Characterization of flame sprayed NiCrBSiMo coatings deposited with different spraying parameters

Haideé Ruiz-Luna^a,*, Karla O. Méndez-Medrano^b, Miguel Montoya Dávila^b, Víctor H. Baltazar-Hernández^b

^aCONACYT – Universidad Autónoma de Zacatecas, Unidad Académica de Ingeniería I, Av. Ramón López Velarde 801, 98000, Zacatecas, Zac., México

^bUniversidad Autónoma de Zacatecas, Unidad Académica de Ingeniería I, Av. Ramón López Velarde 801, 98000, Zacatecas, Zac., México

*Corresponding author: hruizlu@conacyt.mx

1. INTRODUCTIONTOP

Self-fluxing alloys are widely used in chemical, petrochemical and mining industries due to their brazing characteristics and corrosion and wear properties (Miguel et al., 2003; Navas et al., 2006; Zeng et al., 2016). These alloys consist of a Ni or Co matrix with additions of Cr, C, B, Si, Mo, Fe, and others, to improve their mechanical properties and corrosion resistance. For instance, Cr increases the hardness, corrosion and oxidation resistance (Otsubo et al., 2000; Miguel et al., 2003), additions of B and Si decrease the melting temperature and improve the self-fluxing capabilities of the alloys and also supports the formation of hard phases such as borides resulting in high hardness and wear resistance (Grigorescu et al., 1995; Otsubo et al., 2000; Navas et al., 2006). These Ni-based alloys are employed as surface protective coatings to improve material performance and to extend the lifetime of industrial components and parts. Thermal spray technologies are feasible alternatives to deposit high-quality self-fluxing coatings with extensive industrial applications (Miguel et al., 2003; Matsubara et al., 2007; Liyanage et al., 2010; Aussavy et al., 2014; Zeng et al., 2016). Nevertheless, the optimal properties of thermally sprayed coatings greatly depend on the deposition method, processing parameters conditions, and powder characteristics (Navas et al., 2006; Liyanage et al., 2010).

Combustion flame spraying technique is commonly used to deposit a wide range of materials, including Ni-based self-fluxing alloys, due to its low-cost, practical aspects and few environmental issues (Rodríguez et al., 2003; Navas et al., 2006; González et al., 2007a; Amokrane et al., 2011; Culliton et al., 2013; Karimi et al., 2016). Despite the above mentioned advantages; low particle velocity in flame spraying affects the microstructure and behavior of the coatings (Planche et al., 2005). However, the mechanical properties and microstructural defects in terms of porosity, oxide content, presence of unmolten particles and cracks of thermally spray coatings also depend on the proper selection of spraying parameters (Mrdak et al., 2009; Valarezo et al., 2010; Ruiz-Luna et al., 2014). Inappropriate selection and control of the deposition conditions result in a large amount of defects and therefore poor properties which reduces the performance of the coatings. In order to overcome such shortcomings and improve coatings quality; it is important to optimize the spraying parameters and to establish a correlation between the deposition variables and the resultant coatings characteristics. To date, previous investigations reported the microstructure and properties of flame sprayed coatings under a set of working conditions (González et al., 2007b; Culliton et al., 2013; Karimi et al., 2016) without further analysis about the optimization and correlation between the spraying parameters and coatings properties. Therefore, the aim of the present investigation is to analyze the effect of three process variables namely: stand-off distance, flame chemistry, and nozzle diameter, on the porosity content, hardness, and thickness of as-sprayed coatings, and to determine the processing conditions that allow the deposition of coatings with minimal microstructural defects using a more economical technique. The analysis is presented in two steps: in the first stage, the effect of the stand-off distance on the crystal structure and microstructure of the coatings is determined. In the second step, a simple 2² factorial design is employed and the effect of both: nozzle diameter and flame chemistry, on the crystal structure, porosity content, hardness and thickness of the coatings is presented and discussed.

