Materials Processing
Durability of Dense Alumina Coating Deposited by Hybrid Aerosol Deposition under High-Speed Steam-Jet at Elevated Temperatures
2024 Volume 65 Issue 4 Pages 398-404
Details
2024 Volume 65 Issue 4 Pages 398-404
In pursuit of achieving zero-emission power generation, the utilization of carbon-free fuels like H2 holds promise for enhancing the reliability of the next-generation turbines. To realize the necessary of new operational environment, the implementation of dense environmental barrier coatings (EBCs) becomes essential. This study delves into the investigation of the durability of dense Al2O3 coatings deposited using the Hybrid Aerosol Deposition (HAD) method under the harsh condition of a high-speed steam-jet test, operating at approximately 125 m/s and elevated temperatures. The durability assessment encompasses uncoated SUS304 substrates and SUS304 substrates with dense HAD Al2O3 coatings deposited on one side. These samples underwent a rigorous 20-hour steam-jet test within a temperature range of 600–800°C.
Results indicate that the erosion rate of uncoated SUS304 substrates steadily increased with temperature, reaching a recession rate of 4.4 µm/h at the point of impingement. Conversely, the erosion rate was nearly halved following the deposition of HAD Al2O3 coating on one side of the substrates. The dense 8–9 µm Al2O3 coatings applied via HAD exhibited exceptional environmental protection during continuous exposure to the high-speed steam-jet at temperatures up to 800°C for 20 hours. Furthermore, the HAD layer effectively prevented oxygen penetration into the substrate. Post-test analysis revealed no significant features at the coating-substrate interface. Importantly, there were no alterations in the lattice parameter of Al2O3 crystal post-test. The coatings remained exclusively composed of the α-Al2O3 phase, with gradual crystallinity recovery driven by thermal effects during the steam-jet test and increasing temperature. These findings underscore the robust durability of HAD Al2O3 coatings in demanding high-speed steam-jet environments, making them a promising solution for enhancing power generation reliability.
Gas-turbine engines have played a pivotal role in energy, transportation, and defense sectors globally, significantly impacting the global economy.1) However, these engines face challenges related to high operating temperatures that exceed the durability of turbine materials, making material capability a bottleneck. To address this, thermal barrier coatings (TBCs) are widely employed in the hot sections to protect against high-temperature combustion gases.1–4) Ceramic materials with low thermal conductivity are the go-to choose for TBCs in applications ranging from gas turbine blades to internal combustion engines, jet engines, and land-based gas turbines.1–3) Transitioning to carbon-free fuel like H2 is a promising avenue for achieving zero-emission gas turbines and power generation, with the aim of improving next-generation turbine reliability by 2050.5) H2 offers unique potential for replacing the current fuels without CO2 and utilizing existing infrastructure. However, this shift requires the development of protective environmental barrier layers (EBCs) to accommodate the new operating environment in H2 turbine power generation. The characteristics and requirements of EBC layers have been extensively discussed,6–10) with a major focus on achieving dense, crack-free coatings that exhibit self-healing stability and resistance in high-temperature steam environments. While various coating technologies, including thermal spraying, are being actively explored on a global scale,6,7) there is a pressing need for further optimizations of EBC formation processes and the introducing new coating technologies to advance EBC technology. Although research in this field is vibrant and continuously advancing, most of the research is undergoing in the public domain, and much of the findings and development work remains proprietary.
In this context, non-melt deposition technologies like aerosol deposition (AD) and hybrid aerosol deposition (HAD) have played a significant role in enabling the deposition of dense layers at room temperature.11–16) HAD, a novel coating process that incorporates mesoplasma assistance into the conventional AD method, has emerged as a promising approach.12–16) It not only compensates for the deposition rate limitations of AD but also facilitates deposition on complex 3D objects. HAD offers substantial advantages, including the capability to deposit dense, homogeneous nanostructured ceramic coatings, surpassing conventional thermal spray (TS) coatings, and sealing porous materials which were difficult by conventional AD.12,17) Consequently, HAD holds great promise for various novel sustainable applications, including EBCs. Despite the need for dense coatings and new technologies, assessing the durability of EBCs without expensive engine tests remains a significant challenge. There is a pressing need for a cost-effective, lab scale approach to simulate turbine environments, investigate durability, provide size flexibility, and offer extended exposure times with elevated temperatures and H2O vapor. While a few reports exist on laboratory test facilities for characterizing EBC durability and stability,18) further research is essential.
This study aims to investigate the durability of dense Al2O3 coatings fabricated using HAD under the high-speed steam-jet conditions at elevated temperatures. Additionally, we will report on the development of a simplified, laboratory-scale steam-jet test facility for assessing the environmental durability of coatings and materials.
