Degradation study of 7wt.%yttria stabilized zirconia (7YSZ) thermal barrier coatings on aero-engine combustion chamber parts due to infiltration by different CaO-MgO-Al2O3-SiO2 variants

Kurzfassung

A damage assessment of TBC (thermal barrier coatings) coated combustion chamber parts of aero-engines after service in CaO-MgO-Al2O3-SiO2 (CMAS) loaded areas was performed in this study. Parts from different aero-engines were investigated by scanning electron microscopy (SEM) and the extent of damage of the 7 wt.% yttria stabilised zirconia (7YSZ) plasma-sprayed TBC due to CMAS deposition and infiltration was estimated. With the help of SEM and energy-dispersive spectroscopy (EDS), CMAS chemical composition was analyzed and two different typical chemical compositions of CMAS were derived. These model CMAS compositions were synthesized in laboratory and their melting behaviour was determined by means of differential scanning calorimetry (DSC). Both CMAS variants were subsequently deposited on air plasma sprayed yttria stabilized zirconia (APS-7YSZ) TBC specimens. Samples were subjected to isothermal heat treatments in air at temperatures ranging from 1200°C to 1250°C for times between 10 h and 100 h. The phase formations and microstructural changes were examined by SEM and x-ray diffractometry (XRD). Results from the CMAS/TBC interaction experiments could duplicate with the damage patterns on the real combustion chamber parts. Furnace cycle tests (FCT) were conducted at 1135°C on TBC coated buttons with and without CMAS. The extent of TBC damage strongly correlates to the CMAS composition. The presence of Ca-sulfate (CaSO4, anhydrite) in the CMAS plays a large role in damaging the 7YSZ and infiltration depth of CMAS. With raising temperature, the depth of infiltration increases rapidly and the CMAS has penetrated completely throughout the TBC thickness at 1250°C. In addition, the time to failure of the TBCs strongly depended on the type of CMAS deposited. The life time of the samples with CaSO4 containing CMAS was found to be the lowest compared to the samples without CMAS and with the CaSO4 free CMAS variant.