Hardness of thermal sprayed coatings: Relevance of the scale of measurement
Résumé
The coatings obtained by thermal spraying can present a large variety of geometrical parameters (thickness, roughness…), of microstructures (constituents, nature of phases…), of mechanical properties (hardness, elastic modulus…) and of morphological defects (cracks, pores…) depending on the spraying conditions. In order to determine the mechanical properties of the coating, one of the most relevant techniques is probably the instrumented indentation test. Nowadays this technique is very attractive since it allows the determination of numerous parameters. Moreover, recent developments allow the use of a phenomenological approach and modeling at different scales of measurement, from nano (even ultra-nano) to macro scale, i.e. from few milligrams to several kilograms of loading. However, the information, which can be extracted at the different regimes of loading can be the same or lead to different values of the mechanical properties, which can be complementary or contradictory depending on the nature of the coating and the preparation of the sample. For example, roughness, porosity and cracks present in the coating will affect the mechanical characterization since the indentation data analysis is based on how a rigid indenter penetrates into the material. So, an important question arises: Should the influence of these defects to be taken into account, or neglected, for the mechanical characterization?
The present work proposes different methodologies for determining the hardness of coated materials by considering or not the influence of both the porosity and roughness of the surface. In the first part, results of microindentation experiments performed on the rough surface of alumina coatings are compared to those obtained on a polished cross-section. Although the surface of the cross-section is irregular even after caution polishing, the hardness can be measured. A decrease of about 30% of the hardness number on the cross-section is observed. The second part is related to the microstructured yttria-stabilized zirconia analysis. A methodology based on the indentation size effect analysis is presented to avoid the influence of roughness and the defects, which can be crossed by the indenter during the indentation. This methodology allows the hardness determination of the coating exempt of defects. In the last part, a statistical analysis using nanoindentation data resulting from the continuous stiffness measurement mode applied to nanostructured yttria-stabilized zirconia shows that, even if the hardness number varies to a great extent according to the applied load and the location of the indent, the hardness can be represented by means of a unique hardness number independent of the sense of the hardness variation during the indenter displacement.