Abstract

Precision establishment of the raw material influences in thermal spray processing of critical ceramic coating applications is essential for new technology engine performance and durability. This paper presents some of the results of a study examining the interrelationships of Zr02-Y203 powder in the ceramic coating plasma spray process and the performance of the deposit.

Examples of powders manufactured by several different processes, such as (i) spray drying, (ii) spray drying and sintering, (iii) sintering and crushing, (iv) casting and crushing, and (v) casting, crushing, and fusing, were initially characterized for particle size and shape, microstructure and morphology, surface area, density, flow characteristics, and qualitative phase distribution. The powder manufacturing process did have significant effects on all the above characteristics. For example, two extreme cases were: (i) spray dried powders - spherical and chemically inhomogeneous with a large surface area, low density and low mass flow rate, and contained primarily the monoclinic Zr02 phase with additional amounts of free Y203 in most cases; and (ii) fused, cast and crushed powders - dense irregular particles with a low surface area and high mass flow rate and were primarily tetragonal phase with small amounts of monoclinic phase. Both narrow and wide size distributions of particle sizes were found in the the range of the as-received powder samples, even though one size distribution was requested.

The powders were air plasma sprayed using identical spray parameters and fixturing on a Hastelloy-X substrate which had been Vacuum Plasma Sprayed (VPS) NiCrAlY bond coated. The coated samples were characterized for microstructure, erosion resistance, and quantitative phase distribution. The coatings ranged from dense to porous and this appeared to depend primarily upon the initial particle size. The coatings predominantly contained the tetragonal phase, even when the starting, powder material was monoclinic. However, the original monoclinic powders tended to retain a small fraction of the monoclinic phase in the as-coated condition. The coated samples were subjected to thermal cycling, and similar characterizations were carried put on the exposed samples. In general, as observed by other investigators, the more porous coatings.appeared to withstand a greater number of thermal cycles. In these sequences, coarser powders tended to produce longer-lived-coatings than powders with a relatively low mean powder size. Also, coatings exhibiting modest (> 5%) fractions of monoclinic phase appeared to withstand less number of thermal cycles than those containing only tetragonal or cubic phases. In addition, the thermal cycling exposure tended to induce a transformation of all the available phases to tetragonal .form.

Several of the powders were resized and again sprayed with the parameters and fixturing used previously. Additional specimen examples of powders were also introduced at this juncture. These represented modifications of the spray drying and fusion manufacturing techniques. The coatings produced were thermal cycled and erosion tested. Generally, as in the initial test sequence, the most homogeneous materials proved to be superior in thermal cycle performance. Improvements in the thermal cycle performance of certain spray dried examples were achieved after resizing, but the improvement achieved did not approach the levels observed in sintered or cast or fused powder structures.

The desire to improve the performance of aircraft turbine engines has led to higher operating temperatures in the turbine sections of the engines. To accommodate these higher temperatures new superalloys and material systems have been developed, but the practical temperature limits of these materials still limit engine performance. The use of thermal barrier coatings (TBCs) on turbine components has the potential to provide a dramatic increase in engine performance by allowing higher gas temperatures in the turbine and/or by reducing the need for cooling air in the turbine components. The desire to expand TBC use from combustors and other lower risk applications to more critical turbine applications has created the necessity for an improved, more reliable thermal barrier coating.

Earlier work on TBCs has focused on the use of stabilized of partially stabilized forms of zirconia as the insulating material. These materials can be plasma sprayed onto the engine part over a plasma sprayed oxidation resistant bond coat'11. The system currently being developed for employment on turbine hardware is zirconia partially stabilized with 8% yttria. This material provides coatings which have good thermal shock performance, hardness, and erosion characteris-

Past work on plasma sprayed TBCs has demonstrated wide variations in the performance of the coatings l5'61. Increased automation of the thermal spray process has demonstrated that significant differences exist in plasma spray powders/ and that these differences have a pronounced effect on coating performance. The recent work by Keller12' demonstrated that the initial powder size distribution and chemical homogeneity had a dramatic effect on TBC performance. Miller and others "•B' have also reported significant effects of composition and processing on the thermal cycle performance of ceramic coatings.' In addition, work on other thermal spray systems within General Electric has demonstrated that an understanding of the plasma spray raw material is critical to the development of reliable, reproducible coatings.

The purpose of this program is to analyze a number of the various zirconia-yttria powders currently available, to characterize those powders based upon their methods of manufacture, to gain insight into powder-plasma flame interactions, and, ultimately, to determine the effects of the raw material on the resulting thermal barrier coating.

This paper reports the results of this initial effort.

0 0

Post a comment