A Novel Temperature Dynamic Prediction Model for Erosion Risk Mitigation of Ladle

The ladle's refractory lining, directly exposed to molten steel, is susceptible to corrosion which could cause an abnormal increase in ladle temperature and pose potential safety hazards. To mitigate such risks, a novel temperature dynamic prediction (NTDP) model for ladle based on erosion boundary inversion is proposed in this study. In the NTDP model, the ladle transient heat conduction model is established to obtain real-time ladle shell temperatures for erosion boundary inference. An innovative dynamic procedure and weighting strategy is proposed to address the damping and hysteresis inherent problems in the ladle heat transfer process. Such model can estimate erosion boundary estimation and predict future temperature variations. Extensive numerical tests are conducted to scrutinize the NTDP model's performance. Results indicate that the NTDP model can well balance prediction accuracy and anticipation lead time, and exhibits unwavering precision even with measurement noise. Remarkably, the weighting strategy based on sensitivity coefficients significantly enhances early prediction accuracy. The NTDP model can enable a preliminary 275-min advance warning preceding hazardous temperature surges in the ladle shell and an exact safety diagnosis 200 min in advance. Therefore, this study is conducive to effectively achieve timely prediction and preemptive mitigation of ladle-related safety concerns.