PR-DT0521-1 - 2021 Digital Transformations, May

Fully Automated Rating of Slab Segregation Images for Pipeline Steel on a Continuous Scale

A. van der Breggen, K. Dunnett, L. Collins

Download PDF

View Sample Page

Abstract:
For high-strength, low-alloy (HSLA) steel produced for pipelines, the centerline segregation of continuously cast slabs is an important metric of quality. In the current market, measurement of centerline segregation is done by comparison to reference images, by measurement of the fraction of surface area darkened by etching, or by measuring and counting of darkened dots on etched surfaces. To remove sensitivity in these adjustments while balancing precision and understandability, this paper proposes a fully automated image analysis method based on industrially recommended dot measuring methods. A software prototype using this method has been developed; since this method uses images of samples etched for other methods, the method has been tested on thousands of samples.


Keywords: digital transformation, industry 4.0