Abstract

Air plasma sprayed (APS) flash coatings on high velocity oxygen fuel (HVOF) bond coatings are well known to extend the lifetime of thermal barrier coatings (TBCs). Recent work compared flash coatings of NiCoCrAlY and NiCoCrAlYHfSi applied to both rods and disk substrates of alloy 247. For rod specimens, 100 h cycles were used at 1100 °C in wet air. Both flash coatings significantly improved the lifetime compared to HVOF-only and vacuum plasma spray (VPS)-only MCrAlY bond coatings with no statistical difference between the two flash coatings. For disk specimens tested in 1 h cycles at 1100 °C in wet air, the NiCoCrAlY flash coating significantly outperformed an HVOF-only NiCoCrAlYHfSi bond coating and a NiCoCrAlYHfSi flash coating. The flash coatings formed a mixed oxide-metal zone that appeared to inhibit crack formation and therefore extend lifetime. In addition to the flash coating increasing the bond coating roughness, the underlying HVOF layer acted as a source of Al for this intermixed zone and prevented the oxide from penetrating deeper into the bond coating. The lower Y+Hf content in the Y-only flash coating appeared to minimize oxidation in the flash layer, thereby increasing the benefit compared to a NiCoCrAlYHfSi flash coating.

Introduction

Fossil fuels continue to represent the majority of electricity generation in the U.S. (>60%) but natural gas has surpassed coal in recent years [1]. It appears that natural gas-fired combined cycle plants will dominate the power generation market for the foreseeable future. Increased turbine efficiency can help lower both emissions and the cost of electricity. The pathway to higher efficiency is primarily through increasing the turbine inlet temperature. However, achieving higher temperatures must be accomplished while maintaining the reliability and durability (>25 kh between major overhauls) required by the utility industry. Thus, improved materials solutions are essential to improving efficiency. Considerable effort is focused on ceramic composite hot section components, particularly for aircraft engines [2]. However, Si-based composites must overcome the environmental durability issues caused by water vapor in combustion environments [3]. Also, considerable development is needed to address cost concerns and the larger components needed for land-based turbines. For the near term, turbines will continue to rely on Ni-base superalloys, which are able to reliably operate at very high turbine inlet temperatures (≥1400 °C) thanks to film cooling and thermal barrier coatings (TBCs) [413].

New TBC concepts are typically first evaluated in furnace cycle testing (FCT) with thermal gradient testing, such as burner rigs, being used for promising TBC concepts before engine testing. Over time, the TBC research effort at Oak Ridge National Laboratory has developed several core focus areas and concepts: (1) large industrial gas turbines use thermally sprayed coatings [6,9,10], (2) lower cost, polycrystalline superalloys are of interest and have similar TBC lifetimes as single crystal superalloy substrates [14], (3) FCT testing should be conducted in air with water vapor to better simulate turbine exhaust and the negative effects of H2O on TBC performance [1324], and (4) in addition to 1 h FCT, 100 h cycles can simulate base-load performance, which might be relevant for integrated gasification combined cycle operation. Recently, the project began incorporating multilayer top coating architectures [25] and rod specimens [26] as steps toward coating bars for thermal gradient testing. The rod specimens had surprisingly shorter FCT lifetimes compared to those of flat disk specimens [24,27,28]. Based on input from industry, multilayer bond coatings with the incorporation of an Air Plasma Spray (APS) NiCoCrAl “flash” layer [26] above an high velocity oxygen fuel (HVOF) coating layer were evaluated. Preliminary results with multilayer bond coatings were presented previously [28]. The hypothesis was that flash coatings were effective because of their higher roughness [26]. If that is correct, then perhaps the use of MCrAlYHfSi bond coatings [5,29,30] is not necessary since their increased FCT lifetime performance, compared to Y-only bond coatings [31] was attributed to the co-doping effect of Y and Hf [32,33] which improves alumina scale adhesion. To address these observations, this study compares Y-only to MCrAlYHfSi flash coatings on both rod and disk specimens.

Experimental Procedure

Substrates for coating included disk (2 mm thick, ∼16 mm diameter with chamfered edges) and rod specimens (12.5 mm diameter, ∼30 mm length) machined from superalloy 247 with composition shown in Table 1. Prior to coating deposition, the specimen surfaces were grit blasted with alumina and then coated using standard commercial-type, APS, HVOF or vacuum plasma spray (VPS) processes. For HVOF and APS, the powder size was 16–88 μm. The two powder compositions are listed in Table 1. The coated substrates were then annealed in a vacuum of 10−4 Pa (10−6 Torr) for 4 h at 1080 °C. After annealing, the disks were coated with ∼200 μm of standard Y2O3-stabilized ZrO2 (YSZ) deposited by APS (17.1±0.5% porosity). The APS YSZ top coating on the rods was nominally 300 μm thick (1) extra porous YSZ (20.4±1.4% porosity), (2) YSZ with a dense inner layer and extra porous (hi P) outer layer, or (3) YSZ with an outer layer of Gd2Zr2O7 (YSZ-GDZ) top coating. Thus, the ceramic top coating was either a single- or multilayer structure, as described elsewhere [25]. The various coating combinations evaluated are summarized in Table 2. The surface roughness (Ra) of the bond coatings was measured using a Keyence model VR-3000 optical profilometer (Itasca, IL) or by analysis of light microscopy images [27].

