An automatic optical image analyzer, the Quantimet 800 Steelscan, has been applied for microinclusion assessment in bearing steels according to the Stahl Eisen Prufblatt (SEP) 1570/71 method. A brief description of the system is given, including the use of electronic image amendment to distinguish reliably between different types of oxides.
Although the SEP comparison chart is well-formulated and should be sensitive enough for inclusion assessment of present-day bearing steels, results are presented showing that manual assessments with this method should suffer from very poor reproducibility. The automatic method has provided a more reliable basis for discriminating different degrees of inclusion content and, as a consequence, has highlighted the difficulty in obtaining a representative measurement of microinclusion content, especially in worked material such as bar or tube. The effects of deformation by rolling, sampling errors due to material inhomogeneity, and small variations in the orientation of the plane of sectioning on the microinclusion K ratings are discussed.
The limitations of comparison chart methods and automatic image analysis systems are reviewed, with emphasis on the required sensitivity for assessing bearing steels, as compared with an accurate chemical analysis of impurity elements for the prediction of microinclusion content.