Engine Design Strategies to Maximize Ceramic Turbine Life and Reliability

Abstract

Over a billion dollars have been spent on ceramic turbine research and development, but ceramic turbines "have not yet reached bill-of-materials status" [1]. Mature turbine-grade silicon nitrides (e.g., SN282^ AS800^ SN88^ NT154^) have excellent properties, but studies still cite life, reliability, and cost issues, often concluding that the materials themselves require further improvement. However, under less severe operating conditions, ceramics can excel in all these areas. In an ongoing long-term endurance test at NRL, a low cost porcelain vessel has been subjected to the thermal shock of hot coffee several times daily for ten years-over 5000 startup/shutdown cycles-with no perceptible degradation. Ceramic turbocharger rotors have been manufactured for the notoriously cost-sensitive automotive market at rates exceeding 10,000 per month, achieving good life and reliability in Japanese sports cars. These examples suggest that rather than just seeking higherperformance materials, it might also be fruitful to design gas turbines to create a more benign environment for existing ceramics.Foremost among stated concems is water vapor erosion [2-11], a process in which water in the combustion products reacts with silicon nitride at high temperatures, forming gases that get swept away by the freestream. This erodes the blade surfaces at a rate Engine Design Strategies to Maximize Ceramic Turbine Life and ReiiabilityCeramic turbines have long promised to enable higher fuel efficiencies by accommodating higher temperatures without cooling, yet no engines with ceramic rotors are in production today. Studies cite life, reliability, and cost obstacles, often concluding that further improvements in the materiais are required. In this paper, we assume instead that the problems could be circumvented by adjusting the engine design. Detailed analyses are conducted for two key life-iimiting processes, water vapor erosion and slow crack growth, seeking engine design strategies for mitigating their ejfects. We show that highly recuperated engines generate extremely low levels of water vapor erosion, enabling lives exceeding 10,000 hs, without environmental barrier coatings. Recuperated engines are highly efficient at low pressure ratios, making low blade speeds practical. Many ceramic demonstration engines have had design point mean blade speeds near 550 mis. A CARES/Life analysis of an example rotor designed for about half this value indicates vast improvements in slow crack growth-limited life and reliability. Halving the blade speed also reduces foreign object damage partide kinetic energy by a factor of four. In applications requiring very high fuei efficiency that can accept a recuperator, or in short-life simple cycle engines, ceramic turbines are ready for application today.