PSI - Issue 68

Elena Fedorova et al. / Procedia Structural Integrity 68 (2025) 908–914 Elena Fedorova et al./ Structural Integrity Procedia 00 (2025) 000–000

909

2

the Ni-based bond coat (BC), the thermally-grown oxide (TGO) and the yttria-stabilized zirconia top coat (TC). One of major causes for limiting the lifetime of the TBC systems is related to nucleation, propagation and coalescence of cracks along the interfaces. Complex mechanisms controlling the durability and failure modes of TBC have been extensively studied and reviewed by Evans A. et al. (2001), Pindera (2005), Evans H.(2011), Wang (2020).

Nomenclature d m

damage parameter in mixed-mode debonding normal and tangential stress (MPa), respectively

s n , t t

s max , t max

maximum normal and tangential stress (MPa) contact elements can bear, respectively contact gap and tangential slip distance ( µ m) contact gap at s max and completion of debonding, respectively tangential slip distance at t max and completion of debonding, respectively

u n , u t

! " " # $# ! " " # $# ! " # $# ! " # $# !" !# $ %$

normal and tangential contact stiffness (MPa/m), respectively normal and tangential fracture energies (J/m 2 ), respectively normal and tangential critical fracture energies, respectively

Despite numerous theoretical and experimental investigations sought at developing reliable methods for the measurement, some issues related to a thorough understanding of mechanisms behind the interfacial adhesion parameters are open. For example, depending on methods for measuring, the values of strain energy release rate were found to vary from 0.3 to 230 J/m 2 . In previous study (Fedorova et al., 2022), factors that cause large spread of these values were divided into two groups: (1) intrinsic, which are related to conditions of TBCs layers formation influencing the value and distribution of residual strains and stresses as well as the location of failure initiation; (2) extrinsic, which are related to test conditions and approach to process the experimental results. The methodology of interfacial adhesion quantification proposed in this study is based on pullout test geometry, and originates from similar approaches applied in the study of Guo et al. (2005). New designs of samples and test fixtures are proposed to measure the adhesion parameters. Microstructural characterization is employed to study TBC layers thickness, its uniformity and geometry of the interfaces in as-received state as well as to control the occurrence and spread of spallation during testing. Numerical simulation of cooling process is an integral part of assessing the adhesive parameters of TBCs as it provides insights into the residual stress state and failure mechanisms that may affect experimental results. 2. Materials and microstructural analysis Two types of materials were used in this study. First, Ni-based single crystal superalloy samples (ZhS6U) with TBC deposited by EB-PVD supplied in the form of a rod to evaluate the ability of a proposed experimental technique. For this purpose, disk- and cylinder-shape specimens of 12 mm diameter and 2 or 10 mm thick were machined along the [0 0 1] direction from rods. Second, Ni-based single crystal superalloy turbine blades (ZhS32) with TBC deposited by EB-PVD supplied after high temperature cyclic oxidation tests in a purpose-built rig to investigate the influence of TGO thickness on residual stress state and interface cracking behaviour by finite element modelling. The thermal cycle consisted of a heating period of 30 ° C min −1 up to 1100 ° C, followed by a dwell and a cooling ramp with an initial rate of 30 ° C min −1 . The total duration of one cycle under flowing synthetic air was 23h. These specimens were cross-sectioned. Prior to testing or modeling, the microstructural analysis was performed. The specimens were prepared for scanning electron microscopy (SEM) observation using conventional techniques, which include grinding, polishing (up to 1200 grit SiC paper), and fine polishing (up to 1 μm diamond paste). Before mounting, the samples were coated with a thin epoxy resin layer to avoid damage during metallographic preparation. The cross sections were examined by SEM using a JEOL 6490LV microscope.