Materials Transactions, Vol. 52, No. 7 (2011) pp. 1515 to 1521 #2011 The Japan Institute of Metals
EXPRESS REGULAR ARTICLE
Flattening Characteristics of Ni20 Cr Thermal-Sprayed Coating Layers on Preheated SCM415 Substrates Dong Hwan Shin1 , Jae Bin Lee1 , Jae Lyoung Wi2 , Seungil Park1 , Namil Kim1 , Minhaeng Cho1 , Jong Min Kim1 and Seong Hyuk Lee1; * 1 2
School of Mechanical Engineering, Chung-Ang University, 221 Heuksuk-dong, Dongjak-gu, Seoul, 156-756, Korea Seoul Nitrocarburizing Co., 507-1, Dangha-dong, Seo-gu, Incheon, 404-818, Korea
The preheating effect of SCM415 steel substrates on the flattening behavior of Ni20 Cr flame-sprayed particles was investigated in a temperature range from 278 K to 523 K. In the present study, we calculated the particle temperature and velocity distributions of the continuous and discrete phases before collision with the substrate by using a computational fluid dynamics (CFD) code of Fluent (ver. 6.3.26). Particle velocity and gas temperature decreased rapidly in the radial direction while the particles traveled toward the substrate, suggesting that substrate size and distance from the nozzle should be carefully controlled to improve coating quality. The conventional flame spray gun was used to accelerate molten particles and K-type thermocouples were used to monitor the substrate temperature during a preheating process. Commercially available nickel-based Ni20 Cr particles with a mean diameter of 45 mm and 20 mm were used for experiments to examine the particle size effect on the coating characteristics. Herein we present FE-SEM images of coated layers on the substrates. As the substrate preheating temperature increased, the flatter surface morphology was seen at the interface between substrate and coating layer because of better wetting and corresponding higher shear adhesion strength. Moreover, the splat morphology was significantly dependent on the particle size. [doi:10.2320/matertrans.M2011068] (Received February 23, 2011; Accepted April 27, 2011; Published June 15, 2011) Keywords: thermal spraying, coating, preheating, flattening, computational fluid dynamics (CFD)
1.
Introduction
In recent years, the thermal spraying as a generic surface coating technique has been widely used in a variety of steel production industries such as mechanics, aeronautics, chemical and oil, electronic, automotive, mining, and so on.1–5) This technique provides thick coating layers over a large surface at a high deposition rate and improves surface properties of substrates such as appearance, adhesion, wettability, wear resistance, and thermal resistance. While many factors affecting coating quality and corresponding reliability have been reported in the literature,6–9) detailed technologies and fundamentals to control those factors are not comprehensively understood yet. In the thermal spray coating process, adhesive and cohesive quality between molten droplets and target substrates can be sufficiently controlled by three representative factors as follows; thermal properties, impact dynamics and operating conditions, and surface characteristics. For making appropriate bond coats on the substrate, we need better understanding of the detailed physics behind the melting process, deposition and solidification, and the flattening characteristics of thermal-sprayed particles. Moreover, adhesion quality highly depends on some important surface parameters such as the thermochemical properties, the surface state, the preheated temperature, and the existence of oxidation. In fact, deposition of protective coatings against wear, heat corrosion, and oxidation has been one of the major applications in thermal spraying technology.10) Composite particles consisting of a solid phase and a metallic binder play an important role in the development of thermally sprayed coatings with increased wear and corrosion resistance.11) Especially, most of thermal spray proc*Corresponding
author, E-mail:
[email protected]
esses such as conventional flame spray, HVOF, plasma spray, and cold spray have widely used the Ni or nickel alloy particles (e.g., Ni20 Cr, NiCrAlY, and NiCoCrAlY) for the surface treatment such as reclamation for a re-work and repair of damaged parts and bond coats.10,12–14) Therefore, the mechanical, tribological, and thermal properties of the thermal spray coatings were reported in the previous literature.10,12,13) In particular, Ni20 Cr particles dealt with in the present study have been widely used as a bond coat for a ceramic coating which can enhance corrosion resistance. El-Hadj et al.15) conducted numerical analysis to elucidate the influence of gas temperature on the flattening of impinging molten droplets during the thermal spray process. They reported that the gas temperature played an important role in yielding splat morphology and affecting adhesion on the substrates. Bandyopadhyay and Nylen16) carried out three-dimensional computational fluid dynamics simulations which considered several intermediate chemical substances, and they compared the predictions of gas velocity and temperature fields with the results obtained by using the global chemical reaction which considered only. They found that there were little difference between them, indicating that the global chemical reaction model was acceptable in simulating the thermal spraying process. Voyer et al.17) experimentally investigated electrically conductive flame sprayed Al coatings on diverse textile surfaces. They showed the influence of different process parameters and fabric materials on electrical conductivity and the microstructure of metal-fabric composites. Yang et al.18) studied flattening characteristics for copper particle coating on the preheated substrates, and examined wettability changes of different substrate temperatures by measuring static contact angles and the effect of elapsed exposure time during preheating process. In particular, two important issues were discussed as follows: what is the substrate effect of different durations