Table 1

Alloy chemical composition (mass %, balance Ni) determined by plasma and combustion analyses

AlloyCrAlCoTiHfYSaOther
2478.55.79.81.01.4<<39.9W,3.1Ta,0.7Mo,0.2C
Bond coating powders:
YHfSi16.712.321.6<0.250.6820.65Si
Y only16.612.823.0<<0.4280.04Si
AlloyCrAlCoTiHfYSaOther
2478.55.79.81.01.4<<39.9W,3.1Ta,0.7Mo,0.2C
Bond coating powders:
YHfSi16.712.321.6<0.250.6820.65Si
Y only16.612.823.0<<0.4280.04Si
a

S in ppmw < indicates less than 0.01% or 0.0003% for Y.

Table 2

Summary of different coating combinations investigated typically using commercial NiCoCrAlYHfSi (YHS) and YSZ powder except where noted

SubstrateBond CoatingAPS YSZ Top coating
Rod-HVOF (YHS)-High porosity (1 layer)
-HVOF (YHS) + 200 μm APS (YHS)-Bi-layer (inner dense)
-Tri-layer (outer Gd2Zr2O7)
Rod-HVOF (YHS) + 50 μm APS (YHS)-All high porosity (1 layer)
-HVOF (YHS) + 50 μm APS (Y only)
-VPS (YHS)
Disk-HVOF (YHS)-All standard porosity (1 layer)
-HVOF (YHS) + 50 μm APS (YHS)
-HVOF (YHS) + 50 μm APS (Y only)
SubstrateBond CoatingAPS YSZ Top coating
Rod-HVOF (YHS)-High porosity (1 layer)
-HVOF (YHS) + 200 μm APS (YHS)-Bi-layer (inner dense)
-Tri-layer (outer Gd2Zr2O7)
Rod-HVOF (YHS) + 50 μm APS (YHS)-All high porosity (1 layer)
-HVOF (YHS) + 50 μm APS (Y only)
-VPS (YHS)
Disk-HVOF (YHS)-All standard porosity (1 layer)
-HVOF (YHS) + 50 μm APS (YHS)
-HVOF (YHS) + 50 μm APS (Y only)

Average FCT lifetimes of TBC specimens were determined by exposing groups of 3–5 similarly coated rod or disk specimens at each condition. Coating failure was defined as >20% loss of the YSZ layer but, in most cases, the entire top coating spalled in one piece at failure. For 1 h cycle time FCT, the specimens were hung by a Pt-Rh wire in an automated cyclic rig and exposed for 1 h at temperature (1100 °C) in air with 10±1 vol. % H2O followed by cooling for 10 min to <30 °C in laboratory air for each cycle. For 100 h cycle time FCT, the specimens were placed in an alumina boat, heated to temperature over ∼4 h in an alumina tube with flowing argon, held for 100 h in air with 10±1 vol. % H2O, and then furnace cooled after the cycle in argon. To add water vapor, the carrier gas was typically flowed at 500 ml/min with distilled water atomized into the gas stream above its condensation temperature. The injected water was measured to calibrate its concentration. For weighing, specimens were cooled to room temperature and visually examined. Specimen mass change was measured every 20 cycles or each 100 h cycle using a Mettler-Toledo model XP205 balance (Columbus, OH). However, only time to failure is reported for this study.

A Dilor XY800 Raman microprobe (Horiba Scientific, Edison, NJ) with an Innova 308c Ar+ laser (Coherent, Inc., Santa Clara, CA) operating at 5145 Å with a power at the YSZ surface of 10 mW was used to measure residual stresses in the thermally grown alumina scale. Photo-stimulated luminescence piezospectroscopy stress maps were collected on the flat specimens and a line profile was collected along the length of rod specimens. The acquisition time for one spectrum was 0.1–0.5 s and 3468 and 2001 spectra were collected from the disk and rod specimens, respectively. The location of the map or line was identified by a fiducial marker and the procedure is described elsewhere [14,22,24,27,34]. The laser spot size and the mapping step size were both 10 μm. The intensity of the spectrum measured underneath the YSZ was ∼10,000 times weaker compared to a bare bond coating. The stress state was not assumed to be biaxial since the bond coating surface was not flat and the scale cracked, so hydrostatic stress (the average of the trace of the stress tensor) was determined by dividing the frequency shift of the R2 line by 7.61 cm−1/GPa, which corresponds to a zero stress reference (NIST standard reference material 676A) [27]. Due to the large number of spectra collected, no attempt was made to deconvolute multiple stress states and only a single stress (one set of R-lines) was fitted to each spectrum.

For cross-sectional analyses, the specimens were mounted in epoxy and then cut for metallographical polishing and examination by light microscopy. Additional characterization of the specimens is presented elsewhere [35,36].

Results

Rod Specimens.

