Surface defects have a serious detrimental effect on the quality of steel. To address the problems of low efficiency and poor accuracy in the manual inspection process, intelligent detection technology based on machine learning has been gradually applied to the detection of steel surface defects. An improved YOLOv8 steel surface defect detection model called YOLOv8-MGVS is designed to address these challenges. The MLCA mechanism in the C2f module is applied to increase the feature extraction ability in the backbone network. The lightweight GSConv and VovGscsp cross-stage fusion modules are added to the neck network to reduce the loss of semantic information and achieve effective information fusion. The self-attention mechanism is exploited into the detection network to improve the detection ability of small targets. Defect detection experiments were carried out on the NEU-DET dataset. Compared with YOLOv8n from experimental results, the average accuracy, recall rate, and frames per second of the improved model were improved by 5.2%, 10.5%, and 6.4%, respectively, while the number of parameters and computational costs were reduced by 5.8% and 14.8%, respectively. Furthermore, the defect detection generalization experiments on the GC-10 dataset and SDD DET dataset confirmed that the YOLOv8-MGVS model has higher detection accuracy, better lightweight, and speed.

To optimize the submerged entry nozzle (SEN) for an ultra-thick slab mold, a mathematical model has been established. The molten steel flow and solidification, inclusion transports, and meniscus fluctuation have been investigated through the model. Compared with the concave-bottom SEN cases, the convex-bottom SEN decreases the imping depth of the jet flow and increases the horizontal velocity and temperature on the meniscus. However, the remelting of the solidified shell is dramatic for the convex-bottom case. The well depth of the concave-bottom SEN and the SEN’s submerged depth have little influence on molten steel flow and solidification. The effects of SEN port shape and port angle on the molten steel flow are significant. As the port shape changes from rectangle to square or the port downward angle decreases, the imping depth of jet flow decreases, the horizontal velocity and the temperature on the mold free surface increase. For the ultra-thick mold, a square-shaped-port SEN with a −10° downward angle is more beneficial by comprehensive consideration of molten steel flow and solidification, inclusion removal, and mold powder melting. The optimized SEN has been applied to the actual caster and its performance has been assessed, indicating that the SEN optimization is efficient.

The growth rate of the shell thickness in the funnel mold for ultrahigh-speed continuous casting of steel has a significant effect on the shell surface quality. By using the node temperature inheritance algorithm, a three-dimensional transient thermal conductivity model of the liquid steel solidification inside the mold was established, and the effects of casting temperature and speed on the heat transfer behavior of the thin slab were investigated. The results show that with an increase in the casting speed from 4.0 to 6.0 m·min ⁻¹ , the maximum thickness of the shell at mold exit was reduced from 26 to 12.8 mm, and the maximum temperature of the slab surface at mold exit increased from 1,210 to 1,305°C. However, the increase of casting temperature from 1,550 to 1,560°C had little effect on the surface temperature and thickness of the slab shell.

The microstructure, precipitates, and austenite grain in high-strength low-alloy steel were characterized by optical microscope, transmission electron microscope, and laser scanning confocal microscopy to investigate the effect of Mo on the toughness of steel. The microstructure was refined and the toughness was enhanced after the addition of 0.07% Mo in steel. The addition of Mo can suppress the Widmanstätten ferrite (WF) formation and promote the transformation of acicular ferrite (AF), leading to the fine transformed products in the heat-affected zone (HAZ). The chemical composition of precipitates changed from Nb(C, N) to (Nb, Mo)(C, N) because of the addition of Mo. The calculated lattice misfit between Nb(C, N) and ferrite was approximately 11.39%, while it was reduced to 5.40% for (Nb, Mo)(C, N), which significantly affected the size and number density of precipitates. A detailed analysis of the precipitates focusing on the chemical composition, size, and number density has been undertaken to understand the contribution of Mo on the improvement of steel toughness.