COMPOSITES SCIENCE AND TECHNOLOGY ELSEVIER Composites Science and Technology 59(1999)861-872 The application of ceramic-matrix composites to the automotive ceramic gas turbine Hiroshi Kaya' Petroleum Energy Center, Technology Research and Development, 3.9 Toranom 4-Chome, Mintao-ku, Tokyo 105, Japan Received 7 August 1997: received in revised form 24 August 1998: accepted 8 January 1999 Abstract A Japanese 100 kW automotive ceramic gas turbine(CGT) project was started in 1990 and was concluded successfully in 1997. This project was supported by the Ministry of International Trade and Industry and was conducted by the Petroleum Energy Center to achieve the targets of this project such as higher thermal efficiency over 40% at a turbine inlet temperature of 1350.C, lower exhaust emissions to meet Japanese regulations, and multi-fuel capabilities. Ceramic-matrix composites(CMCs)are expected to become one of the most reliable materials for high-temperature use to make up for the deficient properties of monolithic ceramics and heat-resistant alloys. Carbon fiber, silicon nitride fiber, silicon carbide fiber, silicon carbide whisker, in situ silicon nitride, TiB, milled carbon fiber were used as reinforcements for silicon carbide, Si-N-C, SiAION and silicon nitride matrix composites. Higher mechanical properties tested by the developed testing standards, and reliability against thermal shock, particle impact damage and creep resistance were con- firmed to apply these CMCs for engine components. Several screening test steps were performed before the engine tests and these con- firmed that CMC had strong potential for actual engine components. 1999 Elsevier Science Ltd. All rights reserved Keywords: A Ceramic-matrix composites(CMs): Ceramic gas turbine; Ceramic component; Long-fiber reinforced CMC; Short-fiber-reinforced CMC 1. Preface fields such as gas turbines operating under severe condi tions where CMC is considered to work effectively A Japanese 100 kW automotive ceramic gas turbine CMCs were also developed in the seven-year project (CGT)project was started in 1990 and was concluded and the applications of CMC for various components uccessfully in 1997. This project was supported by the have progressed positively in consideration of the lim Ministry of International Trade and Industry and was itations of the monolithic. This report deals with the conducted by the Petroleum Energy Center to achieve technologies for manufacturing components from a heat the targets such as higher thermal efficiency of 40%, resistant composite material for CGt taking into con sideration that the turbine inlet temperature (TIT)is 1350C, and deals with the superiority of the heat-resis- nese regulations, and multi-fuel capabilities [1, 2]. tant co mposite material for CGT, studied in the latte Finally, an output power of 92.3 kw, a thermal effi- stage of the project ciency of 35.6% were achieved [3-5I Ceramic-matrix composites(CMCs)produced by sev- eral kinds of reinforcements such as in fibers, particles and 2. Objectives whiskers with ceramics are known mainly in aerospace industries in Europe and the USA as high fracture resis- The thermal stress to be generated and life prediction tant materials capable of overcoming a fatal defect, speci- were estimated for every component whereby the target fically, brittleness of monolithic ceramic materials(here value for development of the mechanical characteristics after simply called"'monolithics'[6-11]. However, there is of a material, which were required for each CGT com- no precedent for the application of CMC in unknown ponent, was determined. With respect to, for example, a turbine rotor for which the highest strength at high tem- Tel:+8l-3-5402-8506;fax:+8l 8513: e-mail: shingen@ peratures is required, the target value for the monolithic was determined so that the failure probability after 10 0266-3538/99/.see front matter o 1999 Elsevier Science Ltd. All rights reserved. PlI:S0266-3538(99)00016
The application of ceramic-matrix composites to the automotive ceramic gas turbine Hiroshi Kaya1 Petroleum Energy Center, Technology Research and Development, 3-9 Toranom 4-Chome, Mintao-ku, Tokyo 105, Japan Received 7 August 1997; received in revised form 24 August 1998; accepted 8 January 1999 Abstract A Japanese 100 kW automotive ceramic gas turbine (CGT) project was started in 1990 and was concluded successfully in 1997. This project was supported by the Ministry of International Trade and Industry and was conducted by the Petroleum Energy Center to achieve the targets of this project such as higher thermal eciency over 40% at a turbine inlet temperature of 1350C, lower exhaust emissions to meet Japanese regulations, and multi-fuel capabilities. Ceramic-matrix composites (CMCs) are expected to become one of the most reliable materials for high-temperature use to make up for the de®cient properties of monolithic ceramics and heat-resistant alloys. Carbon ®ber, silicon nitride ®ber, silicon carbide ®ber, silicon carbide whisker, in situ silicon nitride, TiB2/milled carbon ®ber were used as reinforcements for silicon carbide, SiÿNÿC, SiAlON and silicon nitride matrix composites. Higher mechanical properties tested by the developed testing standards, and reliability