2. MATERIALS AND METHODSTOP

2.1. Coating depositionTOP

Commercially available Ni-17.3Cr-4.5B-3.5Si-4.6Mo-2.9Fe-2.1Cu-1.2C (wt. %) powder was employed as feedstock material. The characteristics of the powder particles are shown in Fig. 1. The powder consists of mostly spherical particles although some irregular-shaped particles and particles with so-called satellites are observed in the micrograph (Fig. 1a). Distribution particle size ranges between 36 to 133 μm with a mean particle size of 84 μm (Fig. 1b) as measured by laser diffractometer (Helos/BR Sympatec). The SEM-EDS analysis of the cross section of the powder particle (inset of Fig. 1a), indicates that the microstructure consists of an enriched Ni matrix (bright gray phase) and a dark phase, mainly composed of Cr-rich particles, distributed in the Ni matrix.

The coatings were deposited using an oxyacetylene flame spray gun (CastoDyn DS 8000, Castolin Eutectic) onto grit-blasted low carbon steels plates of 50×25×4 mm. As mentioned, two stages were carried out in order to study the effect of the spraying parameters, (stand-off distance, flame chemistry, and nozzle diameter) on the coatings characteristics, including crystal structure, porosity percentage, hardness and thickness. On the first stage, the influence of the stand-off distance on the crystal structure and microstructure of the coatings was evaluated. Four stand-off distances were selected: 100, 150, 200, and 250 mm. At this stage, the coatings were deposited using neutral flame and the nozzle with diameter of 2.4 mm. On the second stage (Table 1), a simple 2² factorial design was set up to investigate the effect of the flame chemistry (two flame combinations, i.e. carburizing and neutral) and nozzle diameter using the porosity content, hardness and thickness of the coatings as response variables. The flame combinations were produced by changing the acetylene gas flow. The nozzles, designed specifically to the torch, are equal in length with a straight duct attached to the outlet of the converging section but its diameter is different (Table 1).

The oxygen and acetylene pressures were fixed to 4 and 0.7 Bar, respectively, in both stages. The number of torch passes over the substrate used in the first and second stages was one and three, respectively.

2.2. Characterization methodsTOP

X–ray diffraction (XRD, D8Advance Bruker diffractometer with Cu-Kα radiation) technique was used to characterize the crystal structure of the coatings. A scanning electron (SEM, Philips XL30) microscope equipped with and energy-dispersive X-ray spectrometer (EDS) and an optical (Union Versamet 3) microscope were employed to analyze the cross-section and thickness of the samples. The porosity content of the coatings was determined by image analysis (ImageJ software) from 10 polished cross-section images. A Vickers indenter (Shimadzu) with a load of 300 g_f and a dwell time of 15 s was used to measure the hardness of the coatings. An average of 10 indentations was carried out on each sample along their cross-section.

3. RESULTSTOP

3.1. Effect of stand-off distanceTOP

Figure 2 shows the diffraction patterns of the feedstock powder and the as-sprayed coatings deposited at various stand-off distances. As observed, the coatings retained the original phases of the powder particles indicating that no phase transformation occurred during deposition. Ni is identified as the main phase, whereas phases corresponding to CrB, Ni₃B, Ni-Si, and Cr₃C₂ are also observed.

Based on microstructure observation in Fig. 3, as well as on the fraction of porosity in Fig 4; the coatings show up markedly differences related to the spraying distance. Samples deposited at short stand-off distances (100 and 150 mm, Fig. 3a and b, respectively) exhibited high porosity (Fig. 4) and significant unmolten and semi-molten particles, whereas a more homogeneous and dense microstructure with few unmolten particles are distinguished for the coating sprayed at a distance of 200 mm (Fig. 3c). However, by further increasing the stand-off distance the porosity increases up to ~15%. According to the above mentioned; it can be stated that coatings with low porosity content and a more homogeneous microstructure were obtained using medium stand-off distances. Therefore, for the second stage, the coatings were deposited using a spraying distance of 200 mm and the experiments were focused on the effect of the nozzle diameter and flame chemistry.

3.2. Influence of flame chemistry and nozzle diameterTOP

The crystal structure of the coating sprayed employing the conditions shown in Table 1 are presented in Fig. 5. Similar to the X-ray results of the first stage, the analysis of the diffraction patterns of the as-sprayed coatings revealed that Ni is still identified as the major phase and the corresponding intensities of CrB, Ni₃B, Ni-Si, and Cr₃C₂ are also detected. A broadening in the 41 to 50° range it is also noted for the C1 and C2 coatings.

