SOM analysis of IMPOC data in relation to process parameters in strip steel production

Mocci, Claudio; Scuola Superiore Sant'Anna; Italy

Mocci, C.; Scuola Superiore Sant'Anna; Italy
Vannucci, M.; Scuola Superiore Sant'Anna; Italy
Colla, V.; Scuola Superiore Sant'Anna; Italy
van den Berg, F.D.; Tata Steel; Netherlands
Schmidt, R.; Arcelor-Mittal Eisenhüttenstadt; Germany

ID: ECNDT-0226-2018
Download: PDF
Session: Modeling and data processing Electromagnetic Techniques 1
Room: J2
Date: 2018-06-12
Time: 09:40 - 10:00

This paper addresses the problem of understanding whether certain process conditions of the Hot Strip Mill, Cold rolling mill and Hot-dip galvanizing lead to non-uniformities in the materials’ microstructure during the production of steel strip.
The work describes the application of an artificial intelligence technique (Self-Organizing Map) in order to relate the several process conditions with inline measurements by a Non-Destructive Testing system called IMPOC©. The influence on non-uniformity of some of these conditions has been investigated through statistical analysis on several different industrial cases.
SOM results can be exploited in Decision Support Systems to help steelwork operators to avoid, during the production, process conditions considered likely to end in materials’ non-uniformities giving them different feasible solutions.