Fine α-Al2O3 powder (Showa Denko K.K., Japan) was utilized as the starting feedstock materials, featuring an average particle size of approximately 1 µm. Figure 1 offers a schematic illustration of the HAD system, and the comprehensive details of the process are reported previously.14,15) To assist aerosol deposition, a radio-frequency (RF) inductively coupled plasma was utilized. The Al2O3 particles were supplied by aerosol flow mixed with helium He as the carrier gas. The generated powder aerosol underwent acceleration through a pressure drops, transitioning from the aerosol unit into the deposition chamber. The powder was injected into the center of the HAD system’s plasma jet. Table 1 offers a summary of the HAD spraying parameters employed in this study. A plasma of 0.5 kW was generated at a reduced pressure of approximately 400 Pa. The coating was applied to type 304 stainless steel, commonly referred to as SUS304 in accordance with the Japanese Industrial Standards (JIS).
Schematic illustration of the hybrid aerosol deposition (HAD) process.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/MT-T2023003tab01.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
To evaluate the durability of the samples under steam exposure, we designed a custom steam-jet apparatus for testing. The set-up is schematically represented in Fig. 2 and is based on using two conventional horizontal tube furnaces. The first, a small tube furnace with a high-purity alumina tube featuring an inner diameter of 30 mm was used to heat and vaporize H2O. This small tube furnace is directly inserted into a larger tube furnace with a high-purity alumina tube boasting an inner diameter of 50 mm. Several modifications were made to enhance the system, including the integration of various components for introducing water steam-jet, thermocouple placement, sample handling at the center of the furnace and modifications to the other end of the second tube furnace to facilitate H2O exhaust.
Schematic illustration of developed steam-jet test.
During testing, purified water was initially injected from a pressurized water tank at 0.1 MPa into the small tube furnace to ensure it was heated, vaporized, and transformed into 100% H2O steam within the furnace. The steam was then directed from the small tube furnace to the second tube furnace using a capillary Pt-10Rh tube nozzle (with an outer diameter of φ1.6 mm and an inner diameter of φ1.0 mm). The large expansion vaporization from this capillary design accelerated the formation of the steam-jet, resulting in a modest flow rate of liquid water at 3 cc/min, producing a high-speed steam-jet with an average velocity of 125 m/s (measured using a pitot tube anemometer).
Figure 3(a) provides a real photograph of the developed H2O steam-jet. The heating and evaporation section ensured that the water entered the hot zone of the main furnace in the form of a high-speed steam-jet directly aimed at the designated samples. The samples were positioned in the center of the hot zone of the furnace using specimen holder as shown in Fig. 3(b). The tested samples were orientated at a 45° angle relative to the steam-jet. This setup guaranteed complete water vaporization and the continuous generation of a high-speed steam-jet on the sample surface. Additionally, the current configuration allows flexibility in varying sample shapes and sizes. In this study, flat plates measuring 20 mm × 20 mm with a thickness of 2 mm were utilized as specimens. This paper will investigate the performance and durability of using HAD dense coatings in comparison to uncoated SUS304 substrates under a 20-hour steam-jet test at different temperatures of 600, 700 and 800°C, respectively.
Photograph images of (a) developed steam-jet and (b) the sample holder to keep the sample in the center of the hot zone in the furnace.
A field emission scanning electron microscopy (FE-SEM, UltraPlus, Carl Zeiss Microscopy, Germany) equipped with an energy dispersive X-ray (EDX) detector, was used to observe the microstructure, and perform elemental analysis. The coating’s phase was identified using an X-ray diffractometer (XRD, RINT-2000 model, Rigaku, Tokyo, Japan) with a Cu Kα radiation. The crystallite size of the feedstock and the coatings was calculated from the XRD spectrum using the Scherrer Equation method. To measure the erosion depth after the steam-jet test, a 3D digital microscopic (RX-100, HIROX, Co., Ltd., Tokyo, Japan) was used.
Figure 4 display photographic images of the uncoated SUS304 substrate before and after the steam-jet test at temperatures of 600, 700 and 800°C, respectively. The damage to the substrate increased gradually with rising the temperature and was particularly significant at 800°C. The 3D digital microscopic analysis was employed to quantify the recession depth. Figure 5 presents the recession depth at the impingement point of uncoated SUS304 as a function of temperature after the 20-hour steam-jet test, along with the 3D surface profile of the impingement point at 800°C. The recession depth gradually increased with the temperature, and it reach to 88.5 µm at 800°C. Flipping the image under the 3D microscope revealed a mountain-like peak measuring 88.5 µm in height, signifying the maximum recession at the impingement point and confirming the erosion depth at that location. This corresponds to a recession rate of 4.4 µm/h at 800°C.