Figure 1 shows the final FCT results from the previous study where the rods with an HVOF-only MCrAlYHfSi bond coating showed very low lifetimes in 100 h cycles at 1100 °C [28]. The addition of an APS MCrAlYHfSi layer over the HVOF layer greatly increased the FCT lifetime under these conditions. The different top coatings did not appear to affect the behavior. Figure 2 shows examples of failed specimens. Figure 2(a) shows one of the bi-layer top coatings after spallation of the top coating after 26, 100 h cycles. A thick alumina layer formed at the bond coating–top coating interface. This layer is shown at higher magnification in Fig. 2(b) for a high porosity YSZ specimen cycled for 36, 100 h cycles before failure. Compared to the as-deposited coating [28], the internal oxidation (dark areas) in the APS layer, particularly near the HVOF interface, has significantly increased with exposure.

Fig. 1
Average coating lifetimes (cumulative time in 100 h cycles to failure at 1100 °C in air + 10%H2O) for rod specimens with two bond coatings and three different top coatings. Whiskers note a standard deviation for three specimens of each type.
Fig. 1
Average coating lifetimes (cumulative time in 100 h cycles to failure at 1100 °C in air + 10%H2O) for rod specimens with two bond coatings and three different top coatings. Whiskers note a standard deviation for three specimens of each type.
Close modal
Fig. 2
Light microscopy of polished cross sections of failed APS+HVOF bond coatings on 247 rods after 100 h cycles at 1100 °C in wet air (a) bi-layer YSZ, 26 × 100 h and (b) high porosity YSZ, 36 × 100 h cycles
Fig. 2
Light microscopy of polished cross sections of failed APS+HVOF bond coatings on 247 rods after 100 h cycles at 1100 °C in wet air (a) bi-layer YSZ, 26 × 100 h and (b) high porosity YSZ, 36 × 100 h cycles
Close modal

Figure 3 compares the residual stress in an HVOF-only disk and APS-coated rod specimens exposed in 100 h cycles. While maps can be produced for the flat specimens and measured at the same location over time [14,22,24,27,34], this is not possible for the rods and only an average compressive stress value is obtained for the rod specimen line profiles. The measurement was further complicated by the 300 μm top coating thickness on the rod specimens. In some cases, the measurements were stopped because no signal could be obtained through the YSZ at longer exposure times. For comparison, a similar value was averaged for the disk specimen. Interestingly, the residual stresses were consistently higher on the disk specimen, perhaps because of the flatter interface compared to the APS-coated rod specimens. Recall from Fig. 1 that the lifetimes were very short for the HVOF-only rod specimens so no similar long-term comparison could be made between disks and rods with the same bond coating. Perhaps the lower residual stress for rod specimens is an indication of more damage to the alumina scale because of the interface geometry.

Fig. 3
Residual compressive stress in the alumina scale measured by photo-stimulated luminescence piezospectroscopy on 247 disk and rod specimens as a function of exposure time in 100 h cycles at 1100 °C. Additional measurements were made on a second high porosity specimen after the first specimen failed.
Fig. 3
Residual compressive stress in the alumina scale measured by photo-stimulated luminescence piezospectroscopy on 247 disk and rod specimens as a function of exposure time in 100 h cycles at 1100 °C. Additional measurements were made on a second high porosity specimen after the first specimen failed.
Close modal

While the results in Fig. 1 showed a large increase in FCT lifetimes when the APS layer was added, the thick APS layer was not like the ∼50 μm APS “flash” coating described previously [26] and there was concern that the lifetime increase was simply due to the thicker bond coating. Therefore, another batch of specimens was made with a thinner APS layer. Also, the hypothesis was that the flash coating was effective because of its high roughness [26]. That led to two questions: (1) VPS coatings can have high roughness, how do they compare on rod specimens? and (2) if roughness is key, are MCrAlYHfSi coating compositions needed? Fig. 4 shows the results of the next phase of FCT testing of rods in 100 h cycles at 1100 °C. For comparison, the average lifetime for HVOF-only disks is also shown. Figure 5 shows the as-deposited VPS coating on a 247 rod with an interface roughness of ∼10 μm (Ra) [36]. However, the lifetime did not match the excellent lifetime observed for the 50 μm thick APS flash coatings, Fig. 4. For the two types of flash coatings, with and without Hf+Si additions, the large variation in lifetimes (large standard deviations) makes their behavior statistically similar. One Y-only flash coated specimen failed after 54 cycles. Figure 6 shows the ∼50 μm flash-coated bond coatings after similar failure times. Very thick alumina formed within the outer bond coating layer in both cases. This mixed metal-alumina layer may have inhibited crack propagation at the interface resulting in a long average FCT lifetime. In both cases, the underlying HVOF layer prevented the internal oxidation from penetrating further and affecting the underlying alloy 247 substrate.

