This study assesses the quality of flame-sprayed alumina coatings produced from recently developed alumina cord using argon and compressed air as atomizing gases. Coatings of different thicknesses were deposited on aluminum substrates and then analyzed using optical microscopy, X-ray diffraction, and resistivity measurements. The coatings, particularly those sprayed with argon, had fine microstructure and higher surface and volume resistivity than flame-spray coatings made from alumina cord in the past. They were also found to have higher alpha phase content than plasma-sprayed coatings, regardless of the atomizing gas used. The effect of humidity and the possible formation of aluminum hydroxides are also addressed.

Alumina and yttria coatings, manufactured by suspension plasma spraying, were investigated to understand the “material effect” in the coating building. Some particle image velocimetry measurements were carried out to evaluate the particle velocities into the plasma. Some particle collections were performed to get information on their molten state. Splats were observed by scanning electron microscopy (SEM) and their dimensions were measured with an interferometric profilometer. Coating cross sections were finally observed by SEM and porosity rates were evaluated by image analysis and ultra-small angle X-scattering. This study revealed no real difference between the two materials concerning particle velocity. However, splat analyses highlighted a better flattening ratio for yttria particles, due to a lower difficulty to melt of this material. This property seems to enhance particle vaporization whose condensates are found on coating surface. These observations explain the difference of pore size distributions observed for both coatings.

Atmospheric Plasma Spray is widely used for tens of years to elaborate protective coatings on parts for several applications. However, our understanding of the APS process can still be improved, requiring a fine modeling of the process in parallel with some corresponding experiments. In the present work, a complete series of models was applied to reinforce our knowledge of the process: the case of an alumina coating was considered. A 3D CFD model was first used to study the internal arc within the torch. Interactions between the external plasma jet and the injected particles were then computed in a second step. At this level, the predicted in-flight particle characteristics were compared with some corresponding measurements recorded with the DPV 2000 diagnostic tool. A third model was then applied to investigate the particle flattening on the substrate/coating material. SEM pictures of coating cross-sections were then captured and a last model was finally applied to estimate the coating effective thermos-mechanical properties based on calculations performed directly on the SEM micrographs. This set of models allows investigating the APS process from the DC arc within the torch to the coating properties.

This study compares the morphology, porosity, and purity of yttria powders produced by spray drying, spray drying and sintering (SDS), and spray drying and plasma fusion (SDPF). The surface morphology of each type of powder is examined by SEM. Pore volume and density are determined by Hg porosimetry, and impurity concentrations are assessed via glow discharge mass spectrometry (GDMS). Coatings made from the powders by means of air plasma spraying are evaluated based on porosity, spray time, powder consumption, and embedded fine particles.

The development of new plasma spray torches reinforced the use of numerical modelling to help in the design steps. Most of the thermal spray material providers are thus now interested in understanding the arc behaviour inside the torch so that CFD studies focussed on this topic become numerous. Our first calculations performed on the ProPlasma HP gun assuming a laminar hypothesis have shown underestimations of the torch voltage and of the thermal losses in the cooling circuit, and a subsequent overestimation of the thermal efficiency of the torch. In the present study, the setup of different turbulence models was performed and a comparison was made between the results obtained using either the laminar assumption or several turbulence models. The calculations indicate that the results obtained using conventional turbulence models such as k-ε or k-ω type models do not significantly differ from those obtained using a laminar assumption, thus only more sophisticated models can be expected to improve the simulations.

Saint-Gobain Coating Solutions new ProPlasma HP gun, based on conventional plasma gun design (single anode-single cathode), has been evaluated for typical demanding applications of the thermal spray industry. A general overview of this gun main features is presented. Aim of its development was to fulfill the whole range of applications of plasma spraying, from high feed rate –high deposit efficiency (DE) to high velocity. First, plasma spraying performance of ProPlasma is exposed through 2 typical examples of materials extensively sprayed: yttria stabilized zirconia (YSZ) and chromia, both needing high DE at high feed rates. A set of spraying parameters was developed for YSZ Saint-Gobain #204 hollow spherical powder showing 57% DE at 150g/min and exhibiting 13 to 15% porosity as typically requested for thick TBCs. Two sets of parameters were developed for chromia at 120g/min leading either to 3% porosity - 51% DE or 2% porosity - 42% DE. Then, the possibility to reach a high velocity mode is shown through the presentation of the anode development made for a typical high velocity application material: WC-Co. After a first design showed experimental promising results with low porosity, the anode design has been further improved based on experiment and CFD modeling.

Thermal spray of Zn, Zn/Al, or Al is extensively used to make anticorrosion coatings on steel structures. Twin arc spray and wire flame spray are the two most practised processes to achieve such coatings. This paper presents measurements of particle emissions generated by these two processes. Sampling and analysis of aerosols generated by both processes have been carried out inside the exhaust ductwork using various instruments: an ELPI impactor, a CNC (Condensation Nucleus Counter), a TEOM microbalance and sampling filters allowing sampling for SEM observations. Electric arc spraying produced much more fumes of ultra fine particles than flame spraying. Aluminum spraying also produces large fume quantities compared to the Zn spraying under the same conditions. The aerosol comprised submicron particles and 95% of the numerical particle size distribution was less than 100 nm. The nanometric nature of the fume particles was confirmed by observations on the SEM. The strong dilution caused by compressed air has the effect of strongly limiting particle coagulation and, in turn, the size of the agglomerated particles. Electric arc spray has taken market share versus wire flame spray for Zn, ZnAl, or Al spraying, but this study shows that it generates much more particle emissions.

The present study is devoted to the modelling of the arc formation in a direct current plasma gun newly commercialized by Saint-Gobain Coating Solutions (Avignon, France). The CFD computations were performed using the FLUENT code and the electromagnetic coupling was taken into account on the basis of a three dimensional model. The main advances compared to previous works performed on the same subject are numerous. First of all, whereas most of earlier models include the arc region only, the CFD domain was here extended to the gas injection region (i.e. upstream part of the gun, including the gas injection sleeve), thus allowing a better description of the effect of the gas injection on the plasma flow. Second, whereas earlier works include the fluid domain only, the present model includes a fluid/solid coupling in the anode. In particular, the thermal and electromagnetic equations are solved not only in the fluid parts but also in the tungsten and copper parts of the anode. This change was found to be important because the internal surface of the anode is no more a boundary of the domain. Thus, its temperature (and electric potential) becomes variable and is thus not necessarily imposed. Finally, the implemented model provides interesting results describing the arc behaviour inside the plasma gun.