Experimental and computational study on hot fatigue process of thermal barrier coatings by EB-PVD

Huibin Xu,Shengkai Gong,Yue Zhang,Chunxia Zhang
DOI: https://doi.org/10.1016/j.intermet.2004.07.031
IF: 4.075
2005-01-01
Intermetallics
Abstract:NiCoCrAlY+(6–8wt%)Y2O3 stabilized ZrO2(YSZ) two-layer thermal barrier coatings (TBCs) were deposited by electron beam physical vapor deposition on Ni3Al or Ni-base superalloy and their high temperature fatigue behavior was investigated experimentally and computationally. It has been found that the pre-treatment of the bond coat has greatly improved the thermal cyclic lifetime of TBCs with respect to those without pre-treatment. The lifetime is limited by the existence of interface between bond coat and top coat, and smoothing the interface can also improve the thermal cyclic lifetime. When the roughness of the interface changed from 3.76 to 0.82μm, testing lifetime increased from about 500 (1000 cycles) to 700h (1400 cycles), and the adhesive strength of the interface increased from 3.5 to 4.7MPam1/2 estimated from the result of surface Rock-well hardness testing. Cracks occurred in the thermally grown oxide (TGO) layer are considered to be the main reason for the failure of two-layer TBCs. It has been found that with pre-oxide layer thickness increasing from 1.0 to 3.0μm, the growth rate of TGO increased during thermal cyclic testing and the thermal cyclic lifetime of TBCs decreased from 730 to 450h. The lower lifetime was caused by the co-action of larger internal stress resulted from thicker TGO layer and the fatigue of the TGO layer. Finite element method (FEM) analysis showed that large tensile stress was introduced at the beginning of rapid cooling process, resulting in the spallation of top coat from bond coat. Ab initio calculation was carried out to evaluate the binding energy of the interface of bond coat and ceramic top coat, and FEM models were used to investigate stress and strain states in TBCs during thermal shock process as a function of cooling rate. The results showed that ZrO2/Ni interface exhibits nonmetallic properties, and that the chemical bonding in ZrO2/Ni is covalent–ionic in nature. The calculation of binding energies between both Ni and O and Zr and O indicates that the binding energy of Ni and O is larger than that of Zr and O, which is helpful to investigate the failure mechanism of thermal barrier coatings.
What problem does this paper attempt to address?