against thermal shock, particle impact damage and creep resistance were con- ®rmed to apply these CMCs for engine components. Several screening test steps were performed before the engine tests and these con- ®rmed that CMC had strong potential for actual engine components. # 1999 Elsevier Science Ltd. All rights reserved. Keywords: A. Ceramic-matrix composites (CMCs); Ceramic gas turbine; Ceramic component; Long-®ber reinforced CMC; Short-®ber-reinforced CMC 1. Preface A Japanese 100 kW automotive ceramic gas turbine (CGT) project was started in 1990 and was concluded successfully in 1997. This project was supported by the Ministry of International Trade and Industry and was conducted by the Petroleum Energy Center to achieve the targets such as higher thermal eciency of 40%, output power of 100 kW at a turbine inlet temperature of 1350C and lower exhaust emissions to meet Japanese regulations, and multi-fuel capabilities [1,2]. Finally, an output power of 92.3 kW, a thermal e- ciency of 35.6% were achieved [3±5]. Ceramic-matrix composites (CMCs) produced by several kinds of reinforcements such as in ®bers, particles and whiskers with ceramics are known mainly in aerospace industries in Europe and the USA as high fracture resistant materials capable of overcoming a fatal defect, speci- ®cally, `brittleness' of monolithic ceramic materials (here after simply called `monolithics' [6±11]. However, there is no precedent for the application of CMC in unknown ®elds such as gas turbines operating under severe conditions where CMC is considered to work eectively. CMCs were also developed in the seven-year project and the applications of CMC for various components have progressed positively in consideration of the limitations of the monolithic. This report deals with the technologies for manufacturing components from a heatresistant composite material for CGT taking into consideration that the turbine inlet temperature (TIT) is 1350C, and deals with the superiority of the heat-resistant composite material for CGT, studied in the latter stage of the project. 2. Objectives The thermal stress to be generated and life prediction were estimated for every component whereby the target value for development of the mechanical characteristics of a material, which were required for each CGT component, was determined. With respect to, for example, a turbine rotor for which the highest strength at high temperatures is required, the target value for the monolithic was determined so that the failure probability after 10 Composites Science and Technology 59 (1999) 861±872 0266-3538/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S0266-3538(99)00016-0 1 Tel: +81-3-5402-8506; fax: +81-3-5402-8513; e-mail: shinnen@ plaza.pecj.or.jp
H. Kaya/Composites Science and Technology 59(1999)861-872 years of operation over 100,000 km, including 10,000 The number of the developing CMC components was starts and stops, specifically, and the failure probability finally limited to five, including a turbine rotor as a after a continuous running test for 300 h under the rated rotational component of high-strength material, a back conditions of 1350 C and 110, 000 rpm(design rotation plate as a stationary component of high-strength mate as reduced to 100,000 rpm afterwards) was 1x10-5 or rial, an orifice liner and an extension liner as stationary less. The target values for CMCs were then determined combustor components of high-temperature and oxida- according to the above designed target values for the tion- resistant material, and an inner scroll support as a monolithic. In addition to such mechanical character- stationary component of high-toughness material. The istic tests, various evaluation tests were performed, since major components(models) which have been developed CMC was expected to exhibit superiority against the at the present time are shown in Fig. 1. The approx thermal shock and particle impact dar mages ex pected imate sizes of the components are shown in Fig. 2. The under actual engine conditions characteristics of a material to be developed must first Extension Liner (Chopped Sic Fiber Reinforced SiNC) Out Inner Scroll Combustion Liner i In-Situ Si3N4 Reinforced Si3N4) (Carbon Fiber Reinforced Sic) Orifice Lin (Milled Carbon Fiber and TiB2 Powder Reinforced Sic) Turbine Rotor (SiC Whisker Reinforced SIAION) (SIC Fiber Reinforced SiNC)(n-Situ SiaNa Reinforced Si3N4) Inner Scroll Support (Carbon Fiber Reinforced Sic) SiN Fiber Reinforced SiNC) Turbine Rotor (Carbon Fiber Reinforced SiC) Back Plate Inner Shroud (SiC Whisker Reinforced SiAJON) Carbon Fiber Reinforced Sic) (In-Situ SiaNa Reinforced Si3N4) Fig 1. Major CMC components(models) for CGT. Orifice Liner Extension Liner Back Plate 21 中177 Turbine Rotor nner Scroll Support ts for cgt
