation",9 bonding of the matrix to the fiber so that the crack deflection mechanism fails to operate before fiber fracture. 0, 3 In cases and in the latter by matrix densification and bonding to the fiber. where the second mechanism alone operates, there is an opportu The lifetime limitations associated with crack deflection mecha- nity to obviate degradation by use of a properly functioning nisms have been a major impediment to the adoption of ceramic coating and to clearly evaluate the efficacy of the coating. composites as engineering materials. The goal of this work was to definitively determine the efficacy Alternative approaches based on oxidation-resistant interfacial coatings have been pursued for some time. ,12 Monazite(LaPO) of monazite coatings by using control specimens to account for process effects on fibers and to provide direct comparisor is arter has been the most thoroughly investigated of these approaches. nominal over-temperature exposures. Alumina tow-reinforced alu- Morgan and Marshall first demonstrated the potential of monazite mina matrix composites, with and without monazite fiber coatings for promoting crack deflection in 1995. Monazite has since been were compared after heat treatment at temperatures that were examined in a variety of forms, and there is good evidence that it excessive for coatingless porous-matrix composite performance can provide the necessary weak interface between fiber and Such heat treatments allowed for matrix consolidation and matrix.--However, due to processing difficulties, its definitive bonding to the fiber in the coatingless control specimens, which demonstration in the prototypical concept of an engineered- resulted in severely degraded properties. As is seen in detail, the interface, high-fiber-fraction ceramic composite(fiber roving/thin monazite fiber coatings were fully functional in providing crack coating/ceramic matrix) has been elusive, with the first stage of deflection even after much longer exposures to the over this work being a rare exception. Two phenomena greatly complicate the evaluation of fiber/ coating performance. The first of these is fiber degradation, can lead to a false negative indication for coating effectivenes Il. Experimental Procedure to premature failure of the fiber.Catastrophic composite consequently occurs at stress levels below those necessary for Nextel 610 fiber tow was desized at 1000%C and then coated five times with a monazite precursor solution. As described in previous work, this sol was washed several times to remove any sidual ions which have been associated with fiber strength loss after coating. Two separate batches of monazite-coated Nextel F buting editor oduced. After coating, the tow was filament wound using a slurry containing alumina powder(AKP-53, Sumitomo Inc, Ashland, MA) boehmite sol as a binder. during the October 1. 2002. winding process, the number of wheel revolutions was recorded to determine the number of fiber tows within a given tape width 325
Effectiveness of Monazite Coatings in Oxide/Oxide Composites after Long-Term Exposure at High Temperature Kristin A. Keller,* ,† Tai-Il Mah,* ,† Triplicane A. Parthasarathy,* ,† Emmanuel E. Boakye,* ,† Pavel Mogilevsky,* ,† and Michael K. Cinibulk* Air Force Research Laboratory, Materials and Manufacturing Directorate, AFRL/MLLN, Wright–Patterson Air Force Base, Ohio 45433 The effectiveness of monazite (LaPO4) in providing an oxidation-resistant weak fiber/matrix interface was evaluated in a fiber roving/thin coating/ceramic-matrix composite with >20% fiber volume fraction. Nextel 610/monazite/alumina composites were fabricated and tensile tested after isothermal exposures of up to 1000 h. Some strength loss was seen after short-term exposures (1100°–1200°C/5–250 h); however, no further loss was observed after 1000 h at 1200°C. Conversely, control samples containing uncoated fiber displayed >70% strength losses after only 5 h at 1200°C. Fiber pullout was seen in monazite-containing samples even after 1000 h at 1200°C. Debonding was predominantly in the coating or at either the fiber/coating or coating/matrix interface. Push-out testing confirmed the weakness of the monazite coating interface. I. Introduction I T IS well-known that ceramic-matrix composites require crack deflection at (or before reaching) the fiber/matrix interface to prevent brittle failure.1,2 Current technology composites rely on carbon,3,4 boron nitride,5,6 or porous matrixes7 for crack deflection. CMC lives are limited in the two former cases by oxidation8,9 and in the latter by matrix densification and bonding to the fiber.10 The lifetime limitations associated with crack deflection mechanisms have been a major impediment to the adoption of ceramic composites as engineering materials. Alternative approaches based on oxidation-resistant interfacial coatings have been pursued for some time.11,12 Monazite (LaPO4) has been the most thoroughly investigated of these approaches. Morgan and Marshall first demonstrated the potential of monazite for promoting crack deflection in 1995.13 Monazite has since been examined in a variety of forms, and there is good evidence that it can provide the necessary weak interface between fiber and matrix.14–21 However, due to processing difficulties, its definitive demonstration in the prototypical concept of an engineeredinterface, high-fiber-fraction ceramic composite (fiber roving/thin coating/ceramic matrix) has been elusive, with the first stage of this work being a rare exception.22 Two phenomena greatly complicate the evaluation of fiber/ coating performance. The first of these is fiber degradation, which can lead to a false negative indication for coating effectiveness due to premature failure of the fiber.23 Catastrophic composite failure consequently occurs at stress levels below those necessary for crack deflection, even though crack deflection would have occurred had the fibers been sound. Related work on the development of fiber coatings has clearly illustrated that fiber strength can be severely degraded by processing if not properly controlled.24 The second complicating point is related to matrix density. It is well-established that crack deflection around fibers occurs in suitably porous-matrix composites without the benefit or need of a fiber coating.25–27 This introduces the possibility of a false positive result if there is significant matrix porosity by confusing the source of the crack deflection: the coating or the matrix porosity. Obtaining a dense crystalline ceramic-matrix (90%) composite, however, has been problematic due to the constraints imposed by the fibers (e.g., see Sudre and Lange28). Additionally, the temperature limitations of small diameter oxide fiber tows (e.g., Nextel, 3M Corp., Minneapolis, MN) limit processing temperatures to below those necessary to attain high-density crystalline ceramics.29 Unfortunately, glassy matrixes that might at least provide a model material wet the monazite grain boundaries, invade the coating, and disrupt coating continuity.30 Over-temperature degradation of coatingless porous-matrix composites occurs via one or two mechanisms: (i) fiber degradation and (ii) some combination of matrix consolidation and strong bonding of the matrix to the fiber so that the crack deflection mechanism fails to operate before fiber fracture.10,31 In cases where the second mechanism alone operates, there is an opportunity to obviate degradation by use of a properly functioning coating and to clearly evaluate the efficacy of the coating. The goal of this work was to definitively determine the efficacy of monazite coatings by using control specimens to account for process effects on fibers and to provide direct comparisons after nominal over-temperature exposures. Alumina tow-reinforced alumina matrix composites, with and