Analysis of Shape Evolution Law in Cooling Process of Tandem Hot Rolling

Focusing on the deformation behavior of strip steel during laminar cooling in tandem hot rolling, a high‐precision weighted integrated temperature prediction model is developed using machine learning techniques and validated through finite element analysis. Based on the strain theory of elastic‐plastic mechanics and the phase transformation kinetics of continuous media, the coupled mechanism of temperature, phase transformation, and residual stress under varying initial conditions during laminar cooling was systematically analyzed. The results demonstrate that the proposed weighted integrated prediction model overcomes the limitations of single‐model approaches. The predicted temperature range is 600–690 °C, with an root mean square error of 3.19, mean absolute error of 2.36, and R 2 of 0.985. During the cooling process, the final cooling temperature difference between the middle and edge of the strip changes with the initial cooling temperature difference. Notably, the phase transformation and residual stress variations at the edges are more pronounced than those in the central region. A 50.54% greater phase transition degree has been observed at the edge relative to the central region. The stress compensation of 60 IU flatness can reduce the stress difference between the middle and edge of the strip by 89.55% and improve the product quality of the strip.