STATUS OF THE AUTOMOTIVE CERAMIC GAS TURBINE DEVELOPMENT PROGRAM -YEAR 2 PROGRESS-

Transcription

1 THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS 345 E. 47th St., New York, N.Y GT-40 The Society shall not be responsible for statements or opinions advanced in papers or discussion at meetings of the Society or of its Divisions or Sections, or printed in its publications. Discussion is printed only if the paper is published in an ASME Journal. Papers are available from ASME for 15 months after the meeting. Printed in U.S.A. Copyright 1993 by ASME STATUS OF THE AUTOMOTIVE CERAMIC GAS TURBINE DEVELOPMENT PROGRAM -YEAR 2 PROGRESS- Takane Itoh Ceramic Gas Turbine R&D Division Japan Automobile Research Institute, Inc. Tsukuba, lbaraki Japan Hidetomo Kimura CGT Development Office Petroleum Energy Center Minato-ku, Tokyo Japan ABSTRACT A seven-year program, designated "Research & Development of Automotive Ceramic Gas Turbine Engine (CGT Program)", was started in June 1990 with the object of demonstrating the advantageous potentials of ceramic gas turbines for automotive use. This CGT-Program is conducted by PEC with the support of MITI. The basic engine is a 100-kW, single-shaft engine having a turbine inlet temperature of a 1350t and a rotor speed of 110,000 rpm. During the second year of the program, experimental evaluation of the various components was started, including a centrifugal compressor, a radial turbine rotor, a high speed rotor system and initial ceramic hot parts. Cold and hot spin testing of ceramic rotors from three different ceramic suppliers was also initiated. INTRODUCTION Presently, Japan is dependent upon foreign countries for over 80% of its total energy requirements and almost all of its petroleum needs. Therefore, the particular sector of industry where prompt improvements must be introduced is in the transportation sector because of its high ratio of dependency on petroleum and its inefficient use of energy resources. From the standpoint of ensuring a stable supply and the effective utilization of energy resources, great expectations have been placed upon the development of an automotive engine which allows the us- of a variety of fuels without any constraints associated with octane or cetane value, in addition to ensuring high thermal efficiency and clean exhaust emissions. In this direction the ceramic gas turbine is believed to be one of the most promising candidates. In consideration of these circumstances, the Petroleum Energy Center(PEC) started, in 1990, the program designated "Research & Development of Automotive Ceramic Gas Turbine Engine (CGT Program)" with the active cooperation of the petroleum, automobile, ceramics, and related industries, and with the financial support of the Agency of Natural Resources and Energy, the Ministry of International Trade and Industry.' The major aims of the program are to cope with fuel diversification and environmental problems within Japan's transportation sector. The goal is to develop an experimental ceramic gas turbine engine as a prospective automotive engine and verify by experimentation its feasibility, as described below: - Ability to operate on a variety of fuels - Greatly increased thermal efficiency - Contribution to ensuring a cleaner environment In order to achieve these goals, research and development efforts are presently under way in the four areas shown below. (1) Research and development of combustion technologies for diverse fuels (2) Research and development of technologies for fabricating components from high-temperature ceramic composite materials (3) Research and development of high temperature lubricating oils technologies (4) Research and development of an automotive ceramic gas turbine engines Of the areas in above-mentioned program, the Japan Automobile Research Institute, Inc.(JARI) is responsible for the fourth one; that is the design, fabrication of an experimental ceramic gas turbine engine and the subsequent evaluation of its performance. This paper describes an overview of the activities carried out along these lines by JARI during the second year of the program. ENGINE STRUCTURE Fig. 1 shows the cutaway of the experimental CGT engine, which is rated at 100kW with a maximum thermal efficiency of 40%. The single-shaft rotating group is composed of a centrifugal compressor, radial turbine and output shaft, supported by oil lubricated ball and roller bearings. Maximum engine speed is Presented at the International Gas Turbine and Aeroengine Congress and Exposition Cincinnati, Ohio May 24-27, 1993