Photograph images of (a) as received SUS304 specimen, and (b)–(d) uncoated SUS304 specimen after steam-jet test for 20 h at different temperature 600–800°C, respectively. Note: the arrow indicates the direction of steam-jet and the impingement point.
The recession depth at the impingement point as a function of the temperature after the steam-jet test for 20 h and the 3D surface profile of the impingement point at 800°C of uncoated SUS304 substrate.
Dense Al2O3 coatings were deposited using the HAD process on one side of SUS304 substrates and investigate the coated substrate’s durability under the same conditions as discussed in the above section. The steam-jet was directed at the coated side of SUS substrates. Figure 6 presents photographic images of the Al2O3 coatings before and after a 20-hour steam-jet test for at temperatures ranging from 600 to 800°C. The images show that no significant damage was observed in the coating samples at lower temperatures, up to 800°C. It’s worth noting that, although there is a dark discoloration around the impingement point at 700°C, SEM observation did not reveal any damages, as shown in Fig. 6(e). This color change is currently under investigation.
(a) Photograph image of as sprayed Al2O3 coating, (b)–(d) Al2O3 coatings after steam-jet test for 20 h at different temperature 600–800°C, respectively, and (e) SEM surface microstructure of the impingement point of the Al2O3 coating at 700°C. Note: the arrows indicate the direction of steam-jet and the impingement points.
The stability of the Al2O3 coating after the steam-jet test was investigating through XRD phase analysis at all temperatures, as descried in Fig. 7. The spectra indicate that the as-sprayed HAD coating consists solely of the α-Al2O3 phase, which matches the phase composition of the starting material powder, in agreement with previous reports.15) After the steam-jet test, the spectra show that the Al2O3 coating remains, indicating the durability and stability of the coating after the steam-jet test. Additionally, the XRD spectra clearly demonstrate no change in the lattice parameter of the Al2O3 crystal, as estimated from the XRD spectra, and the coatings consist exclusively of the α-Al2O3 phase. Furthermore, the peak intensity recovered and gradually improved after the steam test and with increasing temperature.
XRD spectra of the HAD coating before and after steam-jet test at different temperature for 20 h.
The magnified XRD spectra, calculated crystallite size using Scherrer’s method, and the Full Width at Half Maximum (FWHM) values of the (104) peak for the starting Al2O3 powder, HAD coating before and after steam jet test, are illustrated in Fig. 8. The XRD spectra of the as-sprayed HAD coating reveal peak broadening and shifting to lower angles compared to the starting feedstock. The broader Al2O3 peaks in HAD coatings are primarily attributed to reduced crystallite size, grain refinement (nano-scale grains), during HAD deposition through the room temperature impact consolidation mechanism, (RTIC).11,15) The brittleness of the Al2O3 particles leads to fracturing of crystallites during HAD deposition, resulting in distorted crystallinity upon impacting the substrate surface. This is consistent with the reduction in crystallite size from 54.47 nm to 8.69 nm fragments for the (104) peak after HAD. The decrease in grain size and/or induced micro strain causes peak broadening after HAD deposition, indicating indicates particle fracture and deformation during the RTIC mechanism.11,15) Additionally, a slight shift in the XRD spectra to a lower angle after HAD deposition was observed. Currently, we lack a definitive explanation for the observed shift to a lower angle. However, it is noteworthy that this lower angle shift has been reported in prior publications.11,19) The leading hypothesis posits that compressive stress induced during impaction may result in elongation along the out-of-plane direction, potentially causing the observed shift to a lower angle. It is important to note that we have not fully validated this phenomenon. Clarification of this occurrence demands a high-intensity X-ray diffraction analysis, ideally utilizing synchrotron radiation. While we are eager to explore this avenue, it currently lies beyond our present capabilities.
XRD spectra, calculated crystallite size (by Scherrer’s method) and the FWHM values of the (104) peak for the starting Al2O3 powder, HAD coating before and after steam jet test at different temperature for 20 h.
Conversely, after the steam-jet test, the peak intensity and crystallite size gradually increased with rising the temperature. Annealing and increased temperatures led to a decrease in peak width as defects were removed and crystallinity improved. The recovery of crystallinity and observed grain growth in the HAD coating after the steam test were evidenced by sharpened diffraction peaks and increased intensity with increasing temperature. FWHM values, indicative of crystallinity, showed a consistent decrease after the steam-jet test. Figure 8(b) illustrates that, following the test at 800°C, the FWHM values for the Al2O3 phase decreased compared to the as-sprayed coating and continued to decrease with increasing steam test temperature, reaching a 40% decrease at 800°C (from 0.96 in the as-sprayed coating to 0.57 after the 800°C steam-jet test). This suggests improved crystallinity of Al2O3 with thermal effects during the steam test, indicating stable phase retention without transformation, and enhanced grain growth. Notably, the average crystallite size for the (104), (113), and (116) peaks exhibited a similar trend as the (104) peak. Furthermore, it is noteworthy that the diffraction peak position after the 800°C test is almost identical to the position after the 700°C test, warranting further investigation.