years of operation over 100,000 km, including 10,000 starts and stops, speci®cally, and the failure probability after a continuous running test for 300 h under the rated conditions of 1350C and 110,000 rpm (design rotation was reduced to 100,000 rpm afterwards) was 110ÿ5 or less. The target values for CMCs were then determined according to the above designed target values for the monolithic. In addition to such mechanical characteristic tests, various evaluation tests were performed, since CMC was expected to exhibit superiority against the thermal shock and particle impact damages expected under actual engine conditions. The number of the developing CMC components was ®nally limited to ®ve, including a turbine rotor as a rotational component of high-strength material, a back plate as a stationary component of high-strength material, an ori®ce liner and an extension liner as stationary combustor components of high-temperature and oxidation-resistant material, and an inner scroll support as a stationary component of high-toughness material. The major components (models) which have been developed at the present time are shown in Fig. 1. The approximate sizes of the components are shown in Fig. 2. The characteristics of a material to be developed must ®rst Fig. 1. Major CMC components (models) for CGT. Fig. 2. Brief shapes of the developed main CMC components for CGT. 862 H. Kaya / Composites Science and Technology 59 (1999) 861±872
H. Kaya/Composites Science and Technology 59(1999)861-872 e= 已52 曾豆 522g工 aE u 53号6g
Table 1 Evaluation methods of CMCs developed for CGT Classi®cation Developed components Material systems (reinforcement/matrix/CVD) Evaluation items Components Specimens Rotational component High strength material Turbine rotor LRC: C(f)/SiC SRC: SiC(w)/Si-Al-O-N : Si N3 4(is)/Si N3 4 Cold spin test : disk (LRC, SRC) : hub model (LRC)b Hot spin test : rotor (SRC)a Mechanical properties : ¯exural stength (1200C) : fracture toughness : oxidation resistance (1200C200 h) Superiorities Stationary component Back plate SRC: SiC(w)/Si-Al-O-N : Si N3 4(is)/Si N3 4 High temperature stationary component assembly test : 1350C (TIT), 5atmabs : FOD test, : TSFTd *: Only components were evaluated for LRC High temperature oxidation resistant material Ori®ce liner Extension liner SRC: TiB2(p).C(f)/SiC/SiC LRC: Si-C(f)/Si-C/SiC LRC: Si-C(f)/Si-N-C/SiC : screening test for CGT engine Engine combustion enviroment test : 1350C (TIT), 4atmabs : screening test for high temperature stationary component assembly test Mechanical properties : ¯exural strength (1450C) : fracture toughness (SRC) : fracture energy (LRC) : oxidation resistance (1450C200 h) Superiorities : FOD test, : TSFT : engine combustion environment test (1350C) High toughness material Inner scroll support LRC: Si-N(f), Si-C(f)/Si-N-C /SiC : C(f)/Si-C/SiC Hydrostatic pressure loading test : 120% of the expected thermal stress : screening test for high temperature stationary component assembly test Mechanical properties : ¯exural strength (1250C) : fracture energy : oxidation resistance (1250C200 h) Superiorities : FOD test, : TSFT a SRC: short ®ber reinforced CMC. b LRC: long ®ber reinforced CMC. c f: ®ber, w: whisker, p: powder, is: in situ, -: nonstoichometric. d TSFT: thermal shock fracture test. H. Kaya / Composites Science and Technology 59 (1999) 861±872 863
H. Kaya/Composites Science and Technology 59(1999)861-872 x∪寸 258E8≥5 鑼 三 三 2兰 兰8 ≌ v6 之远3 E品王 9g gagE品王
Table 2 Development results of CMCs for CGT Developed components Material systems (reinforcement/matrix/CVD)h Developmental results : components Developmental targets and results : specimens (mechanical properties) Flexural strength (MPa) Fracture toughness (MPa-m1/2) Fracture energy (kj/m2 ) Oxidation resistance (MPa) High strength material (targets) Cold spin test (rpm) Hot spin test 750 (1200C) 8 (SEPB or SEVNB) ± 750 (1200C200h) Turbine rotor LRCb: C(f)/SiC 93700 (Disk: Max) 98600 (Hub: Max) Only components were evaluated SRCa: SiC(w)/Si-Al-O-N Weibull modulus m=23 (Disk) 910 8 (SEVNB)g ± 1000 SRC: Si N3 4(is)/Si N3 4 80100 (Disk: Av) Rotor: 1200C, 70000 rpm 960 7 (SEVNB) ± 890 High strength material (targets) HSATc (1350C, 5atmabs) 750 (1200C) 8 (SEPB or SEVNB) ± 750 (1200C200h) Back plate SRC: SiC(w)/Si-Al-O-N Achieved: 26 h 8 (SEVNB) ± 1000 SRC: Si N3 4(is)/Si N3 4 Achieved: 31 h 7 (SEVNB) ± 890 High temperature oxidation resistant material (targets) HSAT (1350C, 5atmabs) ECETd (1350C, 4atmabs) 400 (1450C) 7(SEPB or SEVNB) 4 400 (1450C200h) Ori®ce liner SRC: TiB2(p)/(f)/SiC/SiC Durable up to 1600C 540 7 (SEPB)f ± 420 LRC: Si-C(f)/Si-C/SiC ± Achieved: 31 h 230 ± 19 170 Extension liner LRC: Si-C(f)/Si-N-C/SiC ± Achieved: 31 h 420 ± 8 360 High toughness material (targets) HPLTe [120% of the expected thermal stress (ETS)] 500 (1250C) ± 10 500 (1250C200h) Inner scroll support LRC: Si-N(f), Si-C(f)/Si-N-C/SiC High oxidation resistant type Achieved: 156%(180 MPa) of the ETS 520 410 ± ± 19 6 310 460 SRC: C(f)/SiC/SiC Achieved: 2000% (20 MPa) of the ETS 390 ± 8 100 a SRC: short ®ber reinforced CMC. b LRC: long ®ber reinforced CMC. c HSAT: high temperature stationary component assembly test. d ECET: engine combustion environment test. e HPLT: hydrostatic pressure loading test. f SEPB: single edge precracked beam. g SEVNB: single edge V-notched beam. h f: ®ber, w: whisker, p: powder, is: in-situ, -: nonstoichiometric. 864 H. Kaya / Composites Science and Technology 59 (1999) 861±872
