A Study of Calcium-Magnesium-Alumino-Silicate (CMAS) Interactions With Gadolinium Zirconate (GZO), Gadolinum Aluminum Perovskite (GAP) and GZO-GAP Composited Thermal Barrier Coatings (TBCs).

Open Access
Author:
Karwa, Saagar
Area of Honors:
Materials Science and Engineering
Degree:
Bachelor of Science
Document Type:
Thesis
Thesis Supervisors:
  • Douglas E. Wolfe, Thesis Supervisor
  • Robert Allen Kimel, Honors Advisor
Keywords:
  • Gadolinium zirconate
  • Gadolinium aluminum perovskite
  • Gehlenite
  • Apatite
  • Gadolinium aluminum garnet
  • rare earth (RE) garnet
  • Calcium-magnesium-aluminum-silicate
  • Electron Beam - Physical Vapor Deposition
  • Air Plasma Spray
  • Thermal Barrier Coatings
  • Thermally grown oxide
  • Bond coat
  • Top coat
Abstract:
Through a pellet-based study, unique TBC compositions were evaluated in terms of their respective calcium magnesium alumino-silicate (CMAS) degradation mechanisms and the subsequently formed reaction products were identified and analyzed. The three TBC compositions used were gadolinium zirconate (GZO), gadolinium aluminum perovskite (GAP) and a 90GZO-10GAP composite. Pellets were synthesized using ground CMAS glass frit and the different TBC compositions loaded at a 25CMAS-75TBC and 50CMAS-50TBC mol. % ratio. They were heat treated (HT) at 1425 °C for 10 minutes, 1 hour and 4 hours to examine correlations between HT duration and relative proportion of reaction products formed. Pellets with pure GZO formed the infiltration resistant, hexagonal structured apatite silicate and minute amounts of spinel, as expected from literature. Higher CMAS loadings (50CMAS-50GZO) formed an additional cubic rare-earth (RE) garnet phase. CMAS-GAP pellets had a much higher affinity for CMAS reaction than the CMAS-GZO based pellets, forming over 2X as much of each corresponding apatite and garnet phase than the CMAS-GZO counterpart with the same loading condition. Additionally, the 50CMAS-50GAP compositions contained gehlenite, a crystalline sorosilicate phase which indicated that an underlying mechanism in GAP caused the crystallization of the residual CMAS melt in the samples. CMAS-(90GZO-10GAP) compositions formed the same reaction products as the GZO based pellets but the addition of 10 wt. % GAP helped increase the amount of precipitated apatite and garnet by 10 – 20 %, depending on the specific composition and HT duration. However, the addition of GAP limited grain coarsening in the 90GZO-10GAP samples when compared to the GZO specimens; for instance, the average grain size for the 50CMAS-50(90GZO-10GAP), 4 hr. HT sample was ~2-2.5 μm compared to ~5 μm grain size for the CMAS-GZO composition with an identical loading and HT duration. Nevertheless, the more rapid reactions with CMAS and higher proportion of precipitates formed indicates the potential for GAP as a material of interest for TBC development in the future.