without monazite fiber coatings, were compared after heat treatment at temperatures that were excessive for coatingless porous-matrix composite performance. Such heat treatments allowed for matrix consolidation and bonding to the fiber in the coatingless control specimens, which resulted in severely degraded properties. As is seen in detail, the monazite fiber coatings were fully functional in providing crack deflection even after much longer exposures to the overtemperature conditions. II. Experimental Procedure Nextel 610 fiber tow was desized at 1000°C and then coated five times with a monazite precursor solution. As described in previous work, this sol was washed several times to remove any residual ions, which have been associated with fiber strength loss after coating.32 Two separate batches of monazite-coated Nextel 610 were produced. After coating, the tow was filament wound using a slurry containing alumina powder (AKP-53, Sumitomo Corp., Tokyo, Japan) and Nyacol (Nyacol Nanotechnologies, Inc., Ashland, MA) boehmite sol as a binder. During the winding process, the number of wheel revolutions was recorded to determine the number of fiber tows within a given tape width. F. W. Zok—contributing editor Manuscript No. 187261. Received December 13, 2001; approved October 1, 2002. *Member, American Ceramic Society. † UES, Inc., Dayton, OH, under USAF Contract No. F33615-01-C-5214. J. Am. Ceram. Soc., 86 [2] 325–32 (2003) 325 journal
Journal of the American Ceramic Sociery-Keller et al. Vol 86. No. 2 In turn, this value was used in calculating the volume percent Il. Results and Discussion age of fibers in the composite. The resultant tapes were cut from the spool, stacked into a mold(0% orientation), and then warm Monazite coatings on the Nextel 610 fibers displayed a"worm- vacuum-molded to form green unidirectional composites. After like"structure with open porosity in the coatings(Fig. 1).32TEM consolidation, the samples were kept in the mold and trans- xamination of the coatings within the composites indicated that ferred to a humidity chamber for further drying. Control he coatings were monazite; no other phase was detected at the mples containing uncoated fiber were produced using the interface, at least within the resolution of the microscope. Samples ame procedure produced using the monazite -coated Nextel 610 fiber exhibited After drying, the first set of Nextel 610/monazite/alumina bulk densities of -2.6 g/cm, with -35 vol% composite porosity samples were subjected to baseline sintering at 1200C/5 h/air, with the exception of one billet that was sintered at 1.C/5 h/air. The fibers were relatively well-dispersed within the matrix. As Control samples were sintered under the same conditions. Portions generally seen in porous-matrix composites, microcracking was of the monazite-coated samples were also subjected to an add- e\anm were et of composites, both monazite-containing and due to both drying and sintering shrinkage tional 100 h heat treatment at 1200 C in air(105 h total ). A second control. were evaluated after heat treatment at 1%C/5 h and set of Nextel 610/monazite/alumina composites was processed by 200%C/5 h. The tensile test results are shown in Fig. 2: numbers sintering for 5 h at 1200oC in air. Specimens from this batch wer in italics indicate the average percentage of failure strain endured then heat-treated for either 250 or 1000 h at 1200%c to evaluate by the samples. Both the control and monazite-containing samples long-term effects. The two sets of monazite-coated samples were displayed reasonably good strengths after the 1100 C sintering. No nsidered independently because of differences in baseline com- strain information was recorded for these samples. The measured ite strengths average tow strength of the Nextel 610 was 1. 47 GPa after or a baseline in evaluating composite properties, Nextel 610 exposure at 1100 C/5 h(Table I). If it is assumed that the fibe fiber tow strength was measured by tensile testing after heat carried the entire load, then the predicted strength of the uncoated treatments simulating those of the composites. Both uncoated and composites, sintered at 1100C/5 h, would be -290 MPa for a monazite-coated tows were exposed at 1.C for 5 h and at 20-vol%-fiber composite The predicted strength of the monazite- 1200 C for 5, 100, 250, and 1000 h Tows were mounted onto thin containing samples would be "270 MPa based on the average tow cardboard frames using epoxy, and a small amount of deionized strength of 1.36 GPa. Approximately 80% of this strength value water was used as a sizing to aid in aligning the fibers within the was obtained in the monazite-containing samples(-220 MPa) tow. The mounted tows were loaded into a test frame(Model while the uncoated samples displayed only -59% of the predicted Synergie, MTS Systems Corp, Eden Prairie, MN) with a 500N value(168 MPa) load cell and then tested at a crosshead speed of 0.0127 cm/min The lower strength of the uncoated samples indicated possible matrix consolidation and excessive fiber/matrix bonding even at Between 15 and 30 tests were completed for each exposure t. this relatively low temperature. For the monazite-containing sam- An average fiber diameter of 12 um was used in evaluating tow trengths, this diameter was also used in determining the volume les, the shortfall in strength might be attributed to a loss of fiber fraction of fibers during multiple monazite coating runs and subsequent composite Straight-sided specimens were sliced from the sintered billets processing and/or to discontinuities in the monazite coating that tabbed, and then tensile tested using a test frame( Model Sintechac emitted isolated fiber/matrix interaction MTS Corp )and a digital strain measuring system(Model Vic- After the 1200C/5 h heat treatment, there was a dramatic Gagen, Correlated Solutions, Inc, West Columbia, SC). Between difference in the strength of the samples. The uncoated sample three and Six specimens were tested for each condition. For retained only -27%of the 1 100C strength(45 MPa), whereas the comparison purposes, fiber volume percentage variations between monazite-containing composite retained -89% of its strengt billets were eliminated by normalizing the tensile test data to iving an average strength of 198 MPa. The uncoated sample howed minimal failure strain, -0.03%, while the monazite 20-vol %-fiber composite. Fracture surfaces were evaluated using SEM(Model Leica Cambridge Stereoscan 360FE, LEO Micros containing sample exhibited an average of 0. 27% failure strain. No further heat treatments or tests were completed on the contro copy, Inc, Cambridge, U. K) samples since substantial degradation occurred after only 5 h at Finally, TEM(Model CM200FEG, Philips Corp, Eindhoven, 1200°C The Netherlands)analysis of the samples was used to determine The fracture surfaces of both the control and monazite samples the effects of time at temperature on interfacial stability. Push-in heat-treated at 1 100C displayed "woody -type failure, with a signif- and push-out tests(Model Micro Measure Machine, Process icant amount of fiber/bundle pullout. After 1200.,however, the Equipment Co., Tipp City, OH) were conducted using a 5 um, control samples showed brittle-type fracture surfaces with no fiber flat-tipped diamond indenter on selected samples to obtain infor allout(Fig 3(a)). This was consistent with the low strength and mation on interfacial properties with the probable strong fiber/matrix bonding. In contrast, the 500nm Fig. 1. Monazite coatings on Nextel fibers