H. Kaya/Composites Science and Technology 59(1999)861-872 be evaluated at a test piece size. Next, the developed Long fiber reinforced CMCs such as silicon-carbide- component is evaluated prior to the actual engine test fiber-reinforced Si-N-C composite for the extension by the tests equivalent to the engine test or a screening liner and silicon-nitride-fiber-reinforced Si-N-Ccom test [12]. Five engine components manufactured, using8 posite for the inner scroll support attained the target kinds of CMCs shown in Table 1, were evaluated values for the flexural strength at high temperatures according to each development situation. and fracture energy, but their oxidation resistances were lower than the target value. It is important that the properties of reinforcements reflect the properties 3. Results of long-fiber-reinforced CMC. The control of the shear strength at the interface between the fiber and 3. 1. Characteristics of CMC a matrix, specifically, how to pull out the reinforcing fiber from the matrix under load is a key point [13] Table 2 shows the target values of the CMCs under Also, it is important to pull out the fiber without elopment and he resulting mechanical properties damage to the interface in a high temperature and Short-fiber-reinforced and /or powder-reinforced CMCs oxidation environment, for improving the oxidation uch as silicon-carbide-whisker reinforced silicon-carbide resistance. A continuous Si-B-c type self-healing composite for the turbine rotor and the back plate and CVD coating technology as shown in Fig 3 has been TiB,/milled-carbon-fiber-reinforced silicon-carbide cor developed for silicon-nitride- fiber-reinforced Si-N-C posite for the orifice finer satisfied all target values. These type CMC and has attracted attention as a technol- CMCs attracted considerable attention as a material in ogy for improving the oxidation resistance at high which toughness and strength are compatible temperatures [14] Reactants Fiber Cleaning Furnace CVD Furnace Coated Fiber Fig 3. Continuous CVD coating method on ceramic fiber(Reactants SiCl4 BCl3. CH4. NH, N2 H2, Reaction conditions temp. <1500C, Press <I atm) c≈o 2。S Fig 4. Thermal shock fracture test result of CMCs
be evaluated at a test piece size. Next, the developed component is evaluated prior to the actual engine test by the tests equivalent to the engine test or a screening test [12]. Five engine components manufactured, using 8 kinds of CMCs shown in Table 1, were evaluated according to each development situation. 3. Results 3.1. Characteristics of CMC Table 2 shows the target values of the CMCs under development and the resulting mechanical properties. Short-®ber-reinforced and/or powder-reinforced CMCs such as silicon-carbide-whisker reinforced silicon-carbide composite for the turbine rotor and the back plate and TiB2/milled-carbon-®ber-reinforced silicon-carbide composite for the ori®ce ®ner satis®ed all target values. These CMCs attracted considerable attention as a material in which toughness and strength are compatible. Long ®ber reinforced CMCs such as silicon-carbide- ®ber-reinforced SiÿNÿC composite for the extension liner and silicon-nitride-®ber-reinforced SiÿNÿC composite for the inner scroll support attained the target values for the ¯exural strength at high temperatures and fracture energy, but their oxidation resistances were lower than the target value. It is important that the properties of reinforcements re¯ect the properties of long-®ber-reinforced CMC. The control of the shear strength at the, interface between the ®ber and a matrix, speci®cally, how to pull out the reinforcing ®ber from the matrix under load is a key point [13]. Also, it is important to pull out the ®ber without damage to the interface in a high temperature and oxidation environment, for improving the oxidation resistance. A continuous SiÿBÿC type self-healing CVD coating technology as shown in Fig. 3 has been developed for silicon-nitride-®ber-reinforced SiÿNÿC type CMC and has attracted attention as a technology for improving the oxidation resistance at high temperatures [14]. Fig. 4. Thermal shock fracture test result of CMCs. Fig. 3. Continuous CVD coating method on ceramic ®ber (Reactants:SiCl4,BCl3,CH4,NH3,N2,H2, Reaction conditions:temp.<1500C,Press.<1 atm). H. Kaya / Composites Science and Technology 59 (1999) 861±872 865
H. Kaya/Composites Science and Technology 59(1999)861-872 3.2. Superiorities of CMC stress(thermal shock) is applied when an engine starts and stops. Circular rings, 100/70 mm OD/ID and 5 mm CMC has been developed to overcome the brittleness thick, made to simulate these components and provided of the monolithic. The thermal shock resistance, particle with a micro-notch so that the circular rings are easily impact resistance, creep resistance, etc. are important broken, were exposed to a high-temperature (1350C) charact eristic required for materials which are reliable and high -speed atel m(s ga oiw and a c ohle d gas fiber reinforced CMCs particularly withstood such severe cir- 3.2.. Thermal shock resistance cumstances without any failure and exhibited superiority Scrolls are components to which the highest thermal to the monolithic. 1200 Long fiber reinforced CMc 1000 (Actual Strength X 2.5) 800 600 Short Fiber Reinforced CMC Monolithic siC 0 0100200300400500600700800 Impact velocity(m/s) Fig. 5. Particle impact test results of CMCs SIAION 0.4 SAON(SiCW5vol‰) SAON(sCw10√ol%‰ 200 Time(hr) Fig. 6. Creep characteristics of short-fiber-reinforced SiAlON composites