2 FUEL INJECTION NOZZLE (secondary) FUEL INJECTION NOZZLE (primary) COMBUSTION CHAMBER TURBINE NOZZLE VIIIIL \ 1 r CVT REDUCTION GEAR BOX CENTER BELLOWS REAR SPRING BULK HEAD TURBINE ROTOR, OUTER SCROLL.4( t, \dmi,: d IR., 111_ ,11=1 =1111TOMMIIII " `, ( NN.,. Fig. 1 CUTAWAY OF THE CGT ENGINE FRONT SPRING i N'''.."1* 1:1-J.1... I \ I r vit i' r \ Poi jw' I :: _ N172 4=06_ \\ - 7 REAR BELLOWS INNER SCROLL SUPPORT 110,000 rpm. The combustion system features a variable geometry, premixed, pre-vaporizing lean-burn combustor. Turbine Inlet Temperature is 1350r. The gases from the turbine diffuser are ducted through the low pressure side of the rotary regenerator core, equipped on each side of the engine, to the engine exhaust. Output power from the gas turbine is delivered to the vehicle via a twostage reduction gear box and an infinitely variable transmission, which is required for the optimization of a single-shaft gas turbine for automotive use. The engine configuration incorporates ceramic structures in the hot flow path, including the combustor and the turbine rotor. Engine design was investigated with emphasis on the structure which would minimize air leakages from interfaces between ceramic components and also from regenerator seals. Inside the bulkhead, a seal mechanism is provided for preventing the leakage of high-pressure air to the low-pressure turbine diffuser outlet. This seal mechanism also serves as a support system for elastically supporting the stacked ceramic stationary component assembly. Fig. 2 show the structure of the high pressure seal mechanism. INITIAL TESTS FOR ENGINE COMPONENTS Compressor The compressor system is composed of a single-stage backswept centrifugal impeller, a single stage radial diffuser and an upstream set of radial type variable inlet guide vanes(vigv). Performance predictions for this compressor system at maximum power operating condition and the geometric design summary for the compressor of the primary design is shown in Table 1. Compressor development testing was conducted on test rigs suitable to aerodynamically evaluate the impeller, diffuser, VIGVs, inlet plenum and exit scroll. First, two kinds of tests were conducted on the compressor. One was testing on the VIGV system and the other was on the compressor stage without the VIGV set. Fig. 3 shows the radial distributions of flow angle at the inlet of an impeller for the three vane setting angles. Although some reverse swirl was noted for the setting angle of 0 degree, the target swirl from -10 to 60 degrees could be obtained, and the pressure loss could be restrained to a level below the target value. FRONT BELLOWS REAR SEAL P TE TURBINE DIFFUSER FRONT SEAL PLATE (CERAMICS) Fig. 2 SEAL AND SUPPORT STRUCTURE Table 1 COMPRESSOR GEOMETRIC PRIMARY DESIGN SUMMARY AIR FLOW RATE = kg/sec., PRESSURE RATIO = 5. 0, SYSTEM EFF. = 81% IMPFI I P12, DIFFUSER Inlet Hub/Tip Ratio Inlet Diameter mm Inlet Hub Diameter 27.0 mm Outlet Diameter 144 mm Inlet Sweep Angle deg. Inlet B Angle 17 deg. Exit Diameter 95 mm Blade Height 4.33 mm Exit Blade height 4.69 mm Numbas of Blade 25 Exit Blade Angle 55 deg. During the first compressor test(without VIGVs), compressor pressure ratio and system efficiency were below design intent In order to reduce impeller blade loading, the number of blades was increased from 18 to 20, and the blade angle distribution and hub line were modified within available hardware constraints. As a result, the maximum efficiency and pressure ratio improved slightly to 77.0% and 4.84 respectively but the pressure ratio, efficiency and surge margin still failed to reach the desired target values. Based on these test results and subsequent analysis of test data, a modification to the diffuser and scroll are also being conducted. Turbine The turbine is a single stage radial inflow design. Table 2 shows the specifications for the turbine rotor and stator of the primary design. Using a metallic stator and a rotor which duplicated the ceramic engine components of the primary design, aerodynamic performance tests were conducted in the cold turbine test rig. The turbine rotor was tested with three turbine stator designs, 2