According to the microstructural features of the coatings deposited upon the conditions listed in Table 1, there is a clear relationship between the nozzle diameter and the resultant microstructure of the coatings (Fig. 6). Coatings sprayed with a higher nozzle diameter, C1 and C2 (Fig. 6a and b), show the presence of unmolten particles and coating defects in terms of splat boundary separation and voids resulting in high levels of porosity (~ 12%). In contrast, a lamellar structure with no visible unmolten or semi-molten particles and low coating porosity (2.7% in average) resulted for samples deposited with a lower nozzle diameter, C3 and C4 (Fig. 6c and d). The microstructure of the samples is quite similar irrespective of the flame chemistry; however, a slight reduction of porosity is observed when neutral flames are used, Fig. 6b and d, compared to the coatings deposited using a carburizing flame (Fig. 6a and c). A high magnification of the cross section of the flame sprayed coatings is presented in Fig. 7. The microstructure is quite similar to that of the powder particles (inset Fig. 1a). The coatings exhibit a homogeneous microstructure consisting of a bright matrix and fine precipitates uniformly distributed in the matrix (dark phase), irrespective of the spraying parameters. Compositional analysis revealed that the matrix is a Ni solid solution with Cr, Si, and Fe while the precipitates contain a high concentration of Cr with some additions of Si, Ni and Mo. Based on phase analysis previously reported for similar alloys, the matrix is identified as γ-Ni/Ni₃B and the precipitates embedded in the matrix primarily correspond to CrB, Cr₅B₃, Ni₃Si, Cr₃C₂, and Cr₇C₃ (Kim et al., 2003; Shrestha et al., 2005; Linayage et al., 2010; Hemmati et al., 2013; Wen et al., 2017). These results are consistent with the X-ray analysis which showed the existence of these phases (Fig. 5).

The results of the factorial design of experiments are presented in Fig. 8. The main effect of each factor on the response variables i.e. porosity content, Vickers hardness, and thickness of the coatings is displayed in Fig. 8a-c, respectively. As published elsewhere (Montgomery, 1991), these values were computed by averaging the response variable performed with the high (or low) level of each variable, irrespective of the value of the other variable.

In accordance with Fig. 8, the nozzle diameter significantly influences the coatings characteristics. The use of small nozzle diameter results in low porosity meanwhile large diameters increase the levels of porosity around 12% (Fig. 8a). Regarding the hardness and thickness of the samples, these responses are also favored by small nozzle diameters as shown in Fig. 8b and c. Hence, coatings deposited with small nozzle diameters (C3 and C4) exhibited higher hardness, thickness and reduced porosity as compared with samples sprayed with large diameters (C1 and C2).

It is important to notice that the flame chemistry barely affects the porosity and hardness, Fig. 8a–b, however an influence on the thickness of the coatings is observed (Figs. 6 and 8c). Thinner coatings with high hardness and low porosity percentage are obtained when using neutral flames.

4. DISCUSSIONTOP

The microstructure of the coatings is significantly influenced by the stand-off distance as pointed out in previous section. It has been reported that the substrate temperature, residence time of the particles in the flame, and the in-flight particle characteristics, i.e. velocity and temperature, depend on the spraying distance (Mrdak et al., 2009; Valarezo et al., 2010; Ruiz-Luna et al., 2014). For instance, at short spraying distances (for example 100 or 150 mm), particles impact the substrate at elevated temperature causing significant levels of splashing and therefore high porosity (Mrdak et al., 2009; Dobler et al., 2000; Valarezo et al., 2010; Ruiz-Luna et al., 2014). At intermediate distance, (200 mm) it seems that the powder particles attained an adequate velocity and temperature resulting in a balance between particle flattening and solidification which decreases the porosity content. At longer spraying distances (≥ 250 mm), the porosity of the coatings and the unmolten particles increase due to the limited particle deformation at impact. Hence, the powder particles reached the substrate in a resolidificated or cooled state because the residence time of the particles in the flame increases resulting in lower particle velocities (He et al., 2001; Valarezo et al., 2010; Ruiz-Luna et al., 2014).

As observed in the XRD patterns in Fig. 5, coatings C1 and C2 showed a peak broadening between 41 to 50° which may be associated to a slight amorphization or slow crystallization of the self-fluxing powder particles due to the high cooling rates attained during solidification as reported previously (Guo et al., 1995; Gruzdys et al., 2009; Zeng et al., 2009). For instance, Planche et al. (2005) have reported that the recrystallization of the NiCrBSi particles after impact is difficult because the presence of B and Si within the alloy.