1516
D. H. Shin et al.
Fig. 1 Schematic and dimension of the calculating domain in the flame spray process. (cooling air through outer ring 10 orifices at 300 m/s, premixed fuel-oxygen through inner ring of 10 orifices at 289 m/s)
after preheating, and what is the influence of the preheating temperature of the substrates on the adhesion and wetting characteristics. Recently, Fukumoto et al.19) conducted extensive free falling experiments to deposit millimetersized molten Cu droplets on the AISI 304 substrates, and discussed the heat transfer characteristics depending on substrate temperature and ambient pressure. By using a visualization technique with a high-speed camera, they also analyzed the spreading behavior of molten droplets on the substrates and showed that enhanced heat transfer from splat to substrate could be achieved by taking preheated surfaces and reducing environmental pressures in the thermal spray process. In this case, the wetting of molten particles is enhanced during splat flattening, affecting solidification behavior. Seo et al.20) investigated the degradation behavior of sprayed CoNi- and CoCrAlY coatings with different particle sizes by using the vacuum plasma spray. They found that the small particles produced a relatively smooth surface, and as the particle size decreased, the oxide growth in the coating layer was slower compared to those using larger particles. Kulkarni et al.21) examined the variation of porosity depending on the particle size and suggested that as the particle size increased, the corresponding porosity increased in the coating layers. Although abundant research15,17–21) has been reported, the flattening behavior of thermal-spayed particles on preheated substrates has been still an open point, and there is a lack of information on flattening characteristics, substantially depending on substrates and coating properties. More effort should be made to yield useful experimental and numerical data in establishing greater reliability and controllability of thermal spraying processes for better coating quality and finding optimized procedures. Therefore, we report on the flattener characteristics of surface morphology and element distributions of flame-sprayed Ni20 Cr alloy particles on commercially available SCM415 substrates, widely used in various industrial applications, which are preheated up to 523.15 K. Moreover, we analyze the particle size effect on the flattening characteristics of coating layer by using commercially available two kinds of Ni20 Cr particles. To analyze the thermal and flow characteristics in a thermal spraying process, we present temperature and velocity distributions of continuous and discrete phases estimated
by computational fluid dynamics (CFD) simulations involving the global chemical reaction model. 2.
Numerical Details
In this study, the CFD simulations were conducted to predict the gas and particle temperatures prior to the impact of the particles on the substrates. We simulated the model of Metco 5P-II powder spray gun (Sulzer Metco Inc.) with the use of commercially available Ni20 Cr particles (43F-NS provided by Sulzer Metco Inc.) with a mean diameter of 45 mm for the numerical calculation. As seen in Fig. 1, an axisymmetric computational grid system was made by ICEMCFD 12.1 and the commercial CFD code (Fluent v.6.3.26) was used to calculate conservation equations for mass, momentum, energy, and species. The total computational time took 1 hour using the computer, 1.6 GHz and 4 GB RAM (Intel Core i7 CPU Q720), with 8 nodes. In particular, the fuel was acetylene (C2 H2 ) and the fuel-oxygen ratio was 3 : 1. The gas inlet pressure was 13 psi and it was converted to the velocity of 289 m/s with initial gas temperature of 278.15 K. The cooling air inlet velocity was 300 m/s with constant temperature of 278.15 K and the sample temperature was initially 523.15 K. Using the volume weighted mixing law of individual heat capacities, the heat capacity of the mixture is calculated.22,23) For the combustion model, we used the chemical reaction formulae of 2C2 H2 + 5O2 ! 4CO2 + 2H2 O and the global reaction. The volumetric eddy dissipation model was used for simulating the combustion process. Also, the standard k-" model was used and the chemical properties were predicted by CHEMKIN 4.1. A theoretical maximum combustion temperature under an adiabatic condition is calculated as Tad ¼ 3358 K. For predicting the particle behavior in the flame spray process, the discrete phase model (DPM) can be expressed as follows: @uP gx ðP Þ ¼ FD ðu uP Þ þ P @t 18 CD Re FD ¼ P dP2 24 dP juP uj Re
ð1Þ ð2Þ ð3Þ
Flattening Characteristics of Ni20 Cr Thermal-Sprayed Coating Layers on Preheated SCM415 Substrates Table 1 Process
Fig. 2
Experimental setup.
