Understanding of microstructures and mechanical properties of thermal sprayed Ni-based coatings with Al and Mo addition

2.1 Twin wire arc spray

Fe-3.5Ni-1.29Cu alloy (HSLA-100), a high-strength low alloy (HSLA) steel, was used as the substrate sample. Its chemical composition was 3.5% Ni, 1.29% Cu, 0.5% Mn, 0.209% Si, 0.048 C, 0.037% Nb, 0.003% V, 0.025% Al, 0.002% S, 0.008% P, and Fe to balance. All substrates were subjected to Ni-based coating after roughening using Alumina Grit blasting to improve the bonding strength of the substrate and coating layer for 2 passes at a grit size of 20 mesh, air pressure of 6 bar, and distance of 200 mm. Twin Arc Spray Coating was performed under the process conditions listed in Table 1 using Auto Arc AVD HD 456 (Thermion, USA) equipment for three types of wire: Ni-5Al solid wires, Ni-20Al cored wires, and Ni-5.5Al-5Mo cored wires.

Table 1:

Process parameter of twin wire arc spray, TWAC process parameters used to spray coating

2.2 Microstructure and microhardness test of Ni-based coatings

For microstructure analysis of the coating layer and the substrate, polishing was performed using a SiC polishing cloth from 200 to 2400 grit and using 9 μm, 3 μm, and 1 μm diamond suspension and 0.04 μm colloidal. The characteristics of the coated layer and substrate were observed using a field emission scanning electron microscope (FE-SEM; MIRA3, TESCAN, accelerating voltage: 10.0 kV) and energy dispersive spectrometer (EDS, EDAX). Image analysis was conducted to measure the porosity and oxide area fraction of the deposited layers. To analyze the crystal with the addition of Al and Mo, x-ray diffraction (XRD; Rigaku, D/MAX 2500 V L/PC, 40 kv, 30 mAh) using Cu Kα radiation (λ =1.54 Å) was performed. Electron backscatter diffraction (EBSD; TESCAN CLARA, Oxford Nordlys Nano, accelerating voltage: 20.0 kV, current: 10 nA, step size: 1 μm) and ATEX-software were also used to analyze the grain size of the coating layers. A High-Resolution 3D Microscope (KEYENCE, VHX-7000) was used to measure the surface roughness of each coating layer. The Ra value was an average of three positions of two readings. The hardness distribution was observed at intervals of 0.1 mm from a point 1 mm from the top surface, according to the ASTM E384 standard, and the load was set to 0.1 kgf load for 10 s using the Micro Vickers hardness test with an HM-122 hardness test machine (Akashi Co., Tokyo, Japan).

3.1. Surface roughness

Figure 1 shows the analysis of the surface roughness of the coated sample on which the twin wire arc spray coating process parameters listed in Table 1 were applied with a High-Resolution 3D Microscope. It can be seen that the surface roughness changed according to the Al content of the Ni based coating and the addition of Mo. Ni-5Al, Ni-20Al, and Ni-5.5Al-5Mo have surface roughness of 13.6375 (±1.21) Ra, 10.8625 (±1.11) Ra, and 11.0125 (±1.05) Ra, respectively. When the molten coating material at the high temperature of the process cools before it comes into close contact with the substrate and exists as an unmelted part, a rough surface occurs, and it changes depending on the twin wire arc spray process parameters such as feed speed and coating voltage [14]. Particles that are not completely melted during the deposition process are solid, so they are coated while maintaining their original shape, resulting in a rough surface. That is, it is coated while minimizing shape change.

Figure 1:

Surface roughness images of (a) Ni-5Al, (b) Ni-20Al, (c) Ni-5.5Al-5Mo coated layers and values of the coating materials

3.2 Microstructure of coating layers

The secondary electron (SE) images of the cross-section of the sprayed sample are shown in Figure 2. As shown in Figure 2, a typical microstructure of arc sprayed coating can be seen, with a continuous deposited layer of splat caused by molten particles colliding with the substrate and the lamellar structure, including porosity, crack, and oxides found on the split boundaries. The lamellar structure is also formed when the coating material particles are not completely melted or melted before reaching the substrate but are re-solidified or unmixed with oxides formed by their reaction with oxygen in the atmosphere.

Figure 2:

Microstructure of the surface and cross section of (a) Ni-5Al, (b) Ni-20Al and (c) Ni-5.5Al-5Mo and EDS results at the red cross markers

The black parts in (a) and (b) of Figure 2 are porosity, cracks, and oxides caused by a failure to completely melt between the split and another split [8], and the gray parts are NiAl intermetallic phase, the light gray parts are the Ni₃Al intermetallic compounds, and the white parts are cracks and oxides. The red cross markers of A(b) and C(a) are NiAl and Ni₃Al, respectively, as verified by EDS analysis. The porosity of the black part is caused by the expansion of trapped air during the melting process. Microcracks are caused by rapid solidification, which occurs through intensive heat transfer to the substrate as it deforms or protrudes from the molten parts that hit the substrate surface at a high speed. In particular, cracks formed at the interface are caused by the difference in coefficients of thermal expansion between the coating material (layer) and the substrate material, and the residual tensile stress generated during cooling. Oxides are formed on the surface of the splats when the molten particles are oxidized and deposited on the substrate. It was confirmed through EDS that the oxide generated at this time was Al₂O₃, as shown by the red cross markers in the surface (c) of Figure 2. As shown in Figure 2, cross section images of (a) Ni-5Al and (c) Ni-5.5Al-5Mo show a lot of porosity at the interface between the deposition layer and the substrate. But the cross section image of (b) Ni-20Al show a relatively low porosity and cracks, and thus it can be seen that excellent bonding was formed on the substrate to make high-quality contact.