3 VA : VANE SETTING ANGLE (DEGREES) (TIP SIDE) (HUB SIDE) 20 ao' 60' FLOW ANGLE Fig. 3 IMPELLER INLET FLOW ANGLE DISTRIBUTION Table 2 TURBINE GEOMETRIC PRIMARY DESIGN SUMMARY GAS FLOW RATE = kg/sec., EXPANSION RATIO = 4. 25, SYSTEM EFF =87.5% STATOR ROTOR Inlet Diameter trim Inlet Diameter 127 mm Outlet Diameter Inlet B-Height 10.4 mm Vane Height 10.4 Mal Exit Diameter 92 mm Numbers of Vane 21 Exit Hub Dia ram Numbers of Blade CORRECTED GAS FLOW RATIO Fig. 4 SYSTEM EFFICIENCY FOR THE PRIMARY TURBINE SECONDARY-FUEL INJECTOR PRIMARY-FUEL INJECTOR PRE-MIXED, PRE-VAPORIZING ZONE including one with a throat area 12% smaller than the baseline stator and one with a throat area 12% larger. Fig. 4 shows an example of system efficiency characteristics as a function of corrected flow rate at the rated speed. The system is rated from combustor exit-to-turbine diffuser exit. In the case of the baseline design stator b, the efficiency was 85% at the design corrected speed. The higher efficiency of 863% was obtained with stator c which provided a 16% larger flow rate. It was found that the optimum flow rate for this turbine was offset to a larger side. Therefore, the stator vane setting angle and the throat area were reduced. At the same time, the turbine rotor outlet blade angle was reduced by 1 degree and the blade outlet outer diameter was also modified so as to reduce the throat area by 10%. Three dimensional flow analysis showed the need of modifications such as increasing the radius of curvature of the shroud line of the rotor for improving velocity distribution. The test results of the modified turbine showed that the interim target value of 86% at the design speed could be successfully achieved. Combustor In the CGT program, emphasis has been placed on gaseous emission, multifuel capability, and fuel economy in an automotive driving cycle. In response to these requirements, and based on experience in previous research, a variable geometry, premixed, pre-vaporizing lean combustion concept(ppl) was introduced for the CGT combustor. FLAME HOLDER Fig. 5 COMBUSTOR SUB-ASSEMBLY The combustor, shown in Fig. 5, is featured by the combined use of the PPL system and diffusion combustion system. More specifically, the PPL combustion system is used from idling to 30% of the rated load, and the mixed combustion of the PPL combustion system and the diffusion combustion system is used from 30% to full rated load. As seen in Fig. 5, the combustor is composed of a premixed, pre-vaporizing zone, a primary lean combustion zone immediately followed by a dilution/quenching zone, a flame holder, combustor assembly supporting parts, and two fuel injection nozzles. The combustor is multiple piece construction, and all of the components are made of sintered silicon nitride except for the primary combustion liner and the following orifice liner which are made of sintered silicon carbide. Both the primary and secondary fuel injection nozzles are of the air-assisted type, and they met the required targets of mean drop size, drop size distribution and cone angle in cold rig testing. 3

4 Initial performance testing for the combustor was conducted in the combustion test rig. The pattern factor at the exit of the combustor was 9.5%, which was larger than expected. The initial emission index for NOx was 3.67 g/kg -Fuel. Although the emission levels were about a fifth of the value of a general diffusion flame combustion system, they were larger than the expected levels. The combustion sub-assembly was disassembled from the test rig and was checked. The following was observed : 1) There were marks indicating that flames had exited in the swirl chamber of the premixed, pre-vaporizing zone. 2) The variable geometry flame holder(bluff body) had been damaged at its neck. 3) A gap had been produced between the ceramic liners of the primary lean combustion zone. From these findings, it appears the reason why the NOx emissions were much larger than the expected values is that substitute metallic holders had been deformed in the hot environment and allowed a gap to be created in the lean combustion zone. A consequent enriched mixture and non uniform flow in the swirl chamber of the premixed, pre-vaporizing zone, resulted in generation of a flame in the partially fuel enriched portion. Preparations are under way to resume performance tests by replacing the substitute metallic holder parts with genuine ceramic holder parts. Regenerator System The regenerator system, as shown in Fig. 6, consists of a regenerator core matrix, inner and outer seals, a drive system and housings. In the development of a low leakage seal system, a major objective was to achieve sufficient contact force to provide effective sealing while maintaining coating wear at a uniform and acceptable rate and without excessive driving torque requirements. Therefore, the following test rigs were fabricated and utilized: 1) Regenerator wear test rig to evaluate coating and wear characteristics. 