Fig. 4
Average lifetime for the next phase of alloy 247 rod specimens with high porosity YSZ top coatings combinations on alloy 247 substrates exposed in 100 h cycles at 1100 °C in air + 10%H2O. Results for disk specimens are shown for comparison. Whiskers note a standard deviation for 3–5 specimens of each type.
Fig. 4
Average lifetime for the next phase of alloy 247 rod specimens with high porosity YSZ top coatings combinations on alloy 247 substrates exposed in 100 h cycles at 1100 °C in air + 10%H2O. Results for disk specimens are shown for comparison. Whiskers note a standard deviation for 3–5 specimens of each type.
Close modal
Fig. 5
Light microscopy of polished cross section of as-coated VPS MCrAlYHfSi bond coating on 247 rod with high porosity YSZ
Fig. 5
Light microscopy of polished cross section of as-coated VPS MCrAlYHfSi bond coating on 247 rod with high porosity YSZ
Close modal
Fig. 6
Light microscopy of polished cross sections of failed APS flash bond coatings on 247 rods after 100 h cycles at 1100 °C in wet air (a) Y-only APS bond coating after 29, 100 h cycles and (b) MCrAlYHfSi APS bond coating after 28, 100 h cycles
Fig. 6
Light microscopy of polished cross sections of failed APS flash bond coatings on 247 rods after 100 h cycles at 1100 °C in wet air (a) Y-only APS bond coating after 29, 100 h cycles and (b) MCrAlYHfSi APS bond coating after 28, 100 h cycles
Close modal

Disk Specimens.

The effectiveness of the APS flash coatings for the rod specimens raised a question about how a flash coating would perform on disk specimens. Thus, a similar experiment was conducted using 1 h cycles [35]. Figure 7 shows the starting flash coatings. It appeared that more oxide formed during deposition of the APS MCrAlYHfSi layer compared to the MCrAlY layer. In addition, a batch of HVOF-only disk specimens was prepared and all of the three types of bond coatings received the same standard porosity YSZ layer, Table 2. Figure 8 shows the measured roughness values for the three bond coatings. In addition to the Ra values, the fractal dimension of the roughness profile was calculated using a method previously reported [26]. The flash coatings did slightly increase the roughness compared to the HVOF-only roughness values.

Fig. 7
Light microscopy of polished cross sections of as-coated 247 disks with an inner HVOF MCrAlYHfSi layer and an outer APS flash coating with (a) Y only and (b) YHfSi
Fig. 7
Light microscopy of polished cross sections of as-coated 247 disks with an inner HVOF MCrAlYHfSi layer and an outer APS flash coating with (a) Y only and (b) YHfSi
Close modal
Fig. 8
Measured coating roughness on disk specimens
Fig. 8
Measured coating roughness on disk specimens
Close modal

The FCT average lifetime results are shown in Fig. 9. Each batch contained five specimens of each coating type. The HVOF-only results were compared to a second batch of three coatings reported previously and exposed in 2015 using the same conditions [37]. Unlike previous 1 h FCT experiments where the results were not statistically significant [24,34], these results showed significant improvements for the flash coatings compared to the HVOF-only lifetimes and the Y-only flash coatings showed a >70% increase in the average FCT lifetime.

Fig. 9
Comparison of average coating lifetimes (cumulative time in 1 h cycles to failure) for alloy 247 disk specimens with various bond coatings tested at 1100 °C in wet air [35]
Fig. 9
Comparison of average coating lifetimes (cumulative time in 1 h cycles to failure) for alloy 247 disk specimens with various bond coatings tested at 1100 °C in wet air [35]
Close modal

Figure 10 shows representative cross section of failed specimens of each type of bond coating. The failed HVOF-only coating appears similar to prior studies [22,31] with a fairly uniform alumina scale formed at the bond coating-top coating interface and minimal internal oxidation, Fig. 10(a). The flash coated specimens show a highly oxidized morphology with the APS layer significantly consumed and thick oxide formed. Despite only being exposed for 680 cycles, the APS MCrAlYHfSi layer appeared to be more heavily oxidized than the Y-only layer, Figs. 10(b) and 10(c), respectively. In both cases, the inner HVOF layer appeared intact with no attack of the 247 substrate.

Fig. 10
Light microscopy of polished cross sections of coated 247 disks exposed in 1 h cycles at 1100 °C in wet air (a) HVOF MCrAlYHfSi after 620 cycles, (b) flash YHfSi bond coating after 680 cycles, and (c) flash Y only after 1060 cycles
Fig. 10
Light microscopy of polished cross sections of coated 247 disks exposed in 1 h cycles at 1100 °C in wet air (a) HVOF MCrAlYHfSi after 620 cycles, (b) flash YHfSi bond coating after 680 cycles, and (c) flash Y only after 1060 cycles
Close modal

Because of the difference in exposure time in Fig. 10, a more uniform comparison was obtained by taking a sixth specimen of each flash coated specimen and exposing it for various cycle intervals and removing a piece for characterization at 0 cycles (Fig. 7) and 100 cycles (Fig. 11). Results after 300 cycles are reported elsewhere [35]. After only 100 cycles, there appears to be a clear difference between the two flash coatings with significantly more oxide formed with the YHfSi coating, which contained significantly more Y+Hf than the Y-only coating, Table 1. It appears that for these thin APS layers, the Y-only powder produced a better FCT lifetime because oxidation was minimized in this outer layer.

