THE LIFING OF CRITICAL ROTATING PARTS: SCIENCE OR TECHNOLOGY?

摘要

Traditionally safe life prediction in the engine has been based on component testing. This has the advantage that not only is the material with the correct surface condition tested, but also representative volumes of material are involved. Correlation of specimen and component testing in a DataBank method allows the statistical variation in all fatigue testing to be captured thus allowing a higher fraction of the life achieved in the component test to be utilised in service. To date, however, these methods are semi-scientific, recognising the difficulty of capturing surface condition and volume effects. Advancing the methods for modern materials and conditions requires improvements in stressing methods, while the life prediction once the temperature and stress cycles are evaluated remains based on the experimental measurement of material and surface condition fatigue capability. This paper examines the methods used to predict the safe cyclic life of such components, taking a particular interest in how these methods can be expected to cope with modern requirements which use higher strength materials at higher temperatures to produce engines which meet today's demands for higher fuel efficiency and lower weight.