In turn, this value was used in calculating the volume percentage of fibers in the composite. The resultant tapes were cut from the spool, stacked into a mold (0° orientation), and then warm vacuum-molded to form green unidirectional composites. After consolidation, the samples were kept in the mold and transferred to a humidity chamber for further drying. Control samples containing uncoated fiber were produced using the same procedure. After drying, the first set of Nextel 610/monazite/alumina samples were subjected to baseline sintering at 1200°C/5 h/air, with the exception of one billet that was sintered at 1100°C/5 h/air. Control samples were sintered under the same conditions. Portions of the monazite-coated samples were also subjected to an additional 100 h heat treatment at 1200°C in air (105 h total). A second set of Nextel 610/monazite/alumina composites was processed by sintering for 5 h at 1200°C in air. Specimens from this batch were then heat-treated for either 250 or 1000 h at 1200°C to evaluate long-term effects. The two sets of monazite-coated samples were considered independently because of differences in baseline composite strengths. For a baseline in evaluating composite properties, Nextel 610 fiber tow strength was measured by tensile testing after heat treatments simulating those of the composites. Both uncoated and monazite-coated tows were exposed at 1100°C for 5 h and at 1200°C for 5, 100, 250, and 1000 h. Tows were mounted onto thin cardboard frames using epoxy, and a small amount of deionized water was used as a sizing to aid in aligning the fibers within the tow. The mounted tows were loaded into a test frame (Model Synergie, MTS Systems Corp., Eden Prairie, MN) with a 500 N load cell and then tested at a crosshead speed of 0.0127 cm/min. Between 15 and 30 tests were completed for each exposure time. An average fiber diameter of 12 m was used in evaluating tow strengths;33 this diameter was also used in determining the volume fraction of fibers. Straight-sided specimens were sliced from the sintered billets, tabbed, and then tensile tested using a test frame (Model Sintech, MTS Corp.) and a digital strain measuring system (Model VicGage, Correlated Solutions, Inc., West Columbia, SC). Between three and six specimens were tested for each condition. For comparison purposes, fiber volume percentage variations between billets were eliminated by normalizing the tensile test data to a 20-vol%-fiber composite. Fracture surfaces were evaluated using SEM (Model Leica Cambridge Stereoscan 360FE, LEO Microscopy, Inc., Cambridge, U.K.). Finally, TEM (Model CM200FEG, Philips Corp., Eindhoven, The Netherlands) analysis of the samples was used to determine the effects of time at temperature on interfacial stability. Push-in and push-out tests (Model Micro Measure Machine, Process Equipment Co., Tipp City, OH) were conducted using a 5 m, flat-tipped diamond indenter on selected samples to obtain information on interfacial properties. III. Results and Discussion Monazite coatings on the Nextel 610 fibers displayed a “wormlike” structure with open porosity in the coatings (Fig. 1).32 TEM examination of the coatings within the composites indicated that the coatings were monazite; no other phase was detected at the interface, at least within the resolution of the microscope. Samples produced using the monazite-coated Nextel 610 fiber exhibited bulk densities of 2.6 g/cm3 , with 35 vol% composite porosity and 45–50 vol% matrix porosity (including large microcracks). The fibers were relatively well-dispersed within the matrix. As generally seen in porous-matrix composites, microcracking was evident due to both drying and sintering shrinkage. The first set of composites, both monazite-containing and control, were evaluated after heat treatment at 1100°C/5 h and 1200°C/5 h. The tensile test results are shown in Fig. 2; numbers in italics indicate the average percentage of failure strain endured by the samples. Both the control and monazite-containing samples displayed reasonably good strengths after the 1100°C sintering. No strain information was recorded for these samples. The measured average tow strength of the Nextel 610 was 1.47 GPa after exposure at 1100°C/5 h (Table I). If it is assumed that the fiber carried the entire load, then the predicted strength of the uncoated composites, sintered at 1100°C/5 h, would be 290 MPa for a 20-vol%-fiber composite. The predicted strength of the monazitecontaining samples would be 270 MPa based on the average tow strength of 1.36 GPa. Approximately 80% of this strength value was obtained in the monazite-containing samples (220 MPa), while the uncoated samples displayed only 59% of the predicted value (168 MPa). The lower strength of the uncoated samples indicated possible matrix consolidation and excessive fiber/matrix bonding even at this relatively low temperature. For the monazite-containing samples, the shortfall in strength might be attributed to a loss of fiber during multiple monazite coating runs and subsequent composite processing and/or to discontinuities in the monazite coating that permitted isolated fiber/matrix interaction. After the 1200°C/5 h heat treatment, there was a dramatic difference in the strength of the samples. The uncoated sample retained only 27% of the 1100°C strength (45 MPa), whereas the monazite-containing composite retained 89% of its strength, giving an average strength of 198 MPa. The uncoated sample showed minimal failure strain, 0.03%, while the monazitecontaining sample exhibited an average of 0.27% failure strain. No further heat treatments or tests were completed on the control samples since substantial degradation occurred after only 5 h at 1200°C. The fracture surfaces of both the control and monazite samples heat-treated at 1100°C displayed “woody”-type failure, with a significant amount of fiber/bundle pullout. After 1200°C, however, the control samples showed brittle-type fracture surfaces with no fiber pullout (Fig. 3(a)). This was consistent with the low strength and with the probable strong fiber/matrix bonding. In contrast, the Fig. 1. Monazite coatings on Nextel fibers. 326 Journal of the American Ceramic Society—Keller et al. Vol. 86, No. 2
February 2003 Monazite Coatings in Oxide/Oxide Composites after Long-Term Exposure at High Tempera 327 N610/Alumina N610/Monazite/Alumina monazite-containing samples showed the same fibrous fracture 250 surface seen after the 1 100 C heat treatment(Fig 3(b) Monazite-containing billets from this first batch of samples were also heat-treated for 100 h at 1200%c in air. The tensile 品 0.27% strengths of these samples were reduced to an average of 143 MPa with 0. 22% strain (Fig. 2). The fracture surfaces of these composites still appeared"woody, with significant fiber pullout 0.22% Fig. 4). SEM analysis revealed that both pulled-out fibers and 万 matrix troughs had regions with and without monazite coating. The location of the first interfacial crack deflection was not directly evident. The presence of monazite on both the fibers and the matrix troughs indicated that debond cracks were predominantly at the coating surface(coating/fiber or coating/matrix) or within the 0.03% coating rather than in the matrix Although the results of the 100 h heat treatment at 1200 C were promising, longer-term heat treatments were desired to further estab- lish the stability of the monazite-containin batch of billets, therefore, was produced and subjected to extended 1100c1200c1200c heat treatments at 1200oC (5, 255, and 1005 h). For this discussion. these samples were separated from the first batch since the averag trength of these samples after 5 h at 1200.C was notably less than the ig. 2. Ultimate strengths for Nextel 610/alumina tes with and first batch(158 versus 198 MPa) for reasons that are unclear at this without monazite coating Strengths normalized to 2 fiber compos- time average de sity of this second batch of samples was slightly ite. Values in italics are average percentage of strains. lower. at -2.4-2.5 g/cm. however we do not believe that this lower density should have affected composite