In eq. (1), FD ðu uP Þ represents the drag force per unit particle mass where u is the fluid phase velocity, uP is the particle velocity, is the fluid density, and P is the density of the particle. Here, is the fluid viscosity and dP is the particle diameter. As shown in eq. (2), Re is the relative Reynolds number that is defined as eq. (3). 3.
Experimental Details
The conventional flame spray process was used in the present study because its main advantages are cost-effectiveness and versatility, and it can easily be scaled up for mass production. In order to study existing conventional flame spray gun equipment, we used a Metco 5P-II and two different nickel-based Ni20 Cr particles (43F-NS and 43VFNS, Sulzer Metco Inc.) with mean diameters of 45 mm and 20 mm, respectively, to investigate the particle size effect on flattening characteristics. As shown in Fig. 2, fuel (acetylene) and oxygen were controlled by the flow-meter, and a spray distance from the nozzle was fixed 150 mm. For the enhancement of adhesion, relatively rough SCM415 substrates with a mean roughness (Ra ) of 6:26 1:04 mm were prepared by sanding process. Sample diameter and thickness were 30 mm and 10 mm, respectively. For a preheating of the targets, a flame spray gun was used without injecting particles. According to some literature,24–26) most of the adsorbed gas/condensation and oxidation layers can be removed by preheating the substrate. We monitored the surface temperature that was controlled in the range from 278.15 to 523.15 K by using K-type thermocouples within an uncertainty of 0:75%. A sample moving speed is very important in determining the coating layer thickness and uniformity. In the present study, a moving speed of the sample was taken as 500 mm/s to obtain an appropriate coating thickness. After thermal spray process, samples were cooled under atmospheric condition. Table 1 summarizes the detailed conditions used in the present experiment. 4.
Results and Discussion
In general, substrate and particle temperatures are key factors in forming an interface between substrate and coated layer because a temperature difference between impacting particles and substrate surface significantly affects initial solidification due to rapid cooling. Moreover, the gas flow and particle impact velocity are important factors that substantially affect the flattening characteristics appearing at an interface flattener between a substrate and a coated
Summary of experimental details. Details
Fuel-oxygen ratio
C2 H2 : O2 ¼ 3 : 1
Fuel-oxygen flowmeter
C2 H2 : O2 ¼ 35 : 35 psi
Gas inlet pressure
13 psi
Cooling air inlet velocity
300 m/s
Mean diameter of particles
45 mm/20 mm
Preheating temperature Sample moving speeds
1517
323.15, 373.15, 423.15, 473.15, and 523.15 K 500 mm/s
Environmental conditions
temperature = 5 1 C
Spray distance
150 mm
Others
43F-NS/43VF-NS K-type thermocouple 10 times cross the sample
humidity = 13 2%
layer. The present study conducted extensively CFD simulations to analyze thermal and flow characteristics of thermal spraying process because in-situ measurement of temperature and velocity is very difficult. Figure 3 shows the estimated temperature and velocity distributions of combustion gas for the thermal spaying procedure. For the temperature fields, we see that target surfaces are sufficiently covered by the hot gas, indicating acceptable spatial uniformity can be obtained in the thermal spraying process. Near the target sample, the gas temperature is approximately 1500 K and the temperature distribution decreases as the flow goes far from the nozzle because the combustion products and the unreacted fuel can absorb the heat generation from the combustion. For velocity fields, a circulation region can be also observed in the vicinity of the target sample, showing that particle velocity prior to impact on the substrate can be altered by the target shape and location relative to the nozzle exit. In the Fig. 4 showing the predicted radial gas temperature and velocity distributions, it is found that nearly uniform temperatures were present in the range of approximately 50 mm, indicating that the present samples are sufficiently covered in the spaying process. However, gas temperature decreases in the radial direction due to the radiation thermal transport of free jet flows. This means that flow and thermal characteristics in flame spray process should be carefully considered to ensure uniformity of surface temperature. Figure 5 illustrates the predicted particle temperature and velocity distributions with respect to the distance from the nozzle exit. In front of the nozzle exit, the estimated particle temperatures increased over 2000 K. Near the targets which were located at 150 mm from the nozzle exit, the particle temperatures decreased to 1500 K. As the particle moved from the nozzle to the target sample, the maximum difference of the particle temperature reached approximately 500 K due to thermal absorption by combustion products including carbon dioxide. The predicted particle velocities show big differences up to 200 m/s, resulting from a rapid decrease in the normal component of the gas velocity induced by a wall proximity effect. Therefore, particle velocities just before impact would be more useful and reasonable in determining