2) Cold static seal leakage rig to evaluate system leakage characteristics related to the mechanical load. In the wear test rig, a test sample with wear coating on one face is loaded against the rotating model ceramic core having a diameter of 230 mm. The coefficient of friction can be determined under an elevated temperature by measuring the reaction force. Samples evaluated in the rig, have not yet demonstrated friction and wear characteristics with acceptable criteria. The seals consist of static metallic diaphragm seals and ceramic coated dynamic seals. The diaphragms were formed from foil material 0.1 mm thick. The dynamic seals are composed of peripheral shoes and a crossarm plasma-sprayed on the surface with a NiCr undercoat and a top coat material. Using this seal assembly, a leakage test was conducted in the cold static seal leakage rig to check the relationship between air leakage, operating pressure and seal working height. It was found that at room temperature the leakages fell within tolerable limits. The regenerator core is driven through a ring gear and a pinion, and is supported by support rollers. The bearings for the pinion and rollers are composed of solid lubricated bushings made of graphite. An initial durability test on the driving system was conducted by Fig. 6 REGENERATOR SYSTEM setting the drive torque at 16kg m and the core ambient temperature at 210t. It was found that the graphite bushings suffered excessive wear. Efforts are now under way to take corrective action for solving this problem. In order to determine the characteristics of the regenerator assembly as a whole, the regenerator hot rig was assembled and an initial evaluation test was conducted. Because of excessive driving torque, the cold side air pressure could not be set to the design condition. Also, air leakage was much larger than expected. The unit was disassembled and it was found that the ceramic core rubbing surface had become roughened. It had also become rubbed and contacted areas on the crossarm surfaces of the inner and the outer seals worse than expected. While efforts are now under way to locate the causes, it is likely that the axial thermal growth and distortion of the metallic test rig housing, the sliding coefficient of friction, and the seal shoe pressure force had been larger than design values. Rotor System The high speed rotating shaft is supported by a ball bearing between the impeller and turbine rotor and a roller bearing in front of impeller. It is coupled by use of a curvic coupling in the back plane, and tightened by the tie shaft at the center of the shaft. Since the ball bearing has a high DN value and is also exposed to a high temperature, it must be effectively lubricated and cooled with a small amount of lubricant. For this purpose, the lubricant is supplied to the ball bearing through the holes in the inner race. The reduction gear input shaft is coupled with the compressor shaft end by a spline coupling, and the reduction gear ratio is In order to test the dynamic characteristics and power loss of the rotor system and to optimize the lubricant supply system, a rotor dynamic test rig was manufactured. Blade-less rotors with the same mass and moment of inertia to simulate an actual impeller and a turbine rotor respectively are mounted on the turbine shaft. They are driven at predetermined speeds by the drive air turbine mounted on the turbine end. The test results showed that the 1st and 2nd critical speeds, which were estimated to be 24,000 and 72,000 rpm respectively, were highly damped, and the 3rd critical speed occurred above the rated speed. 4