Fig. 11
Light microscopy of polished cross sections of flash coated 247 disks after 100, 1 h cycles at 1100 °C in air + 10%H2O (a) Y-only flash coating and (b) YHfSi flash coating
Fig. 11
Light microscopy of polished cross sections of flash coated 247 disks after 100, 1 h cycles at 1100 °C in air + 10%H2O (a) Y-only flash coating and (b) YHfSi flash coating
Close modal

Discussion

The FCT results for both disks and rods confirm the benefit of an APS flash coating previously reported [26]. Increased bond coating roughness could explain the benefit. However, for the rod specimens, VPS bond coatings also have a high roughness but only provided a 3× increase in average FCT lifetime (12 cycles) compared to HVOF-only bond coatings (4 cycles), Figs. 1 and 4. In contrast, the flash coatings increased the lifetime by 7–8×. For the disk specimens, the increase in roughness, measured by Ra or fractal dimension (Fig. 8), did not appear to predict the large increase in lifetime for the Y-only flash coating, Fig. 9. Certainly, high roughness is needed for thermally sprayed coatings, but increased roughness may not capture the entire flash coating benefit.

Considering the failed sections in Fig. 10, the flash coating benefit also might be related to the mixed metal-alumina layer formed at the interface between the HVOF layer and the YSZ top coating. This layer may be extremely difficult for an interfacial crack to propagate through. In the case of an APS-only bond coating, the internal oxidation would propagate to include the entire APS bond coating. However, for this two layer structure, the HVOF layer prevents the oxidation reaction front from reaching the substrate. (A recent experiment did not show a superior performance for a flash coating compared to an APS-only bond coating [38] and that experiment is being repeated.) The thicker APS layer in Fig. 2(b) illustrates how internal oxidation permeates the entire APS layer from top to bottom. One problem with forming a mixed metal-oxide layer in an oxidizing environment is that the increased metal-oxide surface area, Figs. 10(b) and 10(c), increases the rate of Al consumption. Another benefit of the underlying HVOF layer is that it can act as an Al reservoir for the metal in the mixed layer, supplying fast-diffusing Al to the reaction front to maintain the mixed structure, thereby slowing Al depletion and delaying the formation of voluminous, faster growing Ni-rich oxides [35].

An analysis of Figs. 7, 10, and 11 suggests that using MCrAlYHfSi powder to make the APS layer results in much more oxide formation and thus a faster Al consumption rate, which leads to a shorter lifetime, Fig. 9. The Y-only flash layer appeared to have coarser metal particles remaining in Fig. 10(c), however, both coatings were deposited with the same powder size. The increased oxidation with Y and Hf resulted in more oxide formation in the as-deposited coating and after only 100 cycles, Figs. 7 and 11. The smaller Y content in the MCrAlY flash coating appears to be an advantage and suggests the powder composition of the APS flash layer might be further optimized. A similar large increase in average FCT lifetime was not observed for the Y-only flash coating on rod specimens tested in 100 h cycles. This may have been due to the small (3) number of specimens tested in this experiment. The average was higher for the Y-only specimens, but it was not statistically significant as suggested by the high standard deviations.

The majority of FCT testing is conducted on flat specimens. One reason for using rod specimens was to eliminate edge failures. Also, turbine components are not flat. The change in geometry to rod specimens resulted in unexpected complications including the 5–6× decrease in FCT lifetime with HVOF-only bond coatings (compare rods in Fig. 1 with disks in Fig. 4). It is not clear why this decrease in lifetime occurred, but the rapid propagation of the interfacial crack at failure suggests that a round geometry magnifies the importance of roughness in preventing crack propagation. However, switching to the rougher VPS bond coating did not fully recover the 5–6× decrease in lifetime, Fig. 4. The mixed metal-oxide layer created with the flash coating appears ideal for rod specimens. The residual stress data in Fig. 3 suggests that more alumina scale cracking occurs on rod specimens as cracking relieves the residual stress. However, this may also be a product of the higher roughness with the flash coating and the highly nonplanar alumina reaction product formed, Fig. 2. Residual stress is highest for a flat alumina scale where the compressive cooling stress remains in the plane parallel to the interface. For example, 3 GPa was measure for an aluminide bond coating [34], 2–3× higher than in Fig. 3. However, a more sophisticated analysis of the failure mechanism on these various specimens may be needed to fully understand these results and the benefit of flash coatings.

The improvement in rod FCT lifetime is an important step in testing multilayer top coatings. Without a thermal gradient, there is no benefit of switching to a lower thermal conductivity GDZ top coating. This work has identified the need to optimize coating deposition parameters based on substrate geometry. Also, both the flash bond coating and the higher porosity YSZ top coating are parameters that need to be investigated in burner rig testing.