1518
D. H. Shin et al.
(a) gas temperature distribution
(b) gas velocity distribution Fig. 3
Predicted simulation results; (a) gas temperature (b) velocity distributions.
(a)
Fig. 5
The estimated particle temperature and velocity distributions.
(b)
Fig. 4 Axial gas temperature and velocity in the radial direction at the different locations: (a) gas temperature (b) gas velocity.
the thermal spraying conditions which can affect the coating characteristics in the respect that at different inlet gas pressures, the gas and particle velocity distributions can be considerably altered depending on target shapes and sizes. In the case of conventional flame spray, the flame velocities are in the range from 80 to 100 m/s, and flow rate of working and pressures depend on the type of torch,27) indicating that the predicted gas and particle velocities are acceptable for the conventional flame spraying compared to the previous literature.27) Figure 6 shows the EDS (energy dispersive spectroscopy) results using the BSE (backscattered electron) detector for the as-sprayed Ni20 Cr coating on a SCM415 substrate at a preheating temperature of 523.15 K. Well-coated layers can be seen and porous regions are rarely present. The hardness of the coatings was also evaluated using a micro-Vickers
Flattening Characteristics of Ni20 Cr Thermal-Sprayed Coating Layers on Preheated SCM415 Substrates
1519
Fig. 6 The EDS results using the BSE detector for Ni20 Cr layers coated on SCM415 substrate (preheating temperature of 523.15 K); accelerating voltage of 20 kV and magnification of 300.
Fig. 7 FE-SEM images of flame-sprayed Ni20 Cr particles on the different preheated substrates; image of (a) is unheated condition and images from (b) to (f) are preheated with each indicated temperatures.
hardness tester and the reported value is an average. The measured hardness is about 190 Hv for the SCM415 substrate, whereas the hardness value is about 320 Hv for Ni20 Cr coated layers, which is approximately 1.7 times higher. Figure 7 represents cross-sectional FE-SEM images, using SE (secondary electron) detector, of the microstructures from the fabricated coating layers. Usually, most metal surfaces exposed to air will be oxidized to cover a thin oxide film with a thickness of several nanometers. Even with substrate preheating, however, it has been reported that the oxide film thickness is too low for the effect of thermal contact resistance to be significant.24) In addition, Tran and Hyland28) reported that the splat morphology was not affected by the film oxidation thickness on the substrate surface through the simulation method. From Fig. 7, we can observe typical multilayer structures both on the unheated substrate and on the substrates heated from 323.15 K to 523.15 K. At relatively low preheating temperatures, a very rough interface was observed between two dissimilar layers, whereas with an increase in preheating temperatures, more flattened
interfaces are produced. Similar to the present results, Kuroda et al.29) found that the surface roughness was altered from 40 mm to 15 mm in the HVOF process for a preheated surface at 423.15 K, indicating that the preheating of surface can affect the change in surface roughness. Herein, we suggest two possibilities to make the interface flattened between the substrate and the coated layer as follows: one is the normal stress exerted on the substrate by very high gas momentum at the preheating stage. During a preheating process, we used flame torch without injecting particles and consequently very high momentum and temperature of gas phase may affect the change in surface roughness. In fact, the normal stress exerted on the substrate was estimated from the present CFD results and its amount reached approximately 5.0 kPa. It is conjectured that from the microscopic viewpoint, local temperatures of surface peaks seem to be much higher than the averaged surface temperatures measured by K-type thermocouples. Hence, hot gas pressure exerted on the surface would be one of the major factors which affect the change in surface roughness. The other possibility is such a
1520
D. H. Shin et al.
Fig. 8 Microstructure of the coating systems which generated by the different particle size; (a) bigger particle (45 mm); (b) smaller particle (20 mm).