5 Table 3 CGT COMPONENT AND MATERIAL SUMMARY TURBINE ROTOR BACK PLATE Fig. 7 CERAMIC STATIONARY COMPONENT SUB-ASSEMBLY INITIAL EVALUATION OF CERAMIC COMPONENTS Fig. 7 shows the ceramic stationary component assembly. Table 3 shows the materials used for the individual ceramic components and their forming process. Ceramic structural components received from suppliers are qualified before the engine testing phase according to the flow chart shown in Fig.8. First is the fast fracture 3-point or 4-point flexure test of specimens cut from a fabricated component. Its strength is then compared with the strength of test specimens of the same materials to check whether the original material strength has been reproduced. If its strength is too low, the processing has to be improved. Next, each component is tested in mechanical test rigs for evaluating its over load capability. The mechanical tests are of two types, an internal pressure and a local mechanical load test. The internal pressure test is performed by applying an internal pressure to check a component as a whole for large defects that might have occurred during the manufacturing process. The local stress test is performed by applying a mechanical stress of about 120% of the worst mechanical or thermal maximum principal stress encountered in engine operation. The load is applied at the location where the maximum stress is to be generated. The components that pass these mechanical tests are put together as an assembly and are further tested at elevated temperature by use of the static structure test rig. For evaluating the strength of a turbine rotor, cold and hot spin tests were primarily performed. In addition to these tests, 4-point flexure strength tests of the cut-out specimens were performed. The following is a summary of the initial tests for each ceramic component. Photo 1 shows the main ceramic stationary components. Turbine Rotorm Three types of sintered silicon nitride materials ( SN- 252, NGK SN-90 and NTK EC-125) were selected for the initial turbine rotors. The outer diameter is 127 mm, and the rated speed is 110,000rpm. All rotors delivered had previously passed the proof COMPONENTS SUPPLIER MATERIAL PROCESS Combustor PP-Swider PP-Scroll VG-Nozzle Comb-Uner Orifice-Liner Dilution-Uner Holder Outer-Scroll Inner-Scroll Inner-Shroud Outer-Shroud NTK NIX NIX Asahi Glass Asahi Glass NGK NGK NGK EC-152 EC-152 EC-152 C-600A C-600H SN-252 SN-88 Injection Molded Injection Molded Injection Molded Cold Impressed Cold Impressed Shp Casting Cold Isopressed Slip Casting Slip Casting Cold lsopressed Cold Impressed SN-252 SN-252 SN-88 SN-88 Turbine Stator NGK SN-88 Injection Molded Turbine Rotor SN-252 Slip Casting NGK SN-90 Inj. Molded + CIP NIX EC-152 Ini. Molded + CIP Turbine Backplate NIX EC-152 Cold lsopressed Outer-Scroll Support Inner-Scroll Support NTK SN-252 EC-152 Slip Casting Cold Impressed Seal Plate Hitachi KC. HXL Cold lsopressed Regenerator Seal Platform Hitachi K.C. SN-220 HXL Slip Casting Cold lsopressed Regenerator Core NGK MAS Extrusion FLEXURE STRENGTH TEST JIS-TEST SPECIMENS FLEXURE STRENGTH TEST CUT-OUT SPECIMENS MECHANICAL LOADED COMPONENT PROOF TEST INNER PRESSURE LOAD CONCENTRATED STRESS STATIC STRUCTURE THERMAL TEST I ENGINE TEST I FLEXURE STRENGTH TEST CUT-OUT SPECIMENS FROM USED COMPONENTS Fig. 8 CERAMIC STATIONARY COMPONENT QUALIFICATION SEQUENCE tests of 90,000rpm at each supplier. Flexure strength tests on test specimens. To check the material strength, 4-point flexure strength tests on JIS -size test specimens made from CIP processing were conducted for each material at room temperature, 1000 t, 1200 t and 1400 t. Next, 5

6 SN-252 SN-90 - hr-i& EC-152 m : Weibull Modulus 10 1 Photo 1 MAIN CERAMIC STATIONARY COMPONENTS c Maximum Centrifugal Stress (MPa) Fig. 10 WEIBULL PLOTS OF ROTOR COLD SPIN TEST 1.1 ex x 2 x 38 Test Specimen Cut from Outer Portion A Cut from Middle Portion X Cut from Center Portion SN-252 SN-90 EC-152 showed that the fracture occurred at a blade and that failure mode was obviously different from the other rotors. On the other six rotors, fracture speeds ranged from 102,000 to 124,800rpm. The average fracture speed was 111,000rpm, and the Weibull modulus was 6.1. Seven SN-90 rotors were tested. Their fracture speeds were from 116,000 to 125,000rpm. The average fracture speed was 119,000rpm and the Weibull modulus was Seven EC-152 rotors were tested. Two of them were fractured at very low speeds because of troubles