3.2. Superiorities of CMC CMC has been developed to overcome the brittleness of the monolithic. The thermal shock resistance, particle impact resistance, creep resistance, etc. are important characteristics required for materials which are reliable at high temperatures. 3.2.1. Thermal shock resistance Scrolls are components to which the highest thermal stress (thermal shock) is applied when an engine starts and stops. Circular rings, 100/70 mm OD/ID and 5 mm thick, made to simulate these components and provided with a micro-notch so that the circular rings are easily broken, were exposed to a high-temperature (1350C) and high-speed (100 m/s) gas ¯ow and a cooled gas ¯ow (20 m/s) alternately. As shown in Fig. 4, the long-®ber reinforced CMCs particularly withstood such severe circumstances without any failure and exhibited superiority to the monolithic. Fig. 5. Particle impact test results of CMCs. Fig. 6. Creep characteristics of short-®ber-reinforced SiAlON composites. 866 H. Kaya / Composites Science and Technology 59 (1999) 861±872
H. Kaya/Composites Science and Technology 59(1999)861-872 3.2.2. Particle impact resistance 3.3. Development of components Instant fracture of a turbine rotor caused by foreign- object damage(FOD) has led to the severe damage of Developmental results were shown briefly in Table 2 ceramIc componen use the maximun tip speed of the CGT rotor reached 670 m/s. Zirconia balls with a 3.3.1. Turbine rotor(rotational component of high diameter of I mm(3 mg) were shot into the test pieces strength material) of the various CMCs and the monolithic by using a Turbine rotors were developed by using a carbon gas-pressure gun [12] at a high speed to compare fiber-reinforced silicon-carbide composite, silicon-car he residual strength. As shown in Fi bide-whisker-reinforced Sialon composite and in situ long-fiber-reinforced type CMCs maintained their silicon-nitride- reinforced silicon-nitride composite strength even after the impact test at the above Rotating disks, 120 mm outer diameter(OD)x15 mm maximum tip speed. The da er- thick, were made from the long-fiber and short-fiber reinforced type of CMC was confined to part of the reinforced types of CMCs and, as shown in Fig. 7,a area, as opposed to the total loss of the monolithic. As hub model capable of generating a stress equivalent to clarified above, the long-fiber-reinforced CMC was that of a rotor with blades was made from the long- confirmed to be extremely resistant to damage and this fiber-reinforced CMC. With these models, the functions property is something like a metallic material to which of the turbine rotor as a high-speed rotating material the fracture mechanics theory for the monolithic can were evaluated by a cold-spin test(CST) not be applied As indicated in the fracture mode of the long-fiber- reinforced type of CMC in Fig. 8, it fractured at the 3.2.3. Creep resistance average tangential stress like a metal. By contrast, the CMCs can be utilized for their creep-resistant char monolithic and the short-fiber-reinforced type CMC acteristics by making use of the characteristics of the rein- were fractured at the maximum tangential stress [17].A forcing material Carbon-fiber-reinforced silicon-carbide comparison is not easy since the test models differed composites have low creep rates of the order of 10-s from the turbine rotor in the amount of stress gener- at 1700C under loads of 200 MPa or more [15]. Also, ated. However, the great advantage in withstanding the as shown in Fig. 6, it was confirmed that silicon-car- stress required for the turbine rotor can be seen from bide-whisker-reinforced SiAlON composite had con- the results in the fracture model and also from the fact siderably lower deformation characteristics at 1200 C at that the carbon-fiber-reinforced silicon-carbide com- a load of 200 MPa compared with the monolithic even if posite withstood the CST in which the rotation speed the amount of whisker was 10 vol% or less. This allows was up to 90000 rpm for a rotating disk, which was the pressureless sintering and allows for achievement of same as or greater than that of an actual rotor made both high flexural strength and high fracture toughness from a monolithic silicon nitride. Therefore, it was confirmed that the long-fiber-reinforced CMc has Rote Hub-model
3.2.2. Particle impact resistance Instant fracture of a turbine rotor caused by foreignobject damage (FOD) has led to the severe damage of ceramic components because the maximum tip speed of the CGT rotor reached 670 m/s. Zirconia balls with a diameter of 1 mm (3 mg) were shot into the test pieces of the various CMCs and the monolithic by using a gas-pressure gun [12] at a high speed to compare the residual strength. As shown in Fig. 5, particularly long-®ber-reinforced type CMCs maintained their strength even after the impact test at the above maximum tip speed. The damage of the long-®berreinforced type of CMC was con®ned to part of the area, as opposed to the total loss of the monolithic. As clari®ed above, the long-®ber-reinforced CMC was con®rmed to be extremely resistant to damage and this property is something like a metallic material to which the fracture mechanics theory for the monolithic can not be applied. 