Individual splats are the building blocks of any thermal spray coating. Near the coating-substrate interface, they affect coating properties like adhesion strength. This article examines the effect of substrate heating on droplet splashing. Nickel powder was plasma-sprayed onto a polished stainless steel substrate at various temperatures and the resulting splats were analyzed. Droplet splashing was observed experimentally for three different cases: low substrate temperature, high substrate temperature, and droplet-splat interaction. Mechanisms for splashing were explained with the help of computer-generated nickel droplet impacts. The article proposes that the jetting of molten metal is not triggered by the formation of a central splat but rather a solidified ring on the periphery of the splat. It was observed that, on substrates below 350 deg C, splashing is triggered by solidification at the edge of the spreading droplet. Interactions with previously deposited splats also cause droplets to splash.

Molybdenum splats were produced at three plasma conditions on steel substrates preheated to three temperatures. Morphology of splats and corresponding craters formed on substrates were observed; dimensions of splats and craters were measured with an optical non-contact interferometer. It is found that substrate is significantly melted and deformed upon impact of the droplet, which leads to the formation of flower like splats and craters. On average, only about 36 to 53 % of the areas covered by splats were in good metallurgical/mechanical contact with substrate. Normalized crater volume increases with droplet size and the contact is improved for the high particle energy/high substrate temperature condition as compared with low particle energy/medium substrate energy condition. Splat morphology and crater formation is explained based on impinging jet heat transfer model.

The wire arc spraying process, one of several thermal spray processes, gained a sizable part of the thermal spray market, however, more control is needed for this process to be used for high precision coatings. This study is aimed at investigating the liquid metal droplet formation process in order to identify methods for droplet trajectory control. A high speed Kodak imaging system has been used to observe the droplet formation for different operating conditions. Decreasing the upstream pressure and the current levels lead to the reduction in the asymmetric melting of both anode and cathode. By decreasing the interactions of the large eddy structures with the formed metal agglomerates one can achieve better control of the particle trajectories and jet divergence. Thus, coatings can be obtained with higher definition and improved reliability.

Many studies have been devoted to particle flattening and resulting splat cooling. However, if recent models allow to compute the particle flattening time evolution, very few experiments in spraying conditions have been achieved to back such calculations. The aim of this paper is to describe an imaging device allowing the visualization of particle impacts on cold and hot surfaces. This technique makes it possible : • to investigate the "impact mode" : splashing, deposition or rebound, • to link the particle parameters at impact and the substrate parameters to the observed impact mode, • and therefore, to have a better understanding of coating formation. It consists in a controlled atmosphere chamber where is followed the impact of a single particle on a substrate which can be inclined. The particle parameters prior to its impact are measured : its surface temperature by fast (100 ns) two-color pyrometry, its velocity and diameter by Phase Doppler Anenometry (PDA). The particle image during flattening, splashing or rebounding is given by a fast camera (exposure/delay time 100ns to 1ms) with possible multi exposures. The camera is triggered by the PDA and/or the pyrometer. It is then possible to calculate for each molten particle its Sommerfeld parameter characterizing its impact mode (rebounding, deposition or splashing) when no solidification occurs during flattening. The substrate are made of stainless steel 304L rapidly covered by alumina splats resulting in a Ra~5-6µm. They are kept at 300°C, temperature at which splats are disk shaped on smooth substrate (Ra<0.05µm). The very preliminary results obtained show that unmolten or partially molten particles rebound in all directions but not elastically : the rebounding particle velocity is 3 to 5 times lower than that of the impacting one. For fully molten particles, splashing occurs in all cases even for low Sommerfeld numbers. It thus seems that the substrate roughness plays a key role in splashing.

Yttria stabilized zirconia particles are plasma sprayed on polished stainless steel substrate. Starting powders are fused and crushed powder, and hollow spherical powder. Four types of the splat morphology, which are splash, rugged, gravel mounted, and disk splats, are observed. Splash and disk splats are fully melted particles, but rugged and gravel mounted like splats are partially melted particles Gravel mounted like splat is observed from only hollow spherical powder, and disk splat is observed in the case of high substrate temperature. It is found that the ratio of splat morphology changes with spraying parameters. Porosity of the coating from fused and crushed powder is higher and Young's modulus of that is lower than that from hollow spherical powder. The ratio of rugged and gravel mounted splats affect porosity and Young's modulus. Adhesive strength increases with the increase in the ratio of disk splat. So, the coating properties are improved by controlling splat morphology. KEYWORDS: Splat Morphology, Partially Melted Particle, Disk Splat, Porosity, Young's Modulus, Adhesive Strength

This paper is devoted to the study of alumina splats formation and the adhesion/cohesion of alumina coatings on oxidized 1040 steel substrates. When preheating with the plasma torch, a duplex oxide layer is formed with an upper hematite and an under magnetite. Measurements have been performed on oxide layers with 470 nm average thickness corresponding to preheating times of about 10 min. In these conditions, the relative thickness of each sub-layer can be controlled by monitoring the heating rate and the preheating temperature. When the hematite upper layer is thick, it is mostly broken by the droplet impact and correspondingly the resulting coating has a poor adhesion (≈ 34 MPa) the rupture occurring within the oxide layer. When the hematite upper layer is thin, the droplet impact breaks only partly the oxide layers and the rupture of the coating occurs both within the oxide layer and the substrate surface (≈ 40 MPa).

