Rapid Technology Development for High-Speed Compressors and Fans

Abstract

Developing high-speed compressor and fan technology in the gas turbine industry is a costly and time-consuming process. It can take five to ten years for a new technology to reach the required maturity level for use in an engine, with a total program cost in the tens of millions of dollars. This is mainly because of the cost and time needed to conduct high- speed experiments using complex test facilities and processes. This thesis argues that rapid technology development for high-speed compressors and fans can be achieved by simplifying the facility and the process. The approach is important for its potential to unlock creativity allowing designers to ‘play’ at engine-representative conditions. The main body of the thesis is divided into three parts. The first part demonstrates that it is possible to significantly reduce the mechanical complexity of a high-speed compressor aerodynamic test facility while preserving the key fluid dynamic mechanisms being studied. This is achieved by addressing the physical root causes for the complexity, namely the high power input and the high safety requirements, both of which are reduced by an order of magnitude using a combination of downscaling and a novel turboexpander architecture. By doing so, the full Mach number and a half of the full Reynolds number can be attained, ensuring that the test facility remains aerodynamically representative of an engine compressor stage to the TRL5 level. The second part shows that it is possible to reduce the cycle time of the manufacture- build-test process to under 10 working days, as opposed to the conventional cycle time of over 6 months. This is accomplished by analysing the compressor testing process and making changes to eliminate or accelerate the most time-consuming activities. Extensive design changes are made to the test facility in terms of mechanical, instrumentation, and software design simplifications. A formal time trial is then performed, where a new aerodynamic geometry is made and tested in six working days at a cost of less than $7,000. The third part demonstrates rapid technology development in compressors by experimen- tally investigating the aerodynamic effects of compressor leading edge cutbacks. Cutbacks result from field repair of in-service blade damages and their effects are poorly understood. By rapidly testing an array of six rotor geometries, the pressure rise and surge margin erosion caused by the cutbacks are quantified. The results show that the surge margin loss is controlled by the extent of damage in the worst continuous cluster, whereas the pressure loss is controlled by the total amount of flow separation caused by the cutbacks. The results align with the existing low-speed understanding of the topic but discover that the surge margin loss is a strong function of the Mach number. It has been shown that the rapid technology development process can not only generate numerical data but also study the physical origins of aerodynamic phenomena, making it suitable for integration into the iterative design loop and as a tool for fundamental research.