strength so dramatically. The Table l. Average Measured Tow Strengths for cause of the lower strength is currently under consideration and may Nextel 610 be related to insufficient coverage of the monazite coating on the fibers. althe the strengths of these samples were lower, the Strength Weibull Temperature rimary goal of this work was to compare the short-and long-term (GPa modulus Behaviors of these composites and, consequently, determine the 1.51 effectiveness of monazite in retaining composite strength. In that 12.7 sense, the high values of the absolute strengths of the composites were 1100° not essential Figure 5 shows the tensile test results fo second batch of Nextel 610/monazite/alumina composites. After 5 h at 1200 C, these samples were still dramatically better than the control specimens Monazite 17.3 After 250 h at 1200oC, almost 65% of composite strength was retained, with an average strength of -102 MPa. Most significantly, coated the average strength of the composites after heat treatment for 1000 h at 1200oC was essentially the same at -98 MPa. Strain-to-failure 1200°C emained at -0 1% from 5 to 1000 h. The fracture surfaces of these Uncoated specimens displayed fiber/bundle pullout, even after the 1000 h heat Monazite 1.03 treatment. Figure 6 provides both low and high magnification view coated 0.58′ of fractured samples after heat treatment at 1200C for 1000 h. Monazite 0.97 Monazite coatings appeared intermittently on both pulled-out fibers and in matrix troughs, again indicating a debonding fracture path that Strength decrease presumably due to fiber welding during heat was restricted to the coating region but meandered between surfaces As with samples heat-treated for shorter time periods, deformation of the monazite was observed, as shown in Figs. 6(c), (d). This type of deformation has been reported by other researchers and has been extensively studied by Hay la) 50 Soum Fig 3. Fracture surfaces of porous-matrix composites after heat treatment at 1200C/5 h/air(a) control sample surface and (b) monazite-coated samples)
monazite-containing samples showed the same fibrous fracture surface seen after the 1100°C heat treatment (Fig. 3(b)). Monazite-containing billets from this first batch of samples were also heat-treated for 100 h at 1200°C in air. The tensile strengths of these samples were reduced to an average of 143 MPa with 0.22% strain (Fig. 2). The fracture surfaces of these composites still appeared “woody,” with significant fiber pullout (Fig. 4). SEM analysis revealed that both pulled-out fibers and matrix troughs had regions with and without monazite coating. The location of the first interfacial crack deflection was not directly evident. The presence of monazite on both the fibers and the matrix troughs indicated that debond cracks were predominantly at the coating surface (coating/fiber or coating/matrix) or within the coating rather than in the matrix. Although the results of the 100 h heat treatment at 1200°C were promising, longer-term heat treatments were desired to further establish the stability of the monazite-containing composites. A second batch of billets, therefore, was produced and subjected to extended heat treatments at 1200°C (5, 255, and 1005 h). For this discussion, these samples were separated from the first batch since the average strength of these samples after 5 h at 1200°C was notably less than the first batch (158 versus 198 MPa) for reasons that are unclear at this time. The average density of this second batch of samples was slightly lower, at 2.4–2.5 g/cm3 ; however, we do not believe that this lower density should have affected composite strength so dramatically. The cause of the lower strength is currently under consideration and may be related to insufficient coverage of the monazite coating on the fibers. Although the strengths of these samples were lower, the primary goal of this work was to compare the short- and long-term behaviors of these composites and, consequently, determine the effectiveness of monazite in retaining composite strength. In that sense, the high values of the absolute strengths of the composites were not essential. Figure 5 shows the tensile test results for this second batch of Nextel 610/monazite/alumina composites. After 5 h at 1200°C, these samples were still dramatically better than the control specimens. After 250 h at 1200°C, almost 65% of composite strength was retained, with an average strength of 102 MPa. Most significantly, the average strength of the composites after heat treatment for 1000 h at 1200°C was essentially the same at 98 MPa. Strain-to-failure remained at 0.1% from 5 to 1000 h. The fracture surfaces of these specimens displayed fiber/bundle pullout, even after the 1000 h heat treatment. Figure 6 provides both low and high magnification views of fractured samples after heat treatment at 1200°C for 1000 h. Monazite coatings appeared intermittently on both pulled-out fibers and in matrix troughs, again indicating a debonding fracture path that was restricted to the coating region but meandered between surfaces. As with samples heat-treated for shorter time periods, deformation of the monazite was observed, as shown in Figs. 6(c),(d). This type of deformation has been reported by other researchers16 and has been extensively studied by Hay.34 Fig. 2. Ultimate strengths for Nextel 610/alumina composites with and without monazite coating. Strengths normalized to 20-vol%-fiber composite. Values in italics are average percentage of strains. Fig. 3. Fracture surfaces of porous-matrix composites after heat treatment at 1200°C/5 h/air ((a) control sample surface and (b) monazite-coated samples). Table I. Average Measured Tow Strengths for Nextel 610 Temperature Time (h) Strength (GPa) Weibull modulus As-received 1.51 7.6 As-coated 1.68 12.7 1100°C Uncoated 5 1.46 11.5 Monazite 5 1.36 10.1 1200°C Uncoated 5 1.25 8.8 Monazite 5 1.13 17.3 1200°C Uncoated 100 1.1 7.7 Monazite 100 1.0 14.5 1200°C Uncoated 250 0.9 11 Monazite 250 1.03 8 1200°C Uncoated 1000 0.58† 10.7 Monazite 1000 0.97 12.2 † Strength decrease presumably due to fiber welding during heat treatment. February 2003 Monazite Coatings in Oxide/Oxide Composites after Long-Term Exposure at High Temperature 327
328 Journal of the American Ceramic Sociery-Keller et al. Vol 86. No. 2 interface caused by the partial sintering of monazite. After 5 h at 1200C, debond energy was determined to be 5.5 J/m", this parameter decreased to 0.15 J/m2 for samples heat-treated for 100 and 1000 h. The friction coefficient was similar for all th ranging from 0. 12 to 0.15, but the clamping stresses drastically during longer heat treatment(114 MPa to 四 again presumably due to the generation of gaps from the partial sintering of monazite. The push-out sample of Nextel 610/monazite/alumina heat- treated for 1000 h fractured on removal from the mounting stage It broke in the region of the pushed fibers, and Fig. 8 provides a side view of one such fiber. (The pushed fiber is in the center of the photograph. As stated previously, the conical probe used to push the fibers possessed a 5 um flat tip; however, due to the cone angle, the fibers could only be pushed between 3 and 4 um before the probe contacted the matrix, It can be seen in the image that the fiber was pushed 5 um, with slight damage to the matrix. The Fig. 4. Backscattered SEM image fracture surface of Nextel fibers also were slightly raised over the matrix material due to 610/monazite/alumina sample heat-trea 1200C for 100 h/air. Mon- preferential polishing of the porous matrix. Portions of the mon- matrix troughs azite coating remained attached to the fiber, shown by areas in the backscattered image in Fig. 8. Partial cove remained in the matrix trough, as seen in the trough Push-in/push-out testing was completed on the m the pushed