3.2.3. Creep resistance CMCs can be utilized for their creep-resistant characteristics by making use of the characteristics of the reinforcing material. Carbon-®ber-reinforced silicon-carbide composites have low creep rates of the order of 10ÿ9 sÿ1 at 1700C under loads of 200 MPa or more [15]. Also, as shown in Fig. 6, it was con®rmed that silicon-carbide-whisker-reinforced SiAlON composite had considerably lower deformation characteristics at 1200C at a load of 200 MPa compared with the monolithic even if the amount of whisker was 10 vol% or less. This allows pressureless sintering and allows for achievement of both high ¯exural strength and high fracture toughness [16]. 3.3. Development of components Developmental results were shown brie¯y in Table 2. 3.3.1. Turbine rotor (rotational component of high strength material) Turbine rotors were developed by using a carbon- ®ber-reinforced silicon-carbide composite, silicon-carbide-whisker-reinforced SiAlON composite and in situ silicon-nitride-reinforced silicon-nitride composite. Rotating disks, 120 mm outer diameter (OD)15 mm thick, were made from the long-®ber and short-®ber reinforced types of CMCs and, as shown in Fig. 7, a hub model capable of generating a stress equivalent to that of a rotor with blades was made from the long- ®ber-reinforced CMC. With these models, the functions of the turbine rotor as a high-speed rotating material were evaluated by a cold-spin test (CST). As indicated in the fracture mode of the long-®berreinforced type of CMC in Fig. 8, it fractured at the average tangential stress like a metal. By contrast, the monolithic and the short-®ber-reinforced type CMC were fractured at the maximum tangential stress [17]. A comparison is not easy since the test models diered from the turbine rotor in the amount of stress generated. However, the great advantage in withstanding the stress required for the turbine rotor can be seen from the results in the fracture model and also from the fact that the carbon-®ber-reinforced silicon-carbide composite withstood the CST in which the rotation speed was up to 90000 rpm for a rotating disk, which was the same as or greater than that of an actual rotor made from a monolithic silicon nitride. Therefore, it was con®rmed that the long-®ber-reinforced CMC has Fig. 7. Stress distributions generated in a rotor and a hub model. H. Kaya / Composites Science and Technology 59 (1999) 861±872 867
H. Kaya/Composites Science and Technology 59(1999)861-872 a1200 Maximum Stress 600 Average Stress 400F 1250℃ Room Temp 020406080100120140 Rotational Speed(x1, 000 rpm) Fig 8. Fracture concept of long-fiber-reinforced type CMC. excellent potential as a material capable of withstanding This test confirmed that it had excellent thermal resistance, high temperatures and high speed rotation capable of withstanding a temperature of 1600C. The short-fiber-reinforced CMc was broken at the An orifice liner and an extension liner made from maximum tangential stress, as in the case of the mono- long-fiber reinforced type were subjected to an engine lithic. But the results indicated that the in situ silicon- combustion environment test(ECET: combustion gas nitride-reinforced silicon-nitride could have a Weibull of 1350C, 4 atmabs, 100 m/s) to confirm endurance modulus of 20 or more [18] Experimental production of model components of ar actual rotor shape was successful for each type of CMC 3.3.4. Inner scroll support(a stationary component of and a rotor made from in situ silicon-nitride-reinforced high toughness material) silicon-nitride composite confirmed its potential by a Inner scroll supports were developed by using a car hot spin test( HST) at1200°C×70000rpm. bon fiber reinforced silicon carbide composite and a silicon nitride or silicon carbide fiber reinforced 3.3.2. Back plate(a stationary component of high strength material, Each type of CMC was subjected to a hydrostatic Back plates were developed using silicon-carbide- pressure loading test as shown in Fig 9, in which a force whisker reinforced SiAlON and in situ silicon-nitride- simulating the thermal stress was generated, as a reinforced silicon-nitride composites. A high-tempera- screening test prior to a high temperature stationary ture stationary component assembly test(endurance test component assembly test(HSAT). Each type of CMC for actual components as screening test) was performed withstood a screening stress(120% of the expected for around 30 h under the same conditions (TIT: maximum thermal stress), showing the possibility of 1350C, 5 atmabs, 100 m/s) as for an actual CGt application to actual components. In particular, the engine. The model components withstood the endur- generated thermal stress of carbon fiber reinforced ance test without problem. These CMCs were proved to be almost complete materials, showing that they can be ed in actual component Oil Inlet Inner Scroll Support 3.3.3. Orifice liner and extension liner( component of high temperature and oxidation resistant These components were developed by using a TiB,/ milled carbon fiber reinforced silicon carbide composite and silicon carbide fiber reinforced silicon carbide and ilicon carbide fiber reinforced SiN-C composites. forced type was subjected to a high temperature stationary 50mm component assembly test similar to that for the back plate Fig. 9. Hydrostatic pressure loading test for the inner scroll support