Conclusion

The addition of an APS MCrAlX flash coating on a dense HVOF bond coating was shown to increase average lifetime in FCT of both rod and flat disk specimens. The flash coatings were fabricated using similar NiCoCrAlY and NiCoCrAlYHfSi powders. For rod specimens, both coatings were more effective than HVOF-only or VPS-only bond coatings using 100 h cycles at 1100 °C in wet air. For disk specimens tested using 1 h cycles, the ∼50 μm Y-only flash coating increased the FCT lifetime by >70%, while the MCrAlYHfSi flash coating resulted in only a 12% increase compared to an HVOF-only bond coating. The flash coatings formed a mixed oxide-metal zone between the HVOF inner layer and the YSZ top coating that appeared to inhibit crack formation and extend lifetime. In addition to the flash coating increasing the bond coating roughness, the underlying HVOF layer acted as a source of Al for this intermixed zone and prevented the oxide from penetrating deeper into the bond coating. The lower Y+Hf level in the Y-only flash coating appeared to minimize oxidation in the flash layer, thereby increasing the benefit compared to a NiCoCrAlYHfSi flash coating.

Footnotes

Acknowledgment

The authors would like to thank G. W. Garner, T. M. Lowe, M. Stephens, and T. Jordan at ORNL for assistance with the experimental work and A. Hague of Mitsubishi Hitachi Power Systems for the VPS coatings. S. Dryepondt and E. Lara-Curzio provided helpful comments on the paper. This research was sponsored by the U.S. Department of Energy, Office of Fossil Energy, Turbine Program (R. Dennis program manager).

This manuscript has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paid-up, irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government purposes. The Department of Energy will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan.1

Funding Data

  • U.S. Dept. of Energy, Office of Fossil Energy (FEAA122; Funder ID: 10.13039/100006120).

