DEVELOPMENT OF A TOOL FOR TEMPERATURE ESTIMATION FROM MICROSTRUCTURAL CONDITION OF A NICOCRALY+RE COATING APPLIED ON THE SURFACE OF GAS TURBINE HOT COMPONENTS

摘要

The hot gas path components of gas turbines have to withstand to severe conditions in terms of high temperature oxidation, hot corrosion and creep-fatigue phenomena. The evaluation of components residual life is an important matter for gas turbines producers and the estimation of service temperatures is a key tool for this evaluation. The most diffused methods to estimate service temperatures of gas turbines blades and vanes in Ni based superalloys are related to the microstructural evolution of the dispersed intermetallic phase γ', Ni_3Al. The aim of this work has been the determination of a tool to estimate service temperature on the basis of the microstructural evolutions of a NiCoCrAlY+Re coating. In order to obtain a deep characterisation of the coating after exposure at different durations and temperatures, an extensive experimental test program has been planned. Samples of Ni based superalloys, covered by the investigated coating, have been aged in chamber furnaces in the temperature range 700°C - 1000°C with durations up to 20000 hours. The microstructure of this coating is characterised by β phase, NiAl, which is the Al reservoir, embedded in the matrix, that is constituted by γ' phase at low temperature and by γ phase over 900°C. Moreover, electron back scattered diffraction (EBSD) and X-ray diffraction (XRD) measurements on samples have revealed three classes of secondary phases: the first one has been identified as σ-Cr_2Re_3, the second one as Cr carbide-Cr_(23)C_6 and the third one as α-Cr. σ phase is very abundant at the lower temperatures while it disappears after long exposures at temperatures higher than 900°C. The σ phase composition is different at different temperatures and the Re content in particular increases with the temperature. Starting from the σ phase composition determined at different temperatures, a tool has been constructed that relates the service temperature to the Re content in the same phase. The new tool has been applied to the analyses of different components. The results of the new method have been compared to those ones obtained with the method based on γ' features, developed in the past through huge experimental campaigns. The agreement between the two methods is generally good, they can be used in a complementary way due to the fact that the γ' one seems to be more suitable for high temperature ranges (T＞900°C) where it gives a reliable estimation, while the sigma method is more suitable in the temperature range 700°C - 900°C.