excellent potential as a material capable of withstanding high temperatures and high speed rotation. The short-®ber-reinforced CMC was broken at the maximum tangential stress, as in the case of the monolithic. But the results indicated that the in situ siliconnitride-reinforced silicon-nitride could have a Weibull modulus of 20 or more. Experimental production of model components of an actual rotor shape was successful for each type of CMC and a rotor made from in situ silicon-nitride-reinforced silicon-nitride composite con®rmed its potential by a hot spin test (HST) at 1200C 70000 rpm. 3.3.2. Back plate (a stationary component of high strength material) Back plates were developed using silicon-carbidewhisker reinforced SiAlON and in situ silicon-nitridereinforced silicon-nitride composites. A high-temperature stationary component assembly test (endurance test for actual components as screening test) was performed for around 30 h under the same conditions (TIT: 1350C, 5 atmabs, 100 m/s) as for an actual CGT engine. The model components withstood the endurance test without problem. These CMCs were proved to be almost complete materials, showing that they can be used in actual components. 3.3.3. Ori®ce liner and extension liner (a stationary component of high temperature and oxidation resistant material) These components were developed by using a TiB2/ milled carbon ®ber reinforced silicon carbide composite and silicon carbide ®ber reinforced silicon carbide, and silicon carbide ®ber reinforced SiNÿC composites. An ori®ce liner made from particle/milled ®ber reinforced type was subjected to a high temperature stationary component assembly test similar to that for the back plate. This test con®rmed that it had excellent thermal resistance, capable of withstanding a temperature of 1600C. An ori®ce liner and an extension liner made from long-®ber reinforced type were subjected to an engine combustion environment test (ECET: combustion gas of 1350C, 4 atmabs, 100 m/s) to con®rm endurance [18]. 3.3.4. Inner scroll support (a stationary component of high toughness material) Inner scroll supports were developed by using a carbon ®ber reinforced silicon carbide composite and a silicon nitride or silicon carbide ®ber reinforced SiÿNÿC composite. Each type of CMC was subjected to a hydrostatic pressure loading test as shown in Fig. 9, in which a force simulating the thermal stress was generated, as a screening test prior to a high temperature stationary component assembly test (HSAT). Each type of CMC withstood a screening stress (120% of the expected maximum thermal stress), showing the possibility of application to actual components. In particular, the generated thermal stress of carbon ®ber reinforced Fig. 8. Fracture concept of long-®ber-reinforced type CMC. Fig. 9. Hydrostatic pressure loading test for the inner scroll support. 868 H. Kaya / Composites Science and Technology 59 (1999) 861±872
H. Kaya/Composites Science and Technology 59(1999)861-872 Table 3 Testing methods for evaluation of characteristics of CMc Standard No Title .TS CMC Test method for tensile stress-strain behavior of continuous fiber reinforced ceramic matrix composites at room and elevated temperature PECTS CMC 02 Test method for tensile strength of whisker and / or particulate reinforced ceramIc m and eleva PECTS CMC 03 Test method for apparent tensile strength of whisker and/or particulate reinforced ceramic matrix composites with diametrically comp O-ring specimens PEC-TS CMC Test method for flexural strength of conti composites at room and elevated temper PEC-TS CMC Test method for flexural strength of whisker and /or particulate reinforced ceramic matrix composites at room and elevated temperatures PEC-TS CMC 06 Test method for shear strength of continuous fiber reinforced ceramic matrix mposites at room and elevated temperatures PEC-TS CMC Test method for fracture toughness of whisker and/or particulate reinforced PEC-TS CMC 08 Test method for fracture toughness of continuous fiber reinforced ceramic matrix PEC· IS CMO09 Test method for fracture energy of continuous fiber reinforced ceramic matrix PEC-TS CMC 10 Test method for tensile-tensile cyclic fatigue of continuous fiber reinforced ceramic matrix composites at room and elevated temperatures PEC-TS CMC 11 Test method for tensile creep behavior of continuous fiber reinforced ceramic matrix composites at elevated temperature PEC-TS CMC 12 Test method for flexural creep behavior of whisker and/ or particulate reinforced ceramic matrix composites at elevated temperatures PEC-TS CMC 13 Test method for elastic modulus of ceramic matrix composites at room and elevated temperatures PEC.TS CMC 14 Test method for oxidation resistance of non-oxide ceramic matrix composities at elevated temperatures Fabric P Preparation Laminating Bagging 會 Firing Vacuum bagging PCP and Curing Pressure Firing Product Impregnation nd curi Fig. 10. Near-net-shape process using pre-ceramic polymer-impregnation method