fiber(Fig 8) containing samples to evaluate interfacial properties. Control It is apparent that the addition of the monazite coatings samples were also tested. As expected, the uncoated fibers were significantly improved the lifetime and temperature capability of strongly bonded to the matrix so that they could not be debonded these composites. As stated initially, one reason for its effective in push out(Fig. 7(a)). Figures 7(b(d) give the push-in/push-out ness is the prevention of fiber/matrix bonding. A strongly bonded results for Nextel 610/monazite/alumina composites heat-treated at coating could provide this function, with crack deflection being 1200.C for 5, 100, and 1000 h. The solid lines in the figures provided by the porous matrix. However, in the case of monazite, represent the fit of the progressive roughness models to the fracture occurred in the interfacial region (i. e, at the fiber/coating, experimental data to extract interface parameters. In applying the coating/matrix interfaces or within the coating), with little fracture model, assumptions were made for the fiber volume fraction(0.2), occurring in the matrix. This indicates that the monazite was coating thickness(0. 15 um), elastic modulus of the matrix(50 indeed acting as a weak interface material and could be expected GPa), and Poisson's ratio(0. 23). From the best fit of the model, the to function in dense matrix composites 1200C/5 h monazite sample was determined to have an interfacial It is worth noting that Morgan and Marshallobserved that roughness of 25 nm, whereas the roughnesses of the 100 and crack deflection was favored at the monazite/fiber interface but not 1000 h heat-treated samples were determined to be 0.2 nm. The at the matrix/monazite interface. They also showed that this was former roughness amplitude was consistent with measurements by onsistent with the criterion of He and Hutchinson. 37In contrast, Marshall et al. o The lower roughness in annealed specimens wa the mode and location of the first interfacial crack deflection could believed to result from the formation of periodic gaps at the not be identified in this study, although fractography clearly w/o Coating w/Monazite Coatin g 1100C 1200C m s 80 5日40 1200C 三 5h250hl000h 000hat1200°C
Push-in/push-out testing was completed on the monazitecontaining samples to evaluate interfacial properties. Control samples were also tested. As expected, the uncoated fibers were strongly bonded to the matrix so that they could not be debonded in push out (Fig. 7(a)). Figures 7(b)–(d) give the push-in/push-out results for Nextel 610/monazite/alumina composites heat-treated at 1200°C for 5, 100, and 1000 h. The solid lines in the figures represent the fit of the progressive roughness model35 to the experimental data to extract interface parameters. In applying the model, assumptions were made for the fiber volume fraction (0.2), coating thickness (0.15 m), elastic modulus of the matrix (50 GPa), and Poisson’s ratio (0.23). From the best fit of the model, the 1200°C/5 h monazite sample was determined to have an interfacial roughness of 25 nm, whereas the roughnesses of the 100 and 1000 h heat-treated samples were determined to be 0.2 nm. The former roughness amplitude was consistent with measurements by Marshall et al.36 The lower roughness in annealed specimens was believed to result from the formation of periodic gaps at the interface caused by the partial sintering of monazite. After 5 h at 1200°C, debond energy was determined to be 5.5 J/m2 ; this parameter decreased to 0.15 J/m2 for samples heat-treated for 100 and 1000 h. The friction coefficient was similar for all the samples, ranging from 0.12 to 0.15, but the clamping stresses decreased drastically during longer heat treatment (114 MPa to 1 MPa), again presumably due to the generation of gaps from the partial sintering of monazite. The push-out sample of Nextel 610/monazite/alumina heattreated for 1000 h fractured on removal from the mounting stage. It broke in the region of the pushed fibers, and Fig. 8 provides a side view of one such fiber. (The pushed fiber is in the center of the photograph.) As stated previously, the conical probe used to push the fibers possessed a 5 m flat tip; however, due to the cone angle, the fibers could only be pushed between 3 and 4 m before the probe contacted the matrix. It can be seen in the image that the fiber was pushed 5 m, with slight damage to the matrix. The fibers also were slightly raised over the matrix material due to preferential polishing of the porous matrix. Portions of the monazite coating remained attached to the fiber, shown by the bright areas in the backscattered image in Fig. 8. Partial coverage also remained in the matrix trough, as seen in the trough to the left of the pushed fiber (Fig. 8). It is apparent that the addition of the monazite coatings significantly improved the lifetime and temperature capability of these composites. As stated initially, one reason for its effectiveness is the prevention of fiber/matrix bonding. A strongly bonded coating could provide this function, with crack deflection being provided by the porous matrix. However, in the case of monazite, fracture occurred in the interfacial region (i.e., at the fiber/coating, coating/matrix interfaces or within the coating), with little fracture occurring in the matrix. This indicates that the monazite was indeed acting as a weak interface material and could be expected to function in dense matrix composites. It is worth noting that Morgan and Marshall13 observed that crack deflection was favored at the monazite/fiber interface but not at the matrix/monazite interface. They also showed that this was consistent with the criterion of He and Hutchinson.37 In contrast, the mode and location of the first interfacial crack deflection could not be identified in this study, although fractography clearly Fig. 4. Backscattered SEM image showing fracture surface of Nextel 610/monazite/alumina sample heat-treated at 1200°C for 100 h/air. Monazite is evident on both pulled-out fibers and in matrix troughs. Fig. 5. Percentage of strength retention of Nextel 610/alumina and Nextel 610/monazite/alumina composites. The monazite-containing samples exhibited 0.1% strain, while the control specimens displayed only 0.03% failure strain. Monazite-containing samples retained a majority of strength even after 1000 h at 1200°C. 328 Journal of the American Ceramic Society—Keller et al. Vol. 86, No. 2
February 2003 Monazite Coatings in Oxide/Oxide Composites after Long-Term Exposure at High Temperature 329 (b) um mm Deformed Monazite um Undeformed Monazite Deformed Monazite Fig. 6. SEM images of Nextel 610/ monazite/alumina samples heat-treated at 1200C/1000 h/air(a) low-magnification view of broken test sample, (b) higher-magnification view of fracture surface showing fiber pullout(BSD),(c)deformed monazite attached to fiber as been pulled from the matrix (BSD), and(d) deformed monazite within matrix trough). showed that extensive crack deflection occurred. It is possible that overall shape, however, grain growth and pore coarsening oc the crack deflection occurred at the fiber/monazite interface and curred within the coating. Porosity(white regions) could be seen then began to alternate between the two weak interfaces. It is also within the heat-treated coating, at the coating/matrix interface, and worth pointing out that the matrix in this work was porous and at the coating/fiber interface. The densification of monazite led to much weaker in modulus and strength than the matrix used in the the formation of isolated"gaps""in the interfacial region. Previous work of Morgan and marshall work in other systems has demonstrated that"gaps"at the interface The results of push-out testing revealed that the 100 and 1000 h can be beneficial to composite behavior. samples exhibited extremely low debond energies and low sliding The drop in ultimate strength of the Nextel 610/monazite/ friction, which indicated that the monazite was, at a minimum alumina composite with long-term heat treatment can be rational artially debonded in the san This debond ized as follows. The drop in friction stress from 17 MPa(0.15 x occurred during sample preparation and/or from the densification/ 114) after 5 h at 1200 C down to 0. 