Analysis of predictors for modification of alumina inclusions in medium carbon steel

Souza, Raphael Mariano de; Oliveira., Márcia Spelta de; Oliveira, José Roberto de; Junca, Eduardo; Telles, Victor Bridi; Grillo, Felipe Fardin

doi:10.1016/j.jmrt.2021.07.083

Cited by 5 publications

(1 citation statement)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A recent study showed that the use of statistical tools represents an alternative to the analysis of inclusion modification by Ca treatment. De Sousa et al [ 5 ] analyzed a plant database and developed a statistically significant model via multiple linear regression to identify the main variables to be controlled for successful Ca treatment. The model indicated that the oxidation of the steel and the S and Ti contents in the steel were the main influential variables and enabled the design of process strategies leading to the better control of the inclusion modification.…”

Section: Introductionmentioning

confidence: 99%

Analysis of the Steelmaking Process via Data Mining and Pearson Correlation

Carrasco-López,

Herrera-Trejo,

Castro-Román

et al. 2024

Materials

View full text Add to dashboard Cite

The continuous improvement of the steelmaking process is a critical issue for steelmakers. In the production of Ca-treated Al-killed steel, the Ca and S contents are controlled for successful inclusion modification treatment. In this study, a machine learning technique was used to build a decision tree classifier and thus identify the process variables that most influence the desired Ca and S contents at the end of ladle furnace refining. The attribute of the root node of the decision tree was correlated with process variables via the Pearson formalism. Thus, the attribute of the root node corresponded to the sulfur distribution coefficient at the end of the refining process, and its value allowed for the discrimination of satisfactory heats from unsatisfactory heats. The variables with higher correlation with the sulfur distribution coefficient were the content of sulfur in both steel and slag at the end of the refining process, as well as the Si content at that stage of the process. As secondary variables, the Si content and the basicity of the slag at the end of the refining process were correlated with the S content in the steel and slag, respectively, at that stage. The analysis showed that the conditions of steel and slag at the beginning of the refining process and the efficient S removal during the refining process are crucial for reaching desired Ca and S contents.

show abstract

Section: Introductionmentioning

confidence: 99%