Correlation between longitudinal crack occurrence and mold heat transfer in continuous casting based on process data mining

The correlation between the longitudinal crack occurrence and integrated heat transfer of the mold with data mining methods was investigated. Firstly, three kinds of support vector machine models based on principal component analysis with different input features were established to explore the effe...

Full description

Saved in:
Bibliographic Details
Published inJournal of iron and steel research, international Vol. 31; no. 5; pp. 1139 - 1152
Main Authors Peng, Zhi-qiang, Hou, Zi-bing, Yi, Kai, Guo, Kun-hui, Guo, Dong-wei, Wen, Guang-hua
Format Journal Article
LanguageEnglish
Published Singapore Springer Nature Singapore 01.05.2024
Subjects
Applied and Technical Physics
Engineering
Machines
Manufacturing
Materials Engineering
Materials Science
Metallic Materials
Original Paper
Physical Chemistry
Processes
Continuous casting slab
Data mining
Longitudinal surface crack
Quantitative correlation
Heat transfer
Online AccessGet full text

Cover

More Information
Summary:The correlation between the longitudinal crack occurrence and integrated heat transfer of the mold with data mining methods was investigated. Firstly, three kinds of support vector machine models based on principal component analysis with different input features were established to explore the effect of integrated heat transfer on the accuracy of the prediction model for the longitudinal crack. The results show that the accuracy was improved while features including mean and standard deviation of integrated heat transfer were added. Then, the difference in integrated heat transfer between defect and normal samples under the same process parameters was quantitatively compared. Compared with normal samples, the temperature difference of cooling water for defect samples decreased by 0.65%, and the temperature difference fluctuation increased by 31.1%. Finally, the literature data were used to provide support for the quantitative correlation according to defect formation mechanism. A new criterion for the prediction of longitudinal crack and a discovering method for correlation between product quality and process parameters in the manufacturing industry have been provided.
ISSN:1006-706X
2210-3988
DOI:10.1007/s42243-023-01109-x