PSI - Issue 7

E. Vacchieri et al. / Procedia Structural Integrity 7 (2017) 182–189

183

2

E. Vacchieri et al. / Structural Integrity Procedia 00 (2017) 000–000

consider the more demanding service conditions in particular in term of coexistence of phenomena like creep, fatigue and oxidation. The use of service-like TMF tests is fundamental to assess the lifing methods as they represent as close as possible the operating conditions. The e ff ectiveness of new lifing methods is verified through the available field feedback collected in this new service environment. In fact, the blade and vane conditions after operation are fundamental to assess the lifing methodologies and, if a significant dataset is available, the features of operated parts can be statistically treated, not only to verify the lifing tools, but also to predict the admissible scrap rate with respect to the critical part repairing instruction. In the present paper, a creep-fatigue lifing methodology for GT blades and vanes is described and applied to a first stage blade of an F-class GT. The life maps obtained with the lifing methods, that correspond to crack initiation, have been compared to available field feedback for this component obtaining a good agreement in terms of critical area location and of predicted and experienced number of cycles. The field feedback for this blade has been treated using Weibull statistics. In particular, critical locations identified by the developed lifing method have been analysed considering di ff erent crack lengths to calculate the maximum allowable cycle number with respect to the blade re pair instruction. Moreover, a relationship between crack length and crack penetration has been found and a crack propagation law as a function of cycle number has been defined.

Nomenclature

shape parameter of PDFW

β

d c critical crack depth EBPVD electron beam physical vapour deposition EOH equivalent operating hours η characteristic life of PDFW FPI fluorescent penetrating inspection γ minimum life or location parameter of PDFW GT gas turbine k crack size l minimum detectable crack length L

life of a single blade or of the whole set for a fixed crack size

LCF

low cycle fatigue

n

number of blade in a set

NSU normal start-up PDFW probability density function Weibull PYSZ partially yttria stabilised zirconia S survivability or reliability level s 1 , s 2 , s 3 pricipal stress components σ VM Von Mises stress SX single crystal T Temperature TBC thermal barrier coating TMF thermo-mechanical fatigue

2. Materials and Methods

A first stage blade of an Ansaldo F-class GT has been considered in the present paper as a case study for the creep-fatigue lifing methodology and for the statistical analysis of the field feedback. The material of the blade is an SX Ni based superalloy, IN792SX. The blade is coated with metallic and ceramic coatings. The metallic coating is a NiCoCrAlYRe coating while the TBC is EBPVD PYSZ. The creep-fatigue lifing procedure has been defined on the basis of creep, LCF and cyclic hold time tests performed on the SX alloy samples solidified along the 001 crystal