Oxidation‐Induced Stresses in Thermal Barrier Coating Systems

Thermal Conductivity Analysis and Lifetime Testing of Suspension Plasma-Sprayed Thermal Barrier Coatings

“…Similar results for APS versus HVOF bond coats have been reported previously with conventional APS top coats [6]. For all of the coating combinations tested, failure resulted from transverse crack propagation that could be connected to stress produced by a thermally-grown oxide (TGO) layer at the bond coat-top coat interface [28,29]. This layer is an unavoidable by-product during the high temperature exposure of YSZ-MCrAlY TBCs in atmospheric conditions and results from oxygen diffusing to and reacting with the bond coat material.…”

“…As the TGO develops, oxygen must then diffuse through this layer to react with the bond coat; thus, the TGO composition can affect bond coat oxidation. For example, a reaction between the diffusing oxygen and the aluminum of the bond coat can form an alumina layer that produces a slower growing oxide layer than reactions yielding mixed oxides consisting of nickel, chromia spinels, which are considered highly detrimental to coating lifetime [28]. Hence, the bond coat chemistry also influences TGO development by determining which oxides form.…”

“…Hence, the bond coat chemistry also influences TGO development by determining which oxides form. For instance, a β Ni-Al phase in the bond coat acts as an aluminum reservoir that promotes the formation of a passivating alumina layer, so that the beta phase content can be tracked and used to determine the degree of bond coat oxidation and the remaining chemical life of the bond coat [28]. A slower growing, passivating oxide is preferred, because the formation of any TGO introduces stress into the TBC system.…”

“…There is, however, a large amount of lighter colored mixed oxide regions above the alumina layer (labeled C). The mixed oxide consists of nickel, chromia spinels, and due to its faster growth, it is considered highly detrimental to coating lifetime [28]. Investigation of the HVOF bond coat after 100-140 cycles reveals that a substantial amount of the beta phase (β Ni-Al) remained within the bond coat.…”

“…Unlike the TGO layer in the HVOF coating, the HVAF TGO layer is denser in structure and contains very little of the mixed oxides found within the HVOF or plasma bond coats. In some locations, rumpling of the oxide layer has been observed (labeled B); this may contribute to crack initiation and failure [28]. The improved oxidation performance in TCF testing explains the longer cyclic lifetime.…”