direct mechanical deformation of surface peaks due to the impingement of molten particles with very high momentum in the thermal spraying procedure. A possibility of deformation of surface peaks increases with the increase in momentum of molten particles, resulting in variation of surface morphology. Perhaps, this possibility increases with the increase in the preheated temperature. Prior to our primary experiments, we conducted the preliminary tests to examine the preheating effect on variation of surface morphology by using flame torch without injecting particles. From microstructure images acquired at different preheating temperatures, we observed that there were clear differences of surface roughness with respect to the preheating temperature, supporting that the preheating effect on the surface roughness is obviously present. Nevertheless, two reasons referred above are not clear yet and requires further investigation in detail. Meanwhile, the surface flatness depending on the preheating temperature is associated with rapid cooling due to temperature difference between the substrate and molten droplets. As seen in Fig. 5, it is verified that impacting particles on the substrate are almost molten because their temperatures are above 1500 K that is comparable to the melting temperature. When the substrate temperature is very low, the rapid cooling makes a rough solidified layer which substantially affects the flattening of particles. It is also important to discuss the adhesion mechanism between the coating layer and the substrate which is closely associated with the mechanical properties. According to Yang et al.,18) the equilibrium contact angle significantly decreases with the increase in the preheating temperature, and such a good wetting state is generated by removing the adsorbed gas/ condensation through preheating of the substrate. Kuroda et al.29) investigated the preheating effect on adhesion strength, and suggested an empirical correlation between the adhesion strength and the preheating temperature. From our results and their discussion, the flat pattern of interface indicates improvement in adhesion strength, and more precisely, a flattened interface significantly depends on the substrate temperature and wetting degree of molten particles at the early stage of thermal spraying. Better wetting state by preheating produces the higher adhesion strength of the fabricated coating layer. From Fig. 7, in particular, if the substrate temperature exceeds 473.15 K, a remarkable change in the flatness does not occur, indicating that there
is a critical preheating temperature, depending on materials and thermal spray conditions. Figure 8 shows the microstructure images acquired by the FE-SEM to examine the particle size effect on the flattening characteristics. For large particles, un-melted particles are present and they are poorly adhered on the surface, whereas small particles (20 mm) are well adhered on the surface. The morphology of the splats changed from a disk-like shape to fragmented shape with the increase in particle size, as indicated by Kulkarni et al.21) Seo et al.20) also showed that the fragmented splats led to poor splat-splat adherence and consequently the formation of pores. They confirmed that the amount of porosity with bigger particle is greater than that using smaller particle because of un-melted particle and fragmented splat shape. As seen in Fig. 8, when using bigger particles, the interface between the substrate and the coated layer becomes flatter than that using smaller ones because of variation of surface morphology resulting from the increase in kinetic energy of fully or partially molten particles impinging on the substrate. 5.
Conclusions
This study explored the preheating effect and particle size effect of SCM415 steel substrates on the flattening behavior of thermal-sprayed Ni20 Cr particles. In addition, numerical simulation was conducted to estimate the particle velocity and temperature distributions. The FE-SEM images and the EDS analysis at the different preheating temperatures of the substrates were presented. Numerical and experimental conclusions were drawn as follows: The results of numerical CFD simulations show the target substrate surfaces are sufficiently covered by the hot gas, and the corresponding spatial uniformity is acceptably maintained in the thermal spraying process. There is a circulation region observed in the vicinity of the substrate. In addition, prior to the flame spray process, the flow and thermal characteristics should be carefully considered in detail to ensure whether surface temperature becomes uniform in order to product the reliable coating layers. The particle temperatures impinging on the substrates which were located at 150 mm from the nozzle exit decreased a maximum difference of approximately 500 K. The particle velocity near the substrate decreases up to 200 m/s due to the wall proximity effect.