in the test equipment. The fracture speeds of five rotors were from 130,500 to 138,900rpm. The average fracture speed was 135,500rpm and the Weibull modulus was Fig. 9 FLEXURAL STRENGTH OF TEST SPECIMENS CUT FROM ROTOR test specimens were cut from each of the three turbine rotors and were evaluated using 4- point flexure strength tests at 1000 r, which was estimated as the mean material temperature of the rotor operated in the engine. The cut-out test specimen strength decreased by 5 to 10%, compared with the strength of ITS test specimen at 1000 t, as shown in Fig. 9. Cold Spin Tests. Fig 10 shows the Weibull plots of cold spin tests. Seven, SN-252 rotors were tested, but one of them burst at a speed lower than the proof test speed. It was excluded from the Weibull plots because the photograph at the instnt of the burst Hot Spin Tests. In the hot spin test rig, hot gases from the combustor are directly fed to the scroll, and drive a rotor until fracture occurs. Three rotors for each material were subjected to the hot spin test at a turbine inlet temperature of 1200 C. The fracture speeds were 111,000, 112,000 and 124,000ipiit for three SN-252 rotors, and 124,700 for one SN-90 rotor and 101,000, 102,000rpm for two EC-125 rotors. The tests for the remaining four rotors were interrupted because vibration levels of the rotor system exceeded allowable limits. While the number of hot spin tested rotors are limited, the highest fracture speed obtained was 124,700rpm. This speed was about the same as the results of the cold spin tests, and the tip speed at the fracture was 829 m/sec. Turbine Stator The turbine stator consists of six segments, shown in Photo 2, for ease of fabrication. Each of the segments is fabricated from 6

7 Photo. 2 TURBINE STATOR injection molded silicon nitride, and the six segments are placed together to constitute a stator assembly. The dimensions of the delivered turbine stator segments were found to be within tolerable limits, and did not present any dimensional problem. The FEM stress analysis showed that high thermal stresses were created at the vane trailing edge during a cold start condition. For this reason, the thermal shock test rig for the stator was designed and fabricated. In this rig, shown in Fig. 11, the stator can be rotated 180 degrees and subjected to hot gases and cold air alternately. Hot gas temperature, repetitive cycles, etc. are set to produce the conditions that simulate the thermal shock in the actual engine. Durability cyclic tests have just started using this test rig. Turbine Shroud / Turbine diffuser Earlier FEM stress analysis showed that the inner shroud integrated with turbine diffuser suffered considerable thermal stresses under cold start conditions. Therefore, boundary conditions were checked and FEM stress analysis were conducted. A maximum stress of 168 MPa is created at 28 seconds after light off. The failure probability after 10,000 starts is 6.0*104 which satisfied the current target. A strength test of the inner-shroud was performed by applying an axial load which was twice the expected load in the engine. The stresses, measured by strain gauges at the maximum stress generating portion identified by FEM analysis, were in close agreement with the results of the analysis. When the stator was assembled to the inner and outer shroud, it was found that the geometric accuracy of the grooves for positioning the sandwiched turbine stator was insufficient and required improvements in its machining method. Combustor Radial load crushing tests were conducted on the ring-shaped test specimens. The specimens were cut from the cylindrical portion of the primary combustor liner which is subjected to the highest temperature. The fracture strength was 350 to 545 MPa, and the Weibull modulus was 8.9. The mean strength, corrected to the 3-point flexure strength of the JS-size test piece in terms of Fig. 11 THERMAL SHOCK RIG FOR TURBINE STATOR effective volume, was 568 MPa. The test specimen had a strength of 94% with respect to the strength of the JIS test piece of the same material which was 605 MPa. Combustion performance tests were also conducted on the initially fabricated combustor, as described before. Turbine Scroll Outer Scroll. The current problem is that its cylindrical inlet portion, into which the end of the combustor dilution liner is inserted, does