Figures 9 and 10 display the cross-sectional microstructure and EDX elemental mappings of the HAD coating before and after the 20-hour steam-jet test at 800°C. The coating exhibits an average thickness of 8–9 µm with a uniform and dense microstructure, free from obvious cracks or pores, as depicted in Fig. 9. Following the steam-jet test at 800°C for 20-hours, the coating exhibits remarkable stability, maintaining its thickness, as demonstrated in Fig. 10. The elemental mapping analysis confirms that the coating consists solely of the Al2O3 phase, corroborating the phase analysis presented in Fig. 7. Most notably, the EDX analysis confirms the absence of significant features at the coating-substrate interface, underscoring the environmental protection provided the dense HAD coatings during the steam-jet test at 800°C for 20-hours. It is noteworthy that the EDX analysis identified the presence of iron (Fe) and chromium (Cr) impurities in the as-sprayed coatings. These impurities are primarily attributed to contaminations occurring during deposition and originating from the initial feedstock.
Cross-sectional SEM images and EDX mappings of the as sprayed HAD Al2O3 coatings before steam-jet test.
Cross-sectional SEM images and EDX mappings of the HAD Al2O3 coatings after steam-jet test at 800°C for 20 h.
To quantify the influence of the HAD coating on the erosion rate, we investigated the weight change per unit area as a function of the temperature for both of uncoated SUS304 samples and the one-side Al2O3 coated samples. Figure 11 illustrates the weight change as a function of temperature. At lower temperatures of 600 and 700°C, the SUS304 substrates exhibited gradual weight gain. However, at 800°C, the weight significantly decreased with increasing temperature. This behavior can be explained as follows: at low temperatures, the substrate surface oxidized with parabolic oxidation kinetics, resulting in weight gain. The behavior at higher temperatures is presumably attributed to the gradual erosion of the outer oxidized surface. This behavior transitioned from parabolic oxidation kinetics with weight gain to linear kinetics with weight loss. Thus, after an initial increase in oxide thickness, the layer peels off, and the subsequent fresh layer is subsequently oxidized. Upon reaching a significant thickness, this oxidized layer peels off again. The oxidation rates increased with temperature, leading to a gradual increase in the erosion rate at higher temperatures.
Weight change of the uncoated SUS304 substrates and one side Al2O3 HAD coated SUS304 substrates under steam-jet test as a function of the temperature.
On the other hand, the one-side coated samples also gained weight at lower temperatures up to 700°C and then exhibited weight loss with increasing temperature up to 800°C. While the coated samples displayed a similar behavior to the uncoated SUS samples, the values of weight change for the one-side coated samples were approximately half of those for the uncoated samples. This indicates that the HAD Al2O3 coatings on one side significantly reduced the erosion rate of SUS304 to half the values under high-speed steam-jet conditions for 20 hours at elevated temperatures. The observed weight change in the one-side coated samples is presumably associated with the oxidation behavior of the uncoated side (SUS304 side). Therefore, the weight change is related to the oxidation behavior of the uncoated side. It worth mentioning that the methodology involves comparing two sample types distinguished by the front side (coated and uncoated). Both types underwent the same experimental conditions, with the uncoated backside exhibiting uniform erosion behavior. The front sides, coated or uncoated, were treated under identical water vapor conditions and experimental parameters. The weight difference observed, as supported by SEM observations, is attributed to the main distinction in the front side (with or without coating), serving as an indicative measure of coating protection. Future investigations will address the precise quantification of these values by studying the erosion behavior of the uncoated backside (no vapor injection).
In this study, we successfully demonstrated a laboratory-scale simple approach for simulating turbine environments that offer cost-effectiveness, size flexibility, and extended exposure times to H2O vapor at elevated Temperatures. The durability of dense Al2O3 coatings deposited using the hybrid aerosol deposition (HAD) method was investigated under the harsh conditions of a high-speed steam-jet operating at 125 m/s for 20 h at spanning temperatures ranging from 600 to 800°C. A comparative analysis between one-side HAD-coated SUS304 substrates and their uncoated counterparts under identical testing conditions led to the following conclusions:
A part of this study is based on results obtained from a project, JPNP14021, commissioned by the New Energy and Industrial Technology Development Organization (NEDO).