Figure 3 shows the measurement results of the SEM images of the surface and cross-sectional Ni-based coating with oxides and porosity using an image analysis program. The reason that the area fraction of the surface and the cross-section is different is deduced to be that as the Al content increases during the spraying process, the adhesion force of Ni-Al and Ni-Al-Mo particles to the matrix increases relatively, so the porosity area fraction on the surface increases. On the other hand, considering the cross-section, it is believed that a dense coating layer with low porosity was formed by compression during the pass process.

Figure 3:

Area fraction of porosity, oxides, and intermetallic phase of (a) surface of Ni-5Al, Ni-20Al, and Ni-5.5Al-5Mo coating

In addition, as the amounts of intermetallic compounds increase, Al dissolves in Ni and becomes more likely to form a solid solution [15], which means they act as the driving force leading to the secondary phase, resulting in an increase in the number of intermetallic compounds as the aluminum content increases. Additionally, as illustrated in Figure 3, NiAl and Ni₃Al intermetallic phases are present in Ni-Al and Ni-Al-Mo coatings. These intermetallic compounds are also thermodynamically involved in the bonding strength between the coating layer and the substrate. According to thermodynamic Equation 1 and Equation 2 below, an exothermic reaction occurs when the Ni-Al secondary phase is formed. ΔH refers to the heat released during the reaction and depends on the reaction temperature. This is because the exothermic reaction provides a metallic bond between the coating and the substrate [15].

3.3 Crystal analysis

The X-ray diffraction values of Ni-5Al, Ni5.5Al-5Mo, and Ni-20Al are shown in Figure 4. The detected XRD peaks indicate that as the Al content increases, the widening of the diffraction peaks moves to a lower Bragg angle, which means an increase in the lattice parameters. As shown in Figure 4, it is confirmed that FWHM increased as the Al content increased, and it is inferred that the increase in the element content causes fine grains (crystallite), equivalently the grain (crystallite) size decreases. Al and Mo elements cause the diffraction peaks to shift slightly to the left with a small 2-theta value [15]. This is because, according to the Bragg lattice equation (Equation 3), FCC-structured lattice expansion occurs as the interplanar spacing (d) increases leading to a decrease in theta given a constant wave length. When the peak position is shifted left (2 theta smaller), it is interpreted that the d value increases, and when the peak width (full width half maximum) widens, the grain size decreases. It is the change of FWHM that means increase or decrease of grain size. Since the wave length is constant, interplanar spacing (d) increases as theta decreases, that is, it causes FCC-structured lattice expansion, which is caused by a solid solution of Al and Mo, leading to micro-strain in the lattice. In other words, atoms with an atomic radius larger than Ni are dissolved in the matrix, causing FCC lattice expansion. The atomic radius of Ni, which is a solvent, is 1.24 Å, and the atomic radius of Al and Mo, which are solutes in the Ni matrix, are 1.43 Å and 1.39 Å, respectively.

Figure 4:

XRD analysis of Ni-5Al, Ni-20Al and Ni-5.5Al-5Mo

In Figure 5, grain size was analyzed using EBSD to confirm the expansion of the FCC lattice structure through XRD. As a result of analyzing using EBSD, the grain size changed with the addition of Al and Mo. Figure 5 shows that Ni-5Al has a grain size of 4.411 μm, Ni-20Al has a grain size of 3.537 μm, and Ni-5.5Al-5Mo has a grain size of 4.018 μm. The size decreased as the added elements increased, suggesting that an atomic radius of approximately 12% of Ni atom size caused an expansion of the FCC structure. This coincides with the results of changes in the shifted peak and FWHM in the XRD graph in Figure 4, suggesting thath this leads to the refinement of grain size.

Figure 5:

IPF images of (a) Ni-5Al, (b) Ni-20Al, (c) Ni-5.5Al-5Mo, and (d) grain size of coating materials

3.4 Hardness test

Micro Vickers hardness was analyzed to evaluate the mechanical properties of the coating materials. Figure 6 shows the results of the micro Vickers hardness test on the coating layer surface according to the Al and Mo contents. The results show that the coating layer of Ni-5.5Al-5Mo was deposited to be thicker than that of Ni-5Al and Ni-20Al. The hardness test results for Ni-5Al in the horizontal direction is 182.46 (± 11.78) HV, for Ni-20Al is 404.48 (± 14.99) HV, and for Ni-5.5Al-5Mo is 217.72 (± 7.78) HV. These mechanical characteristics are consistent with Figure 5, which show the results of the crystallite size (grain size) of each coating layer using EBSD analysis. This suggests that the addition of Al and Mo caused a decrease in grain (crystallite) size and solid-solution strengthening.

Figure 6:

Micro hardness values of the coating layers

Author Contributions

Conceptualization, J. Bang and E. Lee; Methodology, J. Bang; Software, J. Bang; Validation, J. Bang, H. Kwon, E. Byon, and E. Lee; Formal Analysis, J. Bang; Investigation, J. Bang, H. Kwon, E. Byon, and E. Lee; Resources, J. Bang; Data Curation, J. Bang; Writing—Original Draft, Preparation, J. Bang; Writing—Review & Editing, E. Byon and E. Lee; Visualization, J. Bang; Supervision, E. Lee; Project Administration, E. Lee; Funding Acquisition, E. Lee. All authors have read and agreed to the published version of the manuscript.