not provide sufficient dimensional accuracy. The accuracy is being improved day by day through taking corrective efforts for fabricating processes. The FEM stress analysis showed that the maximum stress occurred in the cylindrical inlet portion just downstream of the outlet end of the inserted dilution liner. Therefore, this portion of a scroll was subjected to a concentrated mechanical load test rig. Fracture occurred at a load of about 1/15 of the expected load. The location where the fracture occurred was not the expected location but was near the fixture contact point, as shown in Photo 3. Although a kind of compliant layer had been inserted between the fixtures and the scroll to compensate for the small thickness and poor dimensional accuracy of that area, excessive and concentrated contact stress caused the scroll to fracture. These tests were suspended, because it was thought that no test would satisfy our goal with the component at the current level of dimensional accuracy. Inner Scroll. FEM analysis showed that the maximum stress was expected to occur at the outer circumference of the flange at the turbine outlet side. To verify the overall strength of this component, internally pressured screening tests were conducted. A fracture occurred at a pressure of about 1/2 of the expected pressure, at the flange root portion. This location was not the 7

8 Photo. 3 FRACTURED PORTION OF OUTER SCROLL portion where a maximum stress had been expected under this test. On further examination, the thickness of the area around the fracture origin was found to be only 2/3 of the design dimension. Fabricating process improvements are now being pursued. Static Structure Thermal Test The static structure thermal test rig utilizes some or all of the ceramic stationary components in an assembly of the engine structure, without the turbine rotor. This rig checks the mechanical functioning of the assembly, evaluates the sealing capability, contact loading, and axial and radial pilots. Fig. 12 shows the static thermal test rig. First, tests on inner and outer scrolls were conducted. The other parts were not the engine parts but were dummy parts made from sintered silicon carbide. After four cold starts and one hour and 50 minutes of operation at 1200t combustor outlet temperature, the components were disassembled, and checked by visual and fluorescent penetrant. None of the silicon nitride test components had been damaged, however the simulated dummy shroud and bacicplate fabricated from silicon carbide were damaged. It was also found that the temperature of the inner shroud was almost the same temperature as the surrounding gas temperature. At this temperature, oxidation problems are a concern for the parts made by silicon nitride. SUMMARY (1) As the 7-year program, Petroleum Energy Center started the program "Research and Development of Automotive Ceramic Gas Turbine Engines" in 1990, under the financial support of M. During the second fiscal year, the major engine components and individual test rigs designed in the first fiscal year were fabricated. (2) Initial evaluation tests on the major components were started using fabricated test rigs. Research efforts are now under way for improving some of components. (3) Proof tests on the major ceramic stationary components were started. The inner and outer-scroll were subjected to static structure thermal test, and passed the first proof test. Dimensional accuracy of some large stationary parts with thin thickness were found to be insufficient. (4) Cold and hot spin tests were started for three kinds of rotors. The maximum speeds to fracture were 138,900 rpm in a cold spin test and 124,700 rpm in a hot spin test at 1200 t. Fig. 12 STATIC THERMAL TEST RIG ACKNOWLEDGMENTS The authors are grateful to the Agency of Natural Resources and Energy of MITI for making this research possible, and would like to thank M. Iwai, I. Sakai, M. Sasaki, K. Sugiyama of JARI Team for technical support of the program efforts. We would also like to thank PEC for permitting of publication of this paper. REFERENCES 1) T. Itoh, H. Kimura: "Status of the Automobile Ceramic Gas Turbine Development Program", Transactions of the ASME(ASME Paper 92-GT-2) 2) H.Ogita, H. Kimura: "Strength Evaluation of Ceramic Rotor for Automotive 100kW CGT", Preprint of JSME Conference held in Nagano in Nov.,