ICMCTF2012 Session A2-1: Thermal and Environmental Barrier Coatings

Failure mechanisms in thermal barrier coatings (TBCs) often involve the propagation of delamination cracks through the ceramic layer. Mode I toughness measurements on air plasma-sprayed, dense, vertically cracked (DVC) 8YSZ TBCs using a double cantilever beam (DCB) test revealed R-curve behavior and steady state toughness values of Γ_Ic ~ 320±30 J/m². This unexpectedly high value has motivated an analysis of the test itself and the possible mechanisms responsible for the toughness elevation. Examination of local displacement data along the entire length of the cantilevers, including at the load point, reveals the location of the crack front. In analyzing the experiment, the compliant foundation of the cantilever (due to the presence of the TBC) and shear effects at the crack tip must be incorporated. Finite element analysis of the DCB specimen includes a layer with reduced stiffness between the beams that simulates the behavior of the compliant foundation. The model produces results that are consistent with those from experiment; thus it is possible to calculate an energy release rate solely with measured parameters, known material properties, and the displacement data. Additional experiments on a variety of air plasma-sprayed coatings show the evolution of the toughness and the possible contributions of multiple toughening mechanisms, including ferroelastic domain switching, crack bridging and pull-out.

Previous work has demonstrated that delamination progression of thermal barrier coatings (TBC) composed of yttria-stabilized zirconia (YSZ) can be monitored by photoluminescence piezospectroscopy (PLPS) and more recently by upconversion luminescence imaging of TBCs composed of YSZ incorporating a thin base layer co-doped with erbium and ytterbium (YSZ:Er,Yb). The recent development of imaging mode PLPS using a tunable filter now allows the comparison of both techniques by direct imaging of the same specimens. In this study, both PLPS imaging and upconversion luminescence imaging were performed to monitor the delamination progression of electron-beam physical vapor deposited (EB-PVD) TBCs at different stages of interrupted furnace cycling to 1163 ° C. In addition, the extent of mechanically induced delamination produced by Rockwell indentation at selected stages of TBC cyclic life was evaluated by both techniques. The TBC damage associated with the imaging results was verified by post-imaging SEM inspection of the specimen cross-sections. While each technique has its own strengths and weaknesses, it is shown that the information provided by both techniques is complementary and provides a better identification of the location/depth at which delamination cracks occur than by either technique alone. The complementary nature of these techniques can be attributed to their very different mechanisms for achieving the contrast between different stages of delamination. In particular, the delamination contrast for PLPS imaging relies on the reduction in stress in the thermally grown oxide (TGO), while upconversion luminescence imaging relies on the total internal reflection that occurs at cracks within or below the YSZ:Er,Yb layer. Therefore, PLPS imaging is more sensitive to damage to the TGO or to the TGO/bond coat interface, whereas upconversion luminescence imaging is more sensitive to damage at the TGO/TBC interface or above.

The influence of steep thermal gradients combined with rapid hot surface cooling on delamination of thermal barrier coatings is investigated. Transient thermal gradients induce stress gradients through the coating and substrate which, in turn, produce overall bending if the substrate is not very thick and if it is not constrained. Substrate thickness and constraint are important aspects of the mechanics of delamination due to transient thermal loading of coating-substrate systems. These aspects must be considered when laboratory tests are designed, and they must be considered for lifetime assessment under in-service conditions.

Zirconia-based ceramics have long been used to provide thermal protection to the structural components of modern gas turbine engines. Economic and environmental considerations have motivated higher engine operating temperatures, potentially leading to more rapid degradation of thermal barrier coating (TBC) systems dependent on a metastable phase, namely t’-8YSZ (ZrO₂+7-8wt%YO_1.5) TBCs. The t’-phase was originally thought to decay slowly into the equilibrium Y-lean tetragonal phase and Y-rich cubic phase, the former of which may undergo further transformation to the monoclinic phase. However, it has recently been shown that the t’-phase destabilizes at a small fraction of the time necessary to form the deleterious monoclinic phase. The rapid decay of t’-YSZ into a modulated microstructure of coherent domains offers additional insight on the importance of microstructural control in the phase evolution of YSZ TBCs. Traditional phase stability characterization techniques have been reevaluated in order to provide a more complete description of this process. In particular, x-ray diffraction (XRD) techniques, both at room temperature and elevated temperatures, have been used to quantify the changing phase fractions and composition of each phase, with additional implications for the equilibrium phase diagram. XRD is more powerful when used in conjunction with microstructural observations. As such, three different starting morphologies will be compared on the basis of phase stability. Stemming from this comparison, potential pathways for further delaying the onset of the monoclinic transformation will be explored. While the effectiveness of these measures is expected to be modest, lessons can be learned from the t’-8YSZ system. In particular, maintaining a high degree of tetragonality over the entire range of relevant temperatures may be key to supporting or improving the in service toughness and durability. A novel TBC system will be introduced in which a relatively large single phase tetragonal field with exceptional tetragonality has been stabilized and has demonstrated comparable or improved toughness, making it a promising alternative for next generation TBCs.

