This study focuses on the relationship between porosity and leak tightness in cold-sprayed aluminum. Aluminum coatings with 0.2-9% porosity were produced by cold spraying and evaluated via helium leak testing. Multiscale porosity was determined through SEM and TEM analyses and shows good correlation with leak test results. The mechanisms involved in the creation of porosity were investigated as well through finite element analysis of single and multi-particle impacts.

In this work, numerical modeling is used to simulate the effects of laser remelting as a post treatment and as an in-situ component of a hybrid plasma spraying process. Initially, a single-pass 2D model is used to simulate the laser post-treatment process in order to obtain relationships between melting pool depth, relative scanning velocity, and laser power. A 3D finite-element model is then used to study temperature variations during multi-layer deposition of a NiCr alloy by plasma spraying with in-situ laser melting. The effects of phase change are taken into account by defining the enthalpy of the material as a function of temperature. Predicted melting pool depth corresponded well with experimental values.

Effective properties of TBCs may be quantified thanks to different measurement techniques. Image-based analysis represents an alternative method for predicting these effective properties. During the last 10 years, 2D modelling was intensively applied to estimate the thermal conductivity from coating cross-sectional images. However, real coatings present a complex 3D architecture so that the use of 2D computations based on cross-sections has to be validated. In the recent decade, 3D imaging approaches were applied for capturing 3D images of thermal spray coatings with relatively high resolution (up to 1 micrometer). Nevertheless, high resolution brings very large computational systems for which finite-element (FE) methods seem to be unsuitable due to high requirements in terms of computer memory (RAM) capacity. In the present study, a three-dimensional finite-difference-based heat transfer model was developed for analyzing the heat transfer mechanisms through a porous structure by saving RAM usage. An artificial 3D coating image, containing 300×300×300 voxels, was generated from microstructural information measured for a real coating cross-sectional image. In particular, this 3D artificial pore network was generated so that calculations performed on its cross-sections present similar results in comparison with those concerning SEM images of real coating cross-sections. Then, the results computed for the 3D image were compared with those obtained from 2D computations performed on cross-sections of the same 3D image, revealing the differences between 2D and 3D image-based analyses. Finally, the results were then compared with those computed by FE packages (OOF2 and ANSYS).