According to the results of the second stage, the coating characteristics are also affected by the nozzle diameter. A variation in nozzle diameter seems to produce distinct energy of the gas which indeed is transmitted to the powder particles. Because in flame spray process the flame velocity (< 100 m·s⁻¹) (Pawloswski, 2008) is lower compared to the speed of sound; then it is to be assumed that the gas flow behaves as incompressible manner. In accordance to the principles of fluid dynamics, ducts with lower area exit increases the velocity of the flow and vice versa (Mott, 1996). Based on this, it is expected that the nozzle with lower diameter produces a higher kinetic energy of the particles than the nozzle with a large diameter and thus, distinct microstructural features due to the flattening degree of the particles is affected by the in-flight particle characteristics i.e., velocity and temperature (Planche et al., 2005). Therefore, particles upon C1 and C2 conditions have lower particle velocity than those of C3 and C4 conditions. Nozzle with large diameter (samples C1 and C2, associated with low velocities) are unable to spread and plastically deform the powder particles at impact. Thus, the microstructure of the coatings resulted with large unmolten and partially molten particles (Fig. 6a-b) that increases the porosity and decreases the hardness of the coatings, Fig. 8a-b, respectively. Thinner coatings probably result as a consequence of debonding of unmolten and cooled particles at impact. On the contrary, coatings C3 and C4 sprayed with lower diameter showed a lamellar structure with reduced porosity, these characteristics are attributed to a higher kinetic energy of the particles, compared with C1 and C2 conditions, which increases the splat cohesion and ensures a balance of plastic deformation and solidification at the impingement onto the substrate (Fig. 6c-d), and thereby low porosity and high hardness and thickness (Fig. 8).

The observation of the hardness reveals that this parameter is in direct correlation with the inherent microstructure of the coatings. As mentioned before, high kinetic energies improved the microstructure of the coatings by increasing the cohesion and density of the sprayed particles resulting in higher coating hardness. These conditions are achieved using small nozzle diameters and neutral flames (Fig. 8b).

It is worth nothing that the use of small nozzles contributes to reduce the porosity by about 4 times (from 12.1% to 2.7%) and increase the hardness from 900 to 1264 HV. Similar porosity (Miguel et al., 2003; Shrestha et al., 2005; Yu et al., 2013; Singh and Kaur, 2016; Wen et al., 2017) and hardness (González et al., 2007a; González et al., 2007b; Singh and Kaur, 2016) values have been reported for thermally sprayed Ni-based coatings. However, the porosity content obtained in the present study for C3 and C4 conditions (2.7% in average), is lower compared to those reported for as-sprayed self-fluxing coatings deposited by flame and plasma spraying (Higuera Hidalgo et al., 2001; Kim et al., 2003; Planche et al., 2005; Culliton et al., 2013; Bergant et al., 2014).

Concerning the flame chemistry, this parameter influences the temperature of combustion spray systems (Pawloswski, 2008); the maximum flame temperatures are reached near the stoichiometric or neutral condition as stated previously (Valarezo et al., 2010; Ruiz-Luna et al., 2014; Jadidi et al., 2015). This statement is in accordance with the microstructural results in this study where higher particle heating results in superior degree of particle flattening and bonding at impact and subsequent high hardness and low porosity and thickness (Fig. 8a-c).

In summary, the microstructure, pore content and hardness of the coatings are dependent on the deposition parameters; namely, spraying distance and nozzle diameter, and an optimum selection of these leads to an improvement of coatings properties. Proper deformation and spreading of the particles is attained when medium stand-off distances, small nozzles diameter and neutral flames are employed.

5. CONCLUSIONSTOP

ACKNOWLEDGMENTSTOP

Authors are gratefully acknowledged to Consejo Nacional de Ciencia y Tecnología (CONACyT - México) for financial support under projects No. 270613-INFRA-2016 and 254731-INFRA-2015. Authors also thank the program Cátedras-Conacyt and Aixa Ibeth Gutiérrez Pérez and Alma Gabriela Mora García for the laser diffractomer and electron microscopy measurements carried out at CENAPROT and LIDTRA National Laboratories, respectively.

REFERENCESTOP