This work was partially supported with funding from the US DoE via Cooperative Agreement DE-FC26-05NT42643 and the NSF via FRG-GOALI Contract NSF/DMR0605700. Any opinions, findings, conclusions or other recommendations expressed are those of the authors and do not necessarily reflect the views of the US Department of Energy or the National Science Foundation.

Coatings designed for high temperature protection, Aluminium rich intermetallic coating as well as thermal barrier coatings (TBC), are prone to damage when exposed to stages of high temperature and cooling. Coupling thermal and mechanical loading tests provide an accurate simulation of in-service loadings and the generation of subsequent realistic damage. Thus, the aim of this study is to clarify the way the mechanical behaviour of the under-layer interacts with surface damage. This point will be examined for room temperature mechanical tests performed up to oxide or TBC spallation. The specimens were initially subjected to high temperature thermal loading for both isothermal and thermal cycling.

When single crystals coated with TBC are subjected to mechanical compression, the strain localisation arising in the single crystal has been already been shown to drive the ceramic coating to spallation. In the same manner, when the oxide is growing on a free surface, the oxide spallation due to mechanical compression is dependent of the local behaviour of the coating. Indeed, for a typical CVD-NiAl coating, strain localisation is related to the coating microstructure. Moreover, the oxide morphology will also contribute to localisation of strain and hence oxide spallation. The chosen experimental methodology will be explained since it offers a complement to thermo-gravimetric analysis. It particularly includes the intensive use of full-field analysis by surface strain field measurement. This technique enables a quantitative characterisation of surface damage and can be used to define intrinsic rupture material parameter. Complementary finite element analysis of both test configuration and principal microstructural features are performed. It allows a close view of the mechanical state leading to rupture to be obtained. Finally, assessment of thermo-mechanical coupling will be discussed for complex loading paths.

Thermal barrier coatings (TBCs) are essential for the effective operation of turbine blades in high temperature gas environments. The drive for next generation TBCs, however, poses new demands that the current TBC, 7.6 mol% YO_1.5 stabilized zirconia (7YSZ) is unlikely to satisfy. At issue is the phase stability of the coating and its resistance to sintering, both of which are explored in the YO_1.5 - TaO_2.5 - ZrO₂ (Y-Ta-Zr) system. This research addresses the issue of high temperature densification and its mitigation using multi-phase compositions within this ternary system. In addition to the baseline 7YSZ, four compositions were selected for investigation representing four different phase constitutions: 16Y-16Ta-Zr (stable tetragonal, t); 20Y-20Ta-Zr (two-phase tetragonal zirconia solid solution, t, and monoclinic yttrium tantalate, m-YTaO₄); 22Y-13Ta-Zr (two-phase non transformable tetragonal, t, and fluorite, c); and finally 18Y-28Ta-Zr (three-phase mixture of t, m-YTaO₄, and orthorhombic Zr₃TaO₈ phases). Densification studies were performed at 1250 °C with dwell times of 1, 4, 9, and 300 h. Pore size distributions have been quantified at each sintering time using BET analysis. Remarkably, the 18Y-28Ta-Zr composition increased only to 55% of its theoretical density from a green body density of 49% after 300 h. Pore size analysis indicates that the pores are relatively stable from 1 to 300 h at 1250°C. This compares with 22Y-13Ta-Zr and 7.6YSZ, which reached 70% and 85% of their theoretical densities, respectively, after the same exposure. The 22Y-13Ta-Zr, along with 18Y-28Ta-Zr, which are also phase stable and offer lower thermal conductivity than 7YSZ, suggests alternate regions within the Y-Ta-Zr system offering promise for future development.