11 MPa (0. 12 X 0.9)in spheroidization of the coating within the composite. Figure 9 100/1000 h causes an increase in the effective gauge length of the shows a monazite-coated Nextel 610 fiber from the fracture fiber in the composite, as per the model of Curtin. The effective urface of a sample heat-treated for 1000 h at 1200C. The coating gauge length increases from nearly 0.55 mm to the composit itself appeared to have sintered during thermal exposure. Figure gauge length of 25.4 mm. As per the model of Curtin, this lowers 10(a)shows a TEM image of a monazite- coated Nextel fiber in the composite strength to(0.55/25.4)( /m). For a Weibull modulus of as-coated condition, while Fig. 10(b) shows the interfacial region 10, this results in a strength of 0.68. The ratio of strength after of a similar sample heat-treated at 1200C for 1000 h in air. During 100 h heat treatment (143 MPa) to that after only 5 h(198 MPa thermal exposure, the monazite coating appeared to retain its is 0. 72, This good agreement indicates that the loss in interfacial
showed that extensive crack deflection occurred. It is possible that the crack deflection occurred at the fiber/monazite interface and then began to alternate between the two weak interfaces. It is also worth pointing out that the matrix in this work was porous and much weaker in modulus and strength than the matrix used in the work of Morgan and Marshall. The results of push-out testing revealed that the 100 and 1000 h samples exhibited extremely low debond energies and low sliding friction, which indicated that the monazite was, at a minimum, partially debonded in the sample. This debonding may have occurred during sample preparation and/or from the densification/ spheroidization of the coating within the composite. Figure 9 shows a monazite-coated Nextel 610 fiber from the fracture surface of a sample heat-treated for 1000 h at 1200°C. The coating itself appeared to have sintered during thermal exposure. Figure 10(a) shows a TEM image of a monazite-coated Nextel fiber in the as-coated condition, while Fig. 10(b) shows the interfacial region of a similar sample heat-treated at 1200°C for 1000 h in air. During thermal exposure, the monazite coating appeared to retain its overall shape; however, grain growth and pore coarsening occurred within the coating. Porosity (white regions) could be seen within the heat-treated coating, at the coating/matrix interface, and at the coating/fiber interface. The densification of monazite led to the formation of isolated “gaps” in the interfacial region. Previous work in other systems has demonstrated that “gaps” at the interface can be beneficial to composite behavior.38 The drop in ultimate strength of the Nextel 610/monazite/ alumina composite with long-term heat treatment can be rationalized as follows. The drop in friction stress from 17 MPa (0.15 114) after 5 h at 1200°C down to 0.11 MPa (0.12 0.9) in 100/1000 h causes an increase in the effective gauge length of the fiber in the composite, as per the model of Curtin.39 The effective gauge length increases from nearly 0.55 mm to the composite gauge length of 25.4 mm. As per the model of Curtin, this lowers composite strength to (0.55/25.4)(1/m) . For a Weibull modulus of 10, this results in a strength of 0.68. The ratio of strength after 100 h heat treatment (143 MPa) to that after only 5 h (198 MPa) is 0.72. This good agreement indicates that the loss in interfacial Fig. 6. SEM images of Nextel 610/monazite/alumina samples heat-treated at 1200°C/1000 h/air ((a) low-magnification view of broken test sample, (b) higher-magnification view of fracture surface showing fiber pullout (BSD), (c) deformed monazite attached to fiber that has been pulled from the matrix (BSD), and (d) deformed monazite within matrix trough). February 2003 Monazite Coatings in Oxide/Oxide Composites after Long-Term Exposure at High Temperature 329
330 Journal of the American Ceramic Sociery-Keller et al. Vol 86. No. 2 G=13J/m2; G=55J/m2;μ=0.15;=114MPa(=0) 0.5 =227MPa(a=0) 0.5h=25nm;2d=0.15μm h=50μm;2d=0.1μm V=0.2,E=50GPa,v=,23 04 V=02,E=50GPa,v=23乙 0.4 0.3 0.3 (a)Nextel 610/Alumina 0.2 (b) Nextel 610/Monazite/A lum in a 0.2 1200C,5h 1200C,5h 00.20.40.60.8 Displacment(u m) Displacement (H m) 0.6 G=0.15J/m;μ=0.12;a=0.9MPa(=0 G=0.15J/m;μ=0.12;=0.9MPa(G=0 05h=02m;2d=0.15μ 0.5h=0.2m;2d=015gm 0.4 0.2,E=50GP V=0.2,E=50GPa,v=23 04r 0.3 0.2 (c) Nextel 610/Monazite/A lum in a (d)Nextel 610/Monazite/Alumina 1200C,100h 1200C,1000h 0.1 0.1 00.20.40 0.8 0.6 Dis placment(u m) Displacment(μm) Fig. 7. Fiber push-in/push-out data for(a)control samples of Nextel 610/alumina and Nextel 610/monazite/alumina samples after heat treatment at 1200oC for(b)5,(c)100, and(d)1000 h(Figs. 7(a)and(b) represent pushin, whereas Figs. 7(c)and(d) represent pushout. Fib Displa 20um 12000. The coating ting on Nextel 610 fiber surface after 1000 h at sintered during heat treatment, resulting in some Fig. 8. Pushed fiber from Nextel 610/monazite/alumina sample heat- treated at 1200 C/1000 h/ai monazite protects the fibers in some way ar effects are due in some part to the fibers in the control shear strength is sufficient to ac for the loss in composite degrading faster than in the coated fiber specimens. For strength. This suggests that future efforts should focus on prevent in earlier experiments, Nextel 610/alumina minicom ing the monazite from sintering and causing interfacial gaps, i.e., posed at 1200 C/100 h in air showed grain growth in the fiber producing dense, continuous coatings whereas minicomposites containing monazite coatings exhibited mhe你 of the nis m. complet lowth atter a similar exposure+图am fibers within the composites are unknown. It is possible that conceivably lead to the results observed here. However, in this
shear strength is sufficient to account for the loss in composite strength. This suggests that future efforts should focus on preventing the monazite from sintering and causing interfacial gaps, i.e., producing dense, continuous coatings. Finally, it should be noted that, while the effects of processing on fibers have been carefully evaluated, the actual strengths of the fibers within the composites are unknown. It is possible that monazite protects the fibers in some way and that the observed effects are due in some part to the fibers in the control specimens degrading faster than in the coated fiber specimens. For example, in earlier experiments, Nextel 610/alumina minicomposites exposed at 1200°C/100 h in air showed grain growth in the fiber, whereas minicomposites containing monazite coatings exhibited no significant growth after a similar exposure.22 Such a mechanism, completely unrelated to interfacial crack deflection, could conceivably lead to the results observed here. However, in this Fig. 7. Fiber push-in/push-out data for (a) control samples of Nextel 610/alumina and Nextel 610/monazite/alumina samples after heat treatment at 1200°C for (b) 5, (c) 100, and (d) 1000 h. (Figs. 7(a) and (b) represent pushin, whereas Figs. 7(c) and (d) represent pushout.) Fig. 8. Pushed fiber from Nextel 610/monazite/alumina sample heattreated at 1200°C/1000 h/air. Fig. 9. Monazite coating on Nextel 610 fiber surface after 1000 h at 1200°C. The coating sintered during heat treatment, resulting in some isolated spheroidization. 330 Journal of the American Ceramic Society—Keller et al. Vol. 86, No. 2
