Development of intelligent mould for online detection of defects in steel billets

Abstract

This study, which is based on six instrumented mould trials conducted at four Canadian minimills, has focused on the development of strategies for the detection of process upsets and defect deformation in the continuous casting of steel billets, using thermocouples installed in the mould wall. Laps and transverse depressions can be detected using two thermocouples installed in the mould wall at the midface location, with the first one located at least 100 mm below the meniscus. With respect to bleeds, it is preferable to monitor metal level fluctuations and mould wall temperatures, maintaining values within a desired range where bleed formation is minimal. In the case of rhomboidity, the average difference between maximum and minimum temperature for the four walls computed for one or more locations situated 25–50 mm below the meniscus (in the vicinity of the peak hot face temperature), for a time period corresponding to the entire billet length, correlates well with the severity of rhomboidity. Local metal level fluctuations are extremely important in generating defects. While the metal level sensor is adequate for monitoring global changes in metal level, it cannot detect local disturbances in metal level fluctuations. Mould wall thermocouples installed above the meniscus (at midface and offcorner locations in the mould wall) can help detect such local fluctuations. The findings of this work are being implemented in the expert system of an ‘intelligent’ mould developed at The University of British Columbia for online monitoring of quality control of the billet casting process.