This article examines the splat shape at off-normal angles in plasma spray and investigates the relationship between the splat elongation ratio and the spray angle on various spray materials and particle sizes. The applicability of the model to the experimental results is investigated. In this study, six powders of aluminum, copper, and nickel as metal spray materials and of alumina, titania, and zirconia as oxide materials were sprayed. It was observed that the model agreed well with the elongation ratio of the experimental results in the range between 90 and 30 degree. The model values were over-estimated than the practical results at spray angles less than 15 degree.

The spreading process of an isothermal droplet impinging on flat substrate surface in plasma spraying is studied numerically in 2D cylindrical coordinate systems by using 'Marker-And-cell (MAC) Technique. The changes and distributions of the transient contact pressures upon substrate surface at flattening are calculated under different droplet conditions with different impacting velocities and densities. The simulated results show that the transient contact pressure is initially high and concentrates at a small contacting area, it then spreads and drops quickly while droplet flattens. The maximum pressure is located at the front of the droplet at early stage of deformation, which pushes the fluid moving quickly along substrate surface and results in lateral flow. The contact pressure is mainly related to the droplet density and impact velocity. The peak pressure reduces consistently along the substrate surface so that the splashing at the periphery of flattening droplet may occur to form a reduced disk like splat because of the falling of contact pressure in this region and the escaping of the evaporated gas from the droplet / substrate interface.

The effect of an alumina shell on stainless steel particles used in plasma spraying has been studied. The mean size of the injected particles is about 65 nm and the thickness of the alumina shell is 3 µm. The composite powder is plasma sprayed using a PTF4 type plasma gun with an internal injection 3 mm upstream of the nozzle exit. The results show that without preheating the substrate splats are extensively fingered and become circular when the substrate surface is preheated over 200°C. EDS analysis of the distribution of the various elements shows that the alumina either uniformly covers the stainless steel splat or is distributed in pieces over the surface. This behavior has been explained by collecting particles in flight and analyzing them. A composite stainless steel/alumina coating sprayed on a rough stainless steel substrate preheated to 400°C has been examined and compared with a pure stainless steel coating. Both hardness and cohesion are improved for the alumina coated particles due to the random distribution of alumina within the steel matrix.

NiCrBSi and Ni-50Cr coatings are deposited using High Velocity Oxygen-Fuel (HVOF) spray process under different spray parameters with two powders of different sizes to clarify the influence of melting state of spray particles on the adhesive strength of the coating. The adhesive strength of coating is estimated according to ASTM C633-79. The melting state of spray droplet is examined from the coating microstructure. It is found that the melting state of spray particles has significant effect on the adhesive strength of HVOF sprayed Ni-based coatings. The significant melting of spray particle does not contribute to the increase in the adhesion of HVOF metallic coatings. On the other hand, the deposition of partially melted large particle contributes to the substantial improvement of adhesive strength of HVOF coating. The subsequent coating presents a dense microstructure and yields an adhesive strength of over 76 MPa, which is doubled compared to the coating deposited with completely molten particles.

A free falling experiment was conducted as a simulation of a thermal spray process. A flattening behavior of the freely fallen metal droplet impinged onto a flat substrate surface was fundamentally investigated. The substrates were kept at various designated temperatures, and the substrates coated with gold by PVD were also prepared in order to investigate the effect of a wetting at the splat/substrate interface on the flattening behavior of the droplet. A falling atmosphere was atmospheric pressure nitrogen to prevent the oxidation of the melted droplet, and the experiments under low-pressure condition were also conducted. A transition of the splat morphology was recognized in atmospheric pressure nitrogen experiments, that is, the splat morphology on a room temperature substrate was a splash type, whereas that on a high temperature substrate was a disk type. The cross-section microstructure of the splat obtained on the room temperature substrate was an isotropic coarse grain, whereas that on the high temperature substrate was a fine columnar. The grain size changed transitionally with increasing the substrate temperature. Transition temperature on the gold-coated substrate was higher than that on the substrate without coating. The cross-section microstructure of the splat obtained under low-pressure was a fine columnar even on the room temperature substrate. The results indicate that the metal droplet wets better under low-pressure condition than in atmospheric pressure nitrogen condition, and the wetting has a significant role in the flattening of the droplet.