Trends and challenges in the morphological study of ceramics
Abstract
Ceramic materials are always fabricated from powders or mix of powders. Their solid state isobtained most often by sintering. Ceramic materials can be classified in two categories:monophased or polyphased ceramics. Cermets (ceramic-metal) are generally obtained by liquidphase sintering; they are made of a metal matrix where a ceramic phase is dispersed.
The study of the morphology of these materials involves, first, the use of a metallographictechnique to reveal the microstructure by an appropriated etching.
The next step consists of isolating phases or grains of material by segmentation and thresholdmethods. Each method has to be adapted to the analysed material.
Then, one proceeds to the analysis itself. Classically, accessible stereological parameters in theR2 space are measured. They allow to estimate corresponding parameters in R3. A finerparametric analysis requires a characterization of the shape, the anisotropy, the state ofdispersion and mainly the size, using various methods (granulometry, granulomorphy, covariance, ...). For these analyses, mathematical morphology will appear as one of the mostefficient tool for morphological purpose.
Finally one will show that the use of probabilistic models is an interesting way to characterize,by a simple but complete manner, the microstructure of ceramics. Their conditions ofapplication to real microstructures and corresponding tests will be specified.