Table 3 Testing methods for evaluation of characteristics of CMC Standard No. Title PEC-TS CMC 01 Test method for tensile stress±strain behavior of continuous ®ber reinforced ceramic matrix composites at room and elevated temperatures PEC-TS CMC 02 Test method for tensile strength of whisker and/or particulate reinforced ceramic matrix composites at room and elevated temperatures PEC-TS CMC 03 Test method for apparent tensile strength of whisker and/or particulate reinforced ceramic matrix composites with diametrically compressed O-ring specimens PEC-TS CMC 04 Test method for ¯exural strength of continuous ®ber reinforced ceramic matrix composites at room and elevated temperatures PEC-TS CMC 05 Test method for ¯exural strength of whisker and/or particulate reinforced ceramic matrix composites at room and elevated temperatures PEC-TS CMC 06 Test method for shear strength of continuous ®ber reinforced ceramic matrix composites at room and elevated temperatures PEC-TS CMC 07 Test method for fracture toughness of whisker and/or particulate reinforced ceramic matrix composites PEC-TS CMC 08 Test method for fracture toughness of continuous ®ber reinforced ceramic matrix composites PEC-TS CMC 09 Test method for fracture energy of continuous ®ber reinforced ceramic matrix composites PEC-TS CMC 10 Test method for tensile-tensile cyclic fatigue of continuous ®ber reinforced ceramic matrix composites at room and elevated temperatures PEC-TS CMC 11 Test method for tensile creep behavior of continuous ®ber reinforced ceramic matrix composites at elevated temperatures PEC-TS CMC 12 Test method for ¯exural creep behavior of whisker and/or particulate reinforced ceramic matrix composites at elevated temperatures PEC-TS CMC 13 Test method for elastic modulus of ceramic matrix composites at room and elevated temperatures PEC-TS CMC 14 Test method for oxidation resistance of non-oxide ceramic matrix composities at elevated temperatures Fig. 10. Near-net-shape process using pre-ceramic polymer-impregnation method. H. Kaya / Composites Science and Technology 59 (1999) 861±872 869
H. Kaya/Composites Science and Technology 59(1999)861-872 silicon carbide type CMC reached a level (0.9 MPa) in Table 3. These methods are expected to be the basis 1/100 of that of an actual component made from of the formal test method in JIS, ISO, ASTM, DIN, monolithic silicon nitride, indicating the superiority of etc. CMC. The silicon carbide fiber reinforced Si-N-O composite also endured the stress up to 156%(180 3.5. Basic technology for the production of CMC MPa)[19 Other than the development of the materials, basic 3.4. Testing standard methods for CMC technologies such as molding, processing, etc. ar required to make components from CMC. In this pro- Testing standard methods for evaluating the char- ject, the development of the following CMC material acteristics of CMC were still not established at the time and base technologies for production of CMc were this project started functioning, therefore, the develop- greatly advanced ment of the testing standard methods for evaluating Examples of the basic technologies for manufacturing ne mechanical properties of CMC was carried out in a material include a technology for manufacturing and parallel with the development of CMC materials and utilizing 2D/3D cloths taking into consideration the components. As a result, the testing standard methods anisotropy of the reinforcing material; technology for consisting of 14 sections were proposed [20] as shown improving the oxidation resistance of a fiber surface Cu Electrode Orifice Liner Cu Electrode Cu Electrode Fig. 11. Manufacturing of orifice liner by electric discharge machining method. 00 mm Fig. 12. Manufacturing of turbine rotor by green machining method
silicon carbide type CMC reached a level (0.9 MPa) 1/100 of that of an actual component made from monolithic silicon nitride, indicating the superiority of CMC. The silicon carbide ®ber reinforced SiÿNÿC composite also endured the stress up to 156% (180 MPa) [19]. 3.4. Testing standard methods for CMC Testing standard methods for evaluating the characteristics of CMC were still not established at the time this project started functioning, therefore, the development of the testing standard methods for evaluating the mechanical properties of CMC was carried out in parallel with the development of CMC materials and components. As a result, the testing standard methods consisting of 14 sections were proposed [20] as shown in Table 3. These methods are expected to be the basis of the formal test method in JIS, ISO, ASTM, DIN, etc. 3.5. Basic technology for the production of CMC Other than the development of the materials, basic technologies such as molding, processing, etc. are required to make components from CMC. In this project, the development of the following CMC materials and base technologies for production of CMC were greatly advanced. Examples of the basic technologies for manufacturing a material include a technology for manufacturing and utilizing 2D/3D cloths taking into consideration the anisotropy of the reinforcing material; technology for improving the oxidation resistance of a ®ber surface Fig. 11. Manufacturing of ori®ce liner by electric discharge machining method. Fig. 12. Manufacturing of turbine rotor by green machining method. 870 H. Kaya / Composites Science and Technology 59 (1999) 861±872
