Just another WordPress site - Ruhr-Universität Bochum

A new method for the quantitative image analysis of complex microstructures in steel is presented. Microstructures in modern steel grades on the one hand become more and more complex while on the other hand the size of the constituents is scaled down ever further. For these reasons it is increasingly difficult or even impossible to access reliable quantitative information on ratios, shapes and spatial distributions of individual phases and constituents by utilising traditional metallographic analysis methods. The presented new analysis approach is object-based. Its basic processing units are so-called "image objects" which reflect the pictured structures by merged groups of likely related pixels. By the use of image objects, information on shape graphical texture and spatial distribution of the microstructural constituents can be processed. With these additional data available, it is possible to quantitatively analyse complex microstructures. This is demonstrated by the application of the method to SEM images of different multiphase steel grades. Because the analysis routine is fully automated, the analysis of large amounts of image data is possible and thereby the investigation of statistically relevant areas of microstructures.