The large demand of high-performance steels to improve the safety and energetic performances in automotive is driving the increasing need of higher uniformity of mechanical properties of rolled products, both within a single coil and in large batch productions of multiple coils. To achieve these targets, we are witnessing an evolution of rolling towards the introduction of sensors and Artificial Intelligence (AI) algorithms to allow the real-time monitoring of both the strip tensile properties and the evolution of microstructure during the annealing process. The work described in the paper is part of a research work that is aimed at developing an AI framework that could allow the full real-time control of the entire mill rolling plant thanks to (i) innovative sensors applied at different steps of the coil processing, (ii) mixed analytical and numerical algorithms for the data computation and analyses, and (iii) IT infrastructure to collect and elaborate the data. The focus of the paper is on the tension-levelling operation which is one of the first that is performed in the rolling mill and can have a relevant influence on the strip mechanical characteristics. To reach this aim, a predictive model applicable to the tension-levelling process was developed, as this process is used to minimize flatness imperfections and residual stresses by means of plastic deformation. During the process, the material is subjected to cyclic tension-compression stresses, obtained with repeated and alternate bends under superposed tension. Consequently, the desired product standards, in terms of thickness and flatness, are achieved, while the mechanism of deformation changes the final mechanical properties of the material, increasing the yield strength and decreasing the maximum elongation. Finally, the predictive model is used to predict the coil mechanical characteristics, by using the data sampled from the sensors placed inside the tension-levelling machine.

An artificial intelligence approach for the in-line evaluation of steels mechanical properties in rolling

Magro T.
;
Ghiotti A.;Bruschi S.;
2021

Abstract

The large demand of high-performance steels to improve the safety and energetic performances in automotive is driving the increasing need of higher uniformity of mechanical properties of rolled products, both within a single coil and in large batch productions of multiple coils. To achieve these targets, we are witnessing an evolution of rolling towards the introduction of sensors and Artificial Intelligence (AI) algorithms to allow the real-time monitoring of both the strip tensile properties and the evolution of microstructure during the annealing process. The work described in the paper is part of a research work that is aimed at developing an AI framework that could allow the full real-time control of the entire mill rolling plant thanks to (i) innovative sensors applied at different steps of the coil processing, (ii) mixed analytical and numerical algorithms for the data computation and analyses, and (iii) IT infrastructure to collect and elaborate the data. The focus of the paper is on the tension-levelling operation which is one of the first that is performed in the rolling mill and can have a relevant influence on the strip mechanical characteristics. To reach this aim, a predictive model applicable to the tension-levelling process was developed, as this process is used to minimize flatness imperfections and residual stresses by means of plastic deformation. During the process, the material is subjected to cyclic tension-compression stresses, obtained with repeated and alternate bends under superposed tension. Consequently, the desired product standards, in terms of thickness and flatness, are achieved, while the mechanism of deformation changes the final mechanical properties of the material, increasing the yield strength and decreasing the maximum elongation. Finally, the predictive model is used to predict the coil mechanical characteristics, by using the data sampled from the sensors placed inside the tension-levelling machine.
File in questo prodotto:
File Dimensione Formato  
CIRP Design 2021.pdf

accesso aperto

Tipologia: Published (publisher's version)
Licenza: Creative commons
Dimensione 498.71 kB
Formato Adobe PDF
498.71 kB Adobe PDF Visualizza/Apri
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11577/3395115
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 2
  • ???jsp.display-item.citation.isi??? ND
social impact