Largest-extreme-value distribution analysis of multiple inclusion types in determining steel cleanliness

There is the need for new inclusion rating methods able to assess the cleanliness of modern steel and to predict the characteristic size of maximum defects in a component. These critical tasks can be extremely difficult, because of the presence of different inclusion types. Therefore, new statistical tools have to be developed for addressing these problems. A new method for the analysis of extreme defects is proposed in this article and is successfully applied to different steels containing multiple inclusions. In particular, the analysis, which allows the assessment of the relative density of the different inclusions, is based on inclusion samplings carried out on different control areas.