February 2003 Monaite Coatings in Oxide/Oxide Composites after Long-Term Exposure at High Temperature composite could be rationalized as being due to the reduction i facial shear strength. In turn, the reduction interfacial es was attributed to densification/spheroidiza tion of the coating during heat treatment, which led formation of isolated Ity at the interface nation of composite fracture surfaces gave no direct ev the mode and location of the first interfacial crack although subsequent cracks wandered in the interfac Monazite sometimes at the matrix/coating interface, sometimes at the coat ing/fiber interface, and possibly within the coating. The near Fiber exclusivity of crack deflection in the interface region and not in the porous matrix is a strong indication that monazite will be effective in dense matrix composites. The authors thank Dr. R. Hay for TEM evaluation, M. Cook and C. Cooke for (a) 500nm mple preparation, and Dr. R J. Kerans for the critical review of the manuscript. References R.J. Kerans, R. S Hay, N. J. Pagano, and T. A. Parthasarathy, "The Role of the Fiber-Matrix Interface in Ceramic Composites, " Am. Ceram. Soc. Bull, 68[2] K. T. Faber, "Ceramic Composite Interfaces: Properties and Design, Annz Re Monazite Mater.Sc.,27,499-524(1997) Glass Matrix Composites,J. Mater. Sci., 15 (2]463-68(1980). AR. L Lehman and C. A Doughan, "Carbon Coated Alumina Fiber/Glass Matrix w. Rice, "BN Coating of Ceramic Fibers for Ceramic Fiber Composites, U.S Pat.No.4642271,Feb.10,1987 a Matrix Cotteret, "Boron Nitride Interphase in Ceramic-Matrix Composites, JAm Ceram Fiber F. Lange, w.C. Tu, and A. G. Evans, "Processing of Damage Toleran Oxidation-Resistant Ceramic-Matrix Composites, Mater. Sci. Eng. A, A195, 145-50 G.N. Morscher, "Tensile Stress Rupture of SiC/SiC Minicomposites with Carbon Sun, H-T. Lin, and J.J. Brennan, "Intermediate-Temperature Environmental Effects on Boron Nitride-Coated Silicon Carbide-Fiber-Reinforced Glass-Ceramic 500nm Composites,J. Am Ceram Soc., 80 3]609-14(1997). 1°M.A. Mattoni,J.Y.Yang,C.G.Levi,andF.W.Zok,“ Effects of matriⅸ on the Mechanical Properties of a Porous-Matrix, All-Oxide Ceramic Composite, " J. Am. Ceram Soc., 84 [1112594-602(2001). Fig. 10. TEM micrographs showing(a) as-coated Nextel fiber and (b) R J. Kerans, R.S. Hay, and T. A. Parthasarathy, "Structural Ceramic Compos- Nextel 610/monazite/alumina composite(1200C/1000 h). The monazite tes, CurT. Opin. Solid State Mater. Sci., 4, 445-51(1999). coating experienced grain growth and pore coarsening during high- 12D.B.Mars nd J. B. Davis. "Ceramics for Future Power Generatio temperature exposure, while maintaining its overall shape Technology: Fiber Reinforced Oxide Composites, " CuT. Opin. Solid State Mater. aP. E. D, Morgan and D. B. Marshall, "Ceramic Composites of Monazite and Alumina,J.A. Ceram Soc., 78[6]1553-63(1995). 14A. Cazzato, M. Colby, D. Daws, J. Davis, P, Morgan, J. Porter, S. Butner, and B. there are dramatic differences in behavior after only 5 h at Jurf, "Monazite Interface Coatings in Polymer and Sol-Gel Derived Ceramic-Matrix and 1200C, wherein grain growth is limited and not a C=且MCmD1B3a.rm significant factor. It seems far more likely that the results are truly "properties of La monazite as an Interphase in Oxide Composites, Z Metalkd, 90 due to monazite ding crack deflection independent of the B. Davis, D B. Marshall, and P E. D Morgan, "Oxide Composites of LapOa and Al-O,, J. Eur. Ceram Soc., 19, 2421-26(1999). IV. Summary and Conclusions ISD.-H. Kuo, W. M. Kriven, and T J Mackin,"Control azite Coatings in Oxide-Oxide Composites,J. The results of this work demonstrate that monazite can provide Ceram.Soc,80212987-96(1997 crack deflection (a weak interface) in fiber roving/thin 9s. M.Johnson,YBlum,CKan eramic-matrix composites and prevent fiber/matrix interac Development for an Oxide-Oxide Composite,"Met. Mater, 4[6]1119-25(1998) M. H. Lewis, S. York, C. Freeman, I. C. Alexander, I. Al-Dawery, E. G. Butler an extended time at high temperature. Nextel 610/me and P. A. Doleman, "Oxide CMCs: Novel Fibres, Coatings and Fabrication Proce alumina composites displayed an initial strength loss with expo dures, "Ceram Eng. Sci. Proc., 21 [3]535-47(2000 sure at high temperature(1100%-1200C/5-250 h); however, this 2K. K. Chawla, H. Liu, J. Janczak-Rusch, and S. Sambasivan, "Microstructure and strength was maintained through 1000 h at 1200%C. Control Properties Of Monazite(LaPO4)Coated Saphikon Fiber/Alumina Matrix Compo ites, "J. Eur. Ceram Soc 20, 551-59(200 samples, with no monazite coatings, displayed a substantial 4K. A. Keller, T. Mah, T.A. Parthasarathy, E. E. Boakye, and M. K. Cinibulk strength loss(70%)after only a short-term exposure (5 h)to "Evaluation of All-Oxide Based on Coated Nextel 610 and Nextel 650 1200.C. Through push-out testing, it was ent that the Fibers, Ceram. Eng. Sci. Proc., 22 [3]667-75(2001) matrix/monazite/fiber interface was weak and became even N. P. Bansal and J. 1. Eldridge, "Hi-Nicalon Fiber-Reinforced Celsian Matrix Composites: Influence of Interface Modification, "J. Mater. Res, 13 16]1530 weaker with longer-term thermal exposure. The initial loss in
work, there are dramatic differences in behavior after only 5 h at 1100° and 1200°C, wherein grain growth is limited and not a significant factor. It seems far more likely that the results are truly due to monazite providing crack deflection independent of the matrix. IV. Summary and Conclusions The results of this work demonstrate that monazite can provide crack deflection (a weak interface) in fiber roving/thin coating/ ceramic-matrix composites and prevent fiber/matrix interaction for an extended time at high temperature. Nextel 610/monazite/ alumina composites displayed an initial strength loss with exposure at high temperature (1100°–1200°C/5–250 h); however, this strength was maintained through 1000 h at 1200°C. Control samples, with no monazite coatings, displayed a substantial strength loss (70%) after only a short-term exposure (5 h) to 1200°C. Through push-out testing, it was apparent that the matrix/monazite/fiber interface was weak and became even weaker with longer-term thermal exposure. The initial loss in composite strength could be rationalized as being due to the reduction in interfacial shear strength. In turn, the reduction in interfacial properties was attributed to densification/spheroidization of the coating during heat treatment, which led to the formation of isolated “gaps” or porosity at the interface. Examination of composite fracture surfaces gave no direct evidence of the mode and location of the first interfacial crack deflection, although subsequent cracks wandered in the interface region, sometimes at the matrix/coating interface, sometimes at the coating/fiber interface, and possibly within the coating. The nearexclusivity of crack deflection in the interface region and not in the porous matrix is a strong indication that monazite will be effective in dense matrix composites. Acknowledgments The authors thank Dr. R. Hay for TEM evaluation, M. Cook and C. Cooke for sample preparation, and Dr. R. J. Kerans for the critical review of the manuscript. References 1 R. J. Kerans, R. S. Hay, N. J. Pagano, and T. A. Parthasarathy, “The Role of the Fiber-Matrix Interface in Ceramic Composites,” Am. Ceram. Soc. Bull., 68 [2] 429–42 (1989). 2 K. T. Faber, “Ceramic Composite Interfaces: Properties and Design,” Annu. Rev. Mater. Sci., 27, 499–524 (1997). 3 K. Prewo and J. J. Brennan, “High-Strength Silicon Carbide Fiber-Reinforced Glass Matrix Composites,” J. Mater. Sci., 15 [2] 463–68 (1980). 4 R. L. Lehman and C. A. Doughan, “Carbon Coated Alumina Fiber/Glass Matrix Composites,” Compos. Sci. Technol., 37, 149–64 (1990). 5 R. W. Rice, “BN Coating of Ceramic Fibers for Ceramic Fiber Composites,” U.S. Pat. No. 4 642 271, Feb. 10, 1987. 