References

1.
Pint
,
B. A.
,
2013
, “
High-Temperature Corrosion in Fossil Fuel Power Generation: Present and Future
,”
JOM
,
65
, pp.
1024
1032
.10.1007/s11837-013-0642-z
2.
Naslain
,
R.
,
2004
, “
Design, Preparation and Properties of Non-Oxide CMCs for Application in Engines and Nuclear Reactors: An Overview
,”
Compos. Sci. Technol.
,
64
(
2
), pp.
155
170
.10.1016/S0266-3538(03)00230-6
3.
Eaton
,
H. E.
, and
Linsey
,
G. D.
,
2002
, “
Accelerated Oxidation of SiC CMC's by Water Vapor and Protection Via Environmental Barrier Coating Approach
,”
J. Eur. Ceram. Soc.
,
22
(
14–15
), pp.
2741
2747
.10.1016/S0955-2219(02)00141-3
4.
Bennett
,
A.
,
1986
, “
Properties of Thermal Barrier Coatings
,”
Mater. Sci. Technol.
,
2
(
3
), pp.
257
261
.10.1179/mst.1986.2.3.257
5.
DeMasi-Marcin
,
J. T.
, and
Gupta
,
D. K.
,
1994
, “
Protective Coatings in the Gas Turbine Engine
,”
Surf. Coat. Technol.
,
68-69
, pp.
1
9
.10.1016/0257-8972(94)90129-5
6.
Nelson
,
W. A.
, and
Orenstein
,
R. M.
,
1997
, “
TBC Experience in Land-Based Gas Turbines
,”
J. Therm. Spray Technol.
,
6
(
2
), pp.
176
180
.10.1007/s11666-997-0009-5
7.
Goward
,
G. W.
,
1998
, “
Progress in Coatings for Gas Turbine Airfoils
,”
Surf. Coat. Technol.
,
108–109
, pp.
73
79
.10.1016/S0257-8972(98)00667-7
8.
Pint
,
B. A.
,
Wright
,
I. G.
,
Lee
,
W. Y.
,
Zhang
,
Y.
,
Prüßner
,
K.
, and
Alexander
,
K. B.
,
1998
, “
Substrate and Bond Coat Compositions: Factors Affecting Alumina Scale Adhesion
,”
Mater. Sci. Eng. A
,
245
(
2
), pp.
201
211
.10.1016/S0921-5093(97)00851-4
9.
Stringer
,
J.
,
1998
, “
Coatings in the Electricity Supply Industry: Past, Present, and Opportunities for the Future
,”
Surf. Coat. Technol.
,
108–109
, pp.
1
9
.10.1016/S0257-8972(98)00642-2
10.
Itoh
,
Y.
,
Saitoh
,
M.
, and
Tamura
,
M.
,
2000
, “
Characteristics of MCrAlY Coatings Sprayed by High Velocity Oxygen-Fuel Spraying System
,”
ASME J. Eng. Gas Turbines Power
,
122
(
1
), pp.
43
49
.10.1115/1.483173
11.
Nicholls
,
J. R.
,
2003
, “
Advances in Coating Design for High-Performance Gas Turbines
,”
MRS Bull.
,
28
(
9
), pp.
659
670
.10.1557/mrs2003.194
12.
Gleeson
,
B.
,
2006
, “
Thermal Barrier Coatings for Aeroengine Applications
,”
J. Propul. Power
,
2
, pp.
375
383
.10.2514/1.20734
13.
Naumenko
,
D.
,
Pillai
,
R.
,
Chyrkin
,
A.
, and
Quadakkers
,
W. J.
,
2017
, “
Overview on Recent Developments of Bond coats for Plasma-Sprayed Thermal Barrier Coatings
,”
J. Therm. Spray Technol.
,
26
(
8
), pp.
1743
1757
.10.1007/s11666-017-0649-z
14.
Lance
,
M. J.
,
Unocic
,
K. A.
,
Haynes
,
J. A.
, and
Pint
,
B. A.
,
2015
, “
APS TBC Performance on Directionally-Solidified Superalloy Substrates With HVOF NiCoCrAlYHfSi Bond Coatings
,”
Surf. Coat. Technol.
,
284
, pp.
9
13
.10.1016/j.surfcoat.2015.08.067
15.
Leyens
,
C.
,
Fritscher
,
K.
,
Gehrling
,
R.
,
Peters
,
M.
, and
Kaysser
,
W. A.
,
1996
, “
Oxide Scale Formation on an MCrAlY Coating in Various H2-H2O Atmospheres
,”
Surf. Coat. Technol.
,
82
(
1–2
), pp.
133
144
.10.1016/0257-8972(95)02665-7
16.
Janakiraman
,
R.
,
Meier
,
G. H.
, and
Pettit
,
F. S.
,
1999
, “
The Effect of Water Vapor on the Oxidation of Alloys That Developed Alumina Scales for Protection
,”
Metall. Mater. Trans.
,
30A
, pp.
2905
2913
.10.1007/s11661-999-0128-3
17.
Onal
,
K.
,
Maris-Sida
,
M. C.
,
Meier
,
G. H.
, and
Pettit
,
F. S.
,
2003
, “
Water Vapor Effects on the Cyclic Oxidation Resistance of Alumina-Forming Alloys
,”
Mater. High Temp.
,
20
(
3
), pp.
327
337
.10.1179/mht.2003.039
18.
Pint
,
B. A.
,
Haynes
,
J. A.
,
Zhang
,
Y.
,
More
,
K. L.
, and
Wright
,
I. G.
,
2006
, “
The Effect of Water Vapor on the Oxidation Behavior of Ni-Pt-Al Coatings and Alloys
,”
Surf. Coat. Technol.
,
201
(
7
), pp.
3852
3856
.10.1016/j.surfcoat.2006.07.244
19.
Smialek
,
J. L.
,
2008
, “
Enigmatic Moisture Effects on Al2O3 Scale and TBC Adhesion
,”
Mater. Sci. Forum
,
595–598
, pp.
191
198
.10.4028/www.scientific.net/MSF.595-598.191
20.
Déneux
,
V.
,
Cadoret
,
Y.
,
Hervier
,
S.
, and
Monceau
,
D.
,
2010
, “
Effect of Water Vapor on the Spallation of Thermal Barrier Coating Systems During Laboratory Cyclic Oxidation Testing
,”
Oxid. Met.
,
73
(
1–2
), pp.
83
93
.10.1007/s11085-009-9170-1
21.
Pint
,
B. A.
,
Garner
,
G. W.
,
Lowe
,
T. M.
,
Haynes
,
J. A.
, and
Zhang
,
Y.
,
2011
, “
Effect of Increased Water Vapor Levels on TBC Lifetime With Pt-Containing Bond Coatings