Flattening Characteristics of Ni20 Cr Thermal-Sprayed Coating Layers on Preheated SCM415 Substrates
The experimental results clearly present the flattening effect caused by the preheating of the substrates. The rough interfaces are formed between the coated layer and the substrate at relatively low preheating temperature of the substrates, whereas more flattened interfaces are produced by increasing the preheating temperature of the substrates. Moreover, better wetting through preheating is closely associated with improvement of adhesion strength of the fabricated coating layer. Note that it was observed that for more than 473.15 K the extent of flattening was seldom changed, which means that there may be in existence a critical preheating temperature, depending on materials and thermal spray conditions. In particular, it is confirmed that the particle size can change in the degree of flattening which may affect the coating adherence at the interface between two dissimilar layers. Acknowledgements This work was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science, and Technology (2010-0028181). REFERENCES 1) Matthews and B. James: J. Therm. Spray Technol. 19 (2010) 1267– 1276. 2) J. Voyer: J. Therm. Spray Technol. 19 (2010) 1013–1023. 3) M. R. Ramesh, S. Prakash, S. K. Nath, P. Kumar Sapra and N. Krishnamurthy: J. Therm. Spray Tehnol. (2010) published online: 30 December (doi: 10.1007/s11666-010-9605-x). 4) M. Zirari, A. Abdellah El-Hadj and N. Bacha: Appl. Surf. Sci. 256 (2010) 3581–3585. 5) R. Ghafouri-Azar, S. Shakeri, S. Chandra and J. Mostaghimi: Int. J. Heat Mass Transfer 46 (2003) 1395–1407. 6) R. L. Williamson, J. R. Fincke and C. H. Chang: Plasma Chem. Plasma Process. 20 (2000) 115–124. 7) M. Fukomoto, M. Shiiba, H. Haji and T. Yasui: Pure Appl. Chem. 77
1521
(2005) 429–442. 8) M. Mellali, P. Fauchais and A. Grimaud: Surf. Coat. Technol. 81 (1996) 275–286. 9) C. Cano, E. Garcia, A. L. Fernandes, M. I. Osendi and P. Miranzo: J. Eur. Ceram. Soc. 28 (2008) 2191–2197. 10) M. Prudenziati and M. L. Gualtieri: J. Therm. Spray Technol. 17 (2007) 385–394. 11) V. V. Sobolev, J. M. Gullemany and A. J. Mrtin: J. Therm. Spray Technol. 6 (1997) 353–360. 12) C. Bartuli and R. W. Smith: J. Therm. Spray Technol. 5 (1996) 335– 342. 13) J. Tuominen, P. Vuoristo, T. Ma¨ntyla¨, S. Ahmaniemi, J. Vihinen and P. H. Andersson: J. Therm. Spray Technol. 11 (2002) 233–243. 14) H. Koivuluoto and P. Vuoristo: J. Therm. Spray Technol. 18 (2009) 555–562. 15) A. A. El-Hadj, M. Zirari and N. Bacha: Appl. Surf. Sci. 257 (2010) 1643–1648. 16) R. Bandyopadhyay and P. Nylen: J. Therm. Spray Technol. 12 (2003) 492–503. 17) J. Voyer, P. Schulz and M. Schreiber: J. Therm. Spray Technol. 17 (2008) 583–588. 18) K. Yang, M. Fukumoto, T. Yasui and M. Yamada: J. Therm. Spray Technol. 19 (2010) 1195–1205. 19) M. Fukumoto, K. Yang, K. Tanaka, T. Usami, T. Yasui and M. Yamada: J. Therm. Spray Technol. 20 (2011) 48–58. 20) D. Seo, K. Ogawa, T. Shoji and S. Murata: J. Therm. Spray Technol. 16 (2007) 954–966. 21) A. Kulkarni, A. Vaidya, A. Goland, S. Sampath and H. Herman: Mater. Sci. Eng. A 359 (2003) 100–111. 22) N. Autissier, D. Larrain, J. Van herle and D. Favrat: J. Power Sources 131 (2004) 313–319. 23) FLUENT 6.3.26 User’s Guide. 24) A. McDonald, C. Moreau and S. Chandra: Surf. Coat Technol. 202 (2007) 23–33. 25) M. Fukumoto, K. Yang, T. Yasui and M. Yamada: J. Solid Mech. Mater. Eng. 4 (2010) 107–118. 26) I. A. Polunina, A. A. Isirikyan, K. E. Polounine and S. S. Mikhailova: Colloids Surf. A 160 (1999) 141–146. 27) L. Pawlowski: The Science and Engineering of Thermal Spray Coatings, (John Wiley & Sons, Ltd., 2008). 28) A. T. T. Tran and M. M. Hyland: J. Therm. Spray Technol. 19 (2010) 11–23. 29) S. Kuroda, T. Inoue, A. Kishi, K. Yokoyama, H. Yumoto and Y. Gotoh: Proc. ITSC, ed. by Ohmori, (2004) pp. 138–142.