6 R. Naslain, O. Dugne, A. Guette, J. Severly, C. R. Brosse, J.-P. Rocher, and J. Cotteret, “Boron Nitride Interphase in Ceramic-Matrix Composites,” J. Am. Ceram. Soc., 74 [10] 2482–88 (1991). 7 F. F. Lange, W. C. Tu, and A. G. Evans, “Processing of Damage Tolerant, Oxidation-Resistant Ceramic-Matrix Composites,” Mater. Sci. Eng. A, A195, 145–50 (1995). 8 G. N. Morscher, “Tensile Stress Rupture of SiC/SiC Minicomposites with Carbon and Boron Nitride Interphases at Elevated Temperatures in Air,” J. Am. Ceram. Soc., 80 [8] 2029–42 (1997). 9 E. Y. Sun, H.-T. Lin, and J. J. Brennan, “Intermediate-Temperature Environmental Effects on Boron Nitride-Coated Silicon Carbide-Fiber-Reinforced Glass-Ceramic Composites,” J. Am. Ceram. Soc., 80 [3] 609–14 (1997). 10M. A. Mattoni, J. Y. Yang, C. G. Levi, and F. W. Zok, “Effects of Matrix Porosity on the Mechanical Properties of a Porous-Matrix, All-Oxide Ceramic Composite,” J. Am. Ceram. Soc., 84 [11] 2594–602 (2001). 11R. J. Kerans, R. S. Hay, and T. A. Parthasarathy, “Structural Ceramic Composites,” Curr. Opin. Solid State Mater. Sci., 4, 445–51 (1999). 12D. B. Marshall and J. B. Davis, “Ceramics for Future Power Generation Technology: Fiber Reinforced Oxide Composites,” Curr. Opin. Solid State Mater. Sci., 5, 283–89 (2001). 13P. E. D. Morgan and D. B. Marshall, “Ceramic Composites of Monazite and Alumina,” J. Am. Ceram. Soc., 78 [6] 1553–63 (1995). 14A. Cazzato, M. Colby, D. Daws, J. Davis, P. Morgan, J. Porter, S. Butner, and B. Jurf, “Monazite Interface Coatings in Polymer and Sol-Gel Derived Ceramic-Matrix Composites,” Ceram. Eng. Sci. Proc., 18 [3] 269–78 (1997). 15D. B. Marshall, J. B. Davis, P. E. D. Morgan, J. R. Waldrop, and J. P. Porter, “Properties of La Monazite as an Interphase in Oxide Composites,” Z. Metallkd., 90 [12] 1048–52 (1999). 16J. B. Davis, D. B. Marshall, and P. E. D. Morgan, “Oxide Composites of LaPO4 and Al2O3,” J. Eur. Ceram. Soc., 19, 2421–26 (1999). 17J. B. Davis, D. B. Marshall, and P. E. D. Morgan, “Monazite-Containing Oxide–Oxide Composites,” J. Eur. Ceram. Soc., 20 [5] 583–87 (2000). 18D.-H. Kuo, W. M. Kriven, and T. J. Mackin, “Control of Interfacial Properties through Fiber Coatings: Monazite Coatings in Oxide–Oxide Composites,” J. Am. Ceram. Soc., 80 [12] 2987–96 (1997). 19S. M. Johnson, Y. Blum, C. Kanazawa, and H.-J. Wu, “Low-Cost Matrix Development for an Oxide–Oxide Composite,” Met. Mater., 4 [6] 1119–25 (1998). 20M. H. Lewis, S. York, C. Freeman, I. C. Alexander, I. Al-Dawery, E. G. Butler, and P. A. Doleman, “Oxide CMCs: Novel Fibres, Coatings and Fabrication Procedures,” Ceram. Eng. Sci. Proc., 21 [3] 535–47 (2000). 21K. K. Chawla, H. Liu, J. Janczak-Rusch, and S. Sambasivan, “Microstructure and Properties Of Monazite (LaPO4) Coated Saphikon Fiber/Alumina Matrix Composites,” J. Eur. Ceram. Soc., 20, 551–59 (2000). 22K. A. Keller, T. Mah, T. A. Parthasarathy, E. E. Boakye, and M. K. Cinibulk, “Evaluation of All-Oxide Composites Based on Coated Nextel 610 and Nextel 650 Fibers,” Ceram. Eng. Sci. Proc., 22 [3] 667–75 (2001). 23N. P. Bansal and J. I. Eldridge, “Hi-Nicalon Fiber-Reinforced Celsian Matrix Composites: Influence of Interface Modification,” J. Mater. Res., 13 [6] 1530–37 (1998). Fig. 10. TEM micrographs showing (a) as-coated Nextel fiber and (b) Nextel 610/monazite/alumina composite (1200°C/1000 h). The monazite coating experienced grain growth and pore coarsening during hightemperature exposure, while maintaining its overall shape. February 2003 Monazite Coatings in Oxide/Oxide Composites after Long-Term Exposure at High Temperature 331
332 Journal of the American Ceramic Sociery-Keller et al. Vol 86. No. 2 E.E. Boakye, R S Hay, P Mogilevsky, and L M. Douglas, "Monazite Coatings 3E. E. Boakye, M. D. Petry, R S. Hay, and L. M. Douglas, "Monazite Coatings n Nextel 720, 610, and Tyranno-SA Fiber Tows: ength,Ceram. Eng. Sci. Proc., 21 [4]229-36(20 3M. D Petry, T Mah, and R, J.Kerans,Validity iameter for and Weibull Modulus of J. Am L. P. Zawada, and S. Mal Soc,800]2741-44(1997) 34R.S Hay, "Monazite and Scheelite Deformation Mechanisms, "Ceram. Eng. Sci. 720/A)Composite at Room and Elevated Temperature, "Ceram. Eng. Proe,214203-18(2000 ST. A Parthasarathy, D. B. Marshall, and R. J. Kerans, "Analysis of the Effect of Advances in Oxide-Oxide CMC,J Eng Gas n Fiber Debonding and Sliding in Brittle matrix o. Sudre and F.E. Lange, "Efect of Inclusions on Densification 1. Microstrnve. Acset etal,421377-84(1994) Inclusions pmemt. Canam. So 5 3 5192-24 (1992)ge volume Fraction of zro D B. Marshall, J.R. Waldrop, and P. E D Morgan,"Thermal Interface between Alumina and Monazite, "Acta Mater, 48[18/19]44 3M, Nextel Textiles Product Description Brochure. St. Paul, MN, 3M 37M. Y. He and J. W. Hutchinson, "Crack Deflection at the Interface between 账AK出TAP功,M时B下DkM面人四m1 Carbon Coatings for Oxide/Oxide Composites,J. Am. Ceram Soc., 83[2] 329-36 V. Carelli, H. Fugita, J. Y Yang, and he Mechanical Properties of a Porous-Matrix Ceramic Composite, J.Am. Ceram rtin,“ Theory of Mecl erties of Ceramic-Matrix Compo Soc,853]595-602(2002) ites, ". Am. Ceram Soc., 74 [1112837-45(1991)
24E. E. Boakye, R. S. Hay, P. Mogilevsky, and L. M. Douglas, “Monazite Coatings on Fibers: II, Coating without Strength Degradation,” J. Am. Ceram. Soc., 84 [12] 2793–801 (2001). 25C. G. Levi, J. Y. Yang, B. J. Dalgleish, F. W. Zok, and A. G. Evans, “Processing and Performance of an All-Oxide Ceramic Composite,” J. Am. Ceram. Soc., 81 [8] 2077–86 (1998). 26S. G. Steel, L. P. Zawada, and S. Mall, “Fatigue Behavior of a Nextel 720/Alumina (N720/A) Composite at Room and Elevated Temperature,” Ceram. Eng. Sci. Proc., 22 [3] 695–702 (2001). 27R. A. Jurf and S. C. Butner, “Advances in Oxide–Oxide CMC,” J. Eng. Gas Turbines Power, 122 [2] 202–205 (2000). 28O. Sudre and F. F. Lange, “Effect of Inclusions on Densification: I. Microstructural Development in an Al2O3 Matrix Containing a Large Volume Fraction of ZrO2 Inclusions,” J. Am. Ceram. Soc., 75 [3] 519–24 (1992). 293M, Nextel Textiles Product Description Brochure. St. Paul, MN, 3M 30K. A. Keller, T. A. Parthasarathy, T. Mah, and E. Boakye, “Evaluation of Monazite Fiber Coatings in Dense Matrix Composites,” Ceram. Eng. Sci. Proc., 20 [3] 451–61 (1999). 31E. A. V. Carelli, H. Fugita, J. Y. Yang, and F. W. Zok, “Effects of Thermal Aging on the Mechanical Properties of a Porous-Matrix Ceramic Composite,” J. Am. Ceram. Soc., 85 [3] 595–602 (2002). 32E. E. Boakye, M. D. Petry, R. S. Hay, and L. M. Douglas, “Monazite Coatings on Nextel 720, 610, and Tyranno-SA Fiber Tows: Effects of Precursors on Fiber Strength,” Ceram. Eng. Sci. Proc., 21 [4] 229–36 (2000). 33M. D. Petry, T. Mah, and R. J. Kerans, “Validity of Using Average Diameter for Determination of Tensile Strength and Weibull Modulus of Ceramic Filaments,” J. Am. Ceram. Soc., 80 [10] 2741–44 (1997). 34R. S. Hay, “Monazite and Scheelite Deformation Mechanisms,” Ceram. Eng. Sci. Proc., 21 [4] 203–18 (2000). 35T. A. Parthasarathy, D. B. Marshall, and R. J. Kerans, “Analysis of the Effect of Interfacial Roughness on Fiber Debonding and Sliding in Brittle Matrix Composites,” Acta Metall., 42 [11] 3773–84 (1994). 36D. B. Marshall, J. R. Waldrop, and P. E. D. Morgan, “Thermal Grooving at the Interface between Alumina and Monazite,” Acta Mater., 48 [18/19] 4471–74 (2000). 37M. Y. He and J. W. Hutchinson, “Crack Deflection at the Interface between Dissimilar Materials,” Int. J. Solid Struct., 25, 1053–67 (1989). 38K. A. Keller, T. Mah, T. A. Parthasarathy, and C. M. Cooke, “Fugitive Interfacial Carbon Coatings for Oxide/Oxide Composites,” J. Am. Ceram. Soc., 83 [2] 329–36 (2000). 39W. A. Curtin, “Theory of Mechanical Properties of Ceramic-Matrix Composites,” J. Am. Ceram. Soc., 74 [11] 2837–45 (1991). 332 Journal of the American Ceramic Society—Keller et al. Vol. 86, No. 2