,”
Surf. Coat. Technol.
,
206
(
7
), pp.
1566
1570
.10.1016/j.surfcoat.2011.06.008
22.
Lance
,
M. J.
,
Unocic
,
K. A.
,
Haynes
,
J. A.
, and
Pint
,
B. A.
,
2014
, “
The Effect of Cycle Frequency, H2O and CO2 on TBC Lifetime With NiCoCrAlYHfSi Bond Coatings
,”
Surf. Coat. Technol.
,
260
, pp.
107
112
.10.1016/j.surfcoat.2014.08.082
23.
Pint
,
B. A.
,
Unocic
,
K. A.
, and
Haynes
,
J. A.
,
2016
, “
The Effect of Environment on TBC Lifetime
,”
ASME J. Eng. Gas Turbines Power
,
138
(
8
), p.
082102
.10.1115/1.4032438
24.
Pint
,
B. A.
,
Haynes
,
J. A.
,
Lance
,
M. J.
,
Aldridge
,
H. L.
, Jr.
,
Viswanathan
,
V.
,
Dwivedi
,
G.
, and
Sampath
,
S.
,
2016
, “
Factors Affecting TBC Furnace Cycle Lifetime: Temperature, Environment, Structure and Composition
,”
Superalloys 2016
,
M. Hardy, E. Huron, U. Glatzel, B. Griffin, B. Lewis, C. Rae, V. Seetharaman, and S. Tin, eds.
,
TMS
,
Warrendale, PA
, pp.
727
734
.10.1002/9781119075646.ch78
25.
Viswanathan
,
V.
,
Dwivedi
,
G.
, and
Sampath
,
S.
,
2014
, “
Engineered Multilayer Thermal Barrier Coatings for Enhanced Durability and Functional Performance
,”
J. Am. Ceram. Soc.
,
97
(
9
), pp.
2770
2778
.10.1111/jace.13033
26.
Nowak
,
W.
,
Naumenko
,
D.
,
Mor
,
G.
,
Mor
,
F.
,
Mack
,
D. E.
,
Vassen
,
R.
,
Singheiser
,
L.
, and
Quadakkers
,
W. J.
,
2014
, “
Effect of Processing Parameters on MCrAlY Bond coat Roughness and Lifetime of APS–TBC Systems
,”
Surf. Coat. Technol.
,
260
, pp.
82
89
.10.1016/j.surfcoat.2014.06.075
27.
Lance
,
M. J.
,
Haynes
,
J. A.
, and
Pint
,
B. A.
,
2016
, “
The Effects of Temperature and Substrate Curvature on TBC Lifetime and Residual Stress in Alumina Scales Beneath APS TBCs
,”
Surf. Coat. Technol.
,
308
, pp.
19
23
.10.1016/j.surfcoat.2016.09.087
28.
Pint
,
B. A.
,
Lance
,
M. J.
, and
Haynes
,
J. A.
,
2019
, “
The Effect of Coating Composition and Geometry on TBC Lifetime
,”
ASME J. Eng. Gas Turbines Power
,
141
(
3
), p.
031004
.10.1115/1.4041309
29.
Gupta
,
D. K.
, and
Duvall
,
D. S.
,
1984
, “
A Silicon and Hafnium Modified Plasma Sprayed MCrAlYating for Single Crystal Superalloys
,”
Superalloys 1984
,
M.
Gell
, ed.,
TMS
,
Warrendale, PA
, pp.
711
720
.
30.
Unocic
,
K. A.
, and
Pint
,
B. A.
,
2010
, “
Characterization of the Alumina Scale Formed on a Commercial MCrAlYHfSi Coating
,”
Surf. Coat. Technol.
,
205
(
5
), pp.
1178
1182
.10.1016/j.surfcoat.2010.08.111
31.
Haynes
,
J. A.
,
Unocic
,
K. A.
, and
Pint
,
B. A.
,
2013
, “
Effect of Water Vapor on the 1100 °C Oxidation Behavior of Plasma-Sprayed TBCs With HVOF NiCoCrAlX Bond Coats
,”
Surf. Coat. Technol.
,
215
, pp.
39
45
.10.1016/j.surfcoat.2012.07.099
32.
Pint
,
B. A.
,
More
,
K. L.
, and
Wright
,
I. G.
,
2003
, “
The Use of Two Reactive Elements to Optimize Oxidation Performance of Alumina-Forming Alloys
,”
Mater. High Temp.
,
20
(
3
), pp.
375
86
.10.1179/mht.2003.044
33.
Unocic
,
K. A.
, and
Pint
,
B. A.
,
2013
, “
Oxidation Behavior of Co-Doped NiCrAl Alloys in Dry and Wet Air
,”
Surf. Coat. Technol.
,
237
, pp.
8
15
.10.1016/j.surfcoat.2013.07.068
34.
Lance
,
M. J.
,
Unocic
,
K. A.
,
Haynes
,
J. A.
, and
Pint
,
B. A.
,
2013
, “
Effect of Water Vapor on Thermally Grown Alumina Scales on Simple and Pt-Modified Aluminide Coatings
,”
Surf. Coat. Technol.
,
237
, pp.
2
7
.10.1016/j.surfcoat.2013.06.041
35.
Lance
,
M. J.
,
Thiesing
,
B. P.
,
Haynes
,
J. A.
,
Parish
,
C. M.
, and
Pint
,
B. A.
,
2019
, “
The Effect of HVOF Bond Coating With APS Flash Coating on TBC Performance
,”
Oxid. Met.
,
91
(
5–6
), pp.
691
704
.10.1007/s11085-019-09903-3
36.
Lance
,
M. J.
,
Thiesing
,
B. P.
,
Haynes
,
J. A.
,
Gildersleeve
,
E. J.
,
Sampath
,
S.
, and
Pint
,
B. A.
,
2019
, “
Effect of APS Flash Bond Coatings and Curvature on TBC Performance on Rod Specimens
,”
Surf. Coat. Technol.
,
378
, p.
124940
.10.1016/j.surfcoat.2019.124940
37.
Lance
,
M. J.
,
Haynes
,
J. A.
, and
Pint
,
B. A.
,
2018
, “
Performance of Vacuum Plasma Spray Bond Coatings at 900 deg and 1100 °C
,”
Surf. Coat. Technol.
,
337
, pp.
136
140
.10.1016/j.surfcoat.2017.12.062
38.
Gildersleeve
,
E. J.
,
Viswanathan
,
V.
,
Lance
,
M. J.
,
Haynes
,
J. A.
,
Pint
,
B. A.
, and
Sampath
,
S.
,
2019
, “
Role of Bond Coat Processing Methods on the Durability of Plasma Sprayed Thermal Barrier Systems
,”
Surf. Coat. Technol.
,
375
, pp.
782
792
.10.1016/j.surfcoat.2019.07.065