Hot Strip Rolling Research Articles

Deep learning for predictive mechanical properties of hot-rolled strip in complex manufacturing systems

Deep learning for predictive mechanical properties of hot-rolled strip in complex manufacturing systems

Analytical model of hot-rolled strip residual stress and flatness in run-out table cooling

The residual stress formed in the run-out table cooling stage of hot-rolled strip has a great influence on flatness and subsequent cutting. Based on the fiber rods model, the formation mechanism of residual stress induced by incompatible deformation caused by temperature and phase transformation is studied in this paper. Meanwhile, the present study makes a first attempt to establishing the analytical model for calculating residual stresses at different run-out table cooling stages. Moreover, in order to further investigate the relationship between residual stress and flatness defects, a strip buckling model is introduced. The traditional buckling equilibrium equation of Lagrange system is introduced into the Hamiltonian system of symplectic space, and the mathematical relationship between residual stress and strip flatness is obtained directly without assuming stress and deflection function. Then the strip flatness is calculated by using this mathematical relationship and residual stress analytical model. Furthermore, to verify the accuracy of the analytical model, a multi-field coupled finite element model based on the validation of the measured data is established. Finally, the distribution of residual stress calculated by the analytical model and the finite element model is basically consistent. The analytical model calculated that the strip would have edge wave defects during run-out table cooling, which is highly consistent with the results of finite element model.

Strip Surface Defect Detection Algorithm Based on YOLOv5.

In order to improve the detection accuracy of the surface defect detection of industrial hot rolled strip steel, the advanced technology of deep learning is applied to the surface defect detection of strip steel. In this paper, we propose a framework for strip surface defect detection based on a convolutional neural network (CNN). In particular, we propose a novel multi-scale feature fusion module (ATPF) for integrating multi-scale features and adaptively assigning weights to each feature. This module can extract semantic information at different scales more fully. At the same time, based on this module, we build a deep learning network, CG-Net, that is suitable for strip surface defect detection. The test results showed that it achieved an average accuracy of 75.9 percent (mAP50) in 6.5 giga floating-point operation (GFLOPs) and 105 frames per second (FPS). The detection accuracy improved by 6.3% over the baseline YOLOv5s. Compared with YOLOv5s, the reference quantity and calculation amount were reduced by 67% and 59.5%, respectively. At the same time, we also verify that our model exhibits good generalization performance on the NEU-CLS dataset.

Steel Strip Defect Sample Generation Method Based on Fusible Feature GAN Model under Few Samples.

Due to the shortage of defect samples and the high cost of labelling during the process of hot-rolled strip production in the metallurgical industry, it is difficult to obtain a large quantity of defect data with diversity, which seriously affects the identification accuracy of different types of defects on the steel surface. To address the problem of insufficient defect sample data in the task of strip steel defect identification and classification, this paper proposes the Strip Steel Surface Defect-ConSinGAN (SDE-ConSinGAN) model for strip steel defect identification which is based on a single-image model trained by the generative adversarial network (GAN) and which builds a framework of image-feature cutting and splicing. The model aims to reduce training time by dynamically adjusting the number of iterations for different training stages. The detailed defect features of training samples are highlighted by introducing a new size-adjustment function and increasing the channel attention mechanism. In addition, real image features will be cut and synthesized to obtain new images with multiple defect features for training. The emergence of new images is able to richen generated samples. Eventually, the generated simulated samples can be directly used in deep-learning-based automatic classification of surface defects in cold-rolled thin strips. The experimental results show that, when SDE-ConSinGAN is used to enrich the image dataset, the generated defect images have higher quality and more diversity than the current methods do.

Refinement mechanism and optimization for compact strip production process

To avoid the fracture of cold-rolled sheets, this study investigated the refinement mechanism for the compact strip production process and proposed an optimised method. The primary austenite grain size was 50–475 µm, and the grain size of the hot-rolled strips was 3.6–4.8 µm. The Gleeble-3500 simulator was applied to simulate the hot rolling process. Grain refinement was realised through dynamic recrystallization and the supply of more nucleation sites for transformation by sub-structures. Increasing the coiling temperature from 620 to 730 °C increased the ferrite fraction from 3.95 to 24.16%. The adjacent cementite distance changed from 0.33 to 0.84 µm, resulting in decreased yield and tensile strengths and increased elongation. Thus, the fracture of cold-rolled sheets was avoided.

Homogeneous cooling control technology for high-strength low-alloy hot-rolled strip

ABSTRACT Heterogeneities of transverse temperature during laminar cooling of high-strength low-alloy hot-rolled strip can cause heterogeneous distribution of microstructure, mechanical property and internal stress. Edge masking is a control technology that can be used to improve laminar cooling heterogeneities. In this paper, different edge masking schemes were set up based on the transverse transformation characteristics of the strip. The results show that a slow cooling of 7.1 s during transformation could reduce the heterogeneity of grain size, internal stress and microhardness at the edge and middle of the strip. In addition, the relationship model between grain size and six variables such as transformation start temperature was established by PCA. It provided a new idea for the construction of a microstructure prediction model under nonlinear cooling.

A joint model for hot-rolled strip quality in advance prediction

ABSTRACT Aiming at the problem that previous product quality prediction models cannot directly assess product quality before hot rolling, this paper presents a joint model for process parameters generation and quality prediction based on Conditional Generative Adversarial Nets and Artificial Neural Network. After generated actual process parameters by the generation module, this model would predict the quality of products ahead of schedule according to the generated parameters before hot rolling, which do not rely on inputting actual process parameter online anymore. Finally, the model has been trained and tested with the actual data from a certain hot rolling plant. The experimental results show that the generated process parameters agree well with actual production and quality prediction accuracy can meet the production requirements, which confirms the proposed model can be applied to simulate the actual rolling process and predict strip quality ahead of hot rolling production, providing a reference for the adjustment of planning and scheduling in the future.

Kinetics and Microstructure Transformations of Oxide Scale on Hot Rolled Steel Strip for Hot‐Dip Galvanizing during Reduction Annealing in 30% H2–N2 Atmosphere

The understanding of oxide scale reduction mechanisms after microstructure transformation under different preheating conditions is incomplete. Herein, the reduction kinetics and microstructure transformations of hot rolled steel strip oxide scale during reduction (30% H2–N2) annealing are studied in detail using thermogravimetric and microstructure analyses, and mathematical modeling based on the gas–solid reaction. The results show that the rate‐determining steps of the reduction reaction are nucleation and growth of new phase between 500 and 600 °C, and gas–solid interface reactions between 700 and 800 °C. Microstructure transformation of the oxide scale occurs during the preheating process: at 500 °C, the eutectoid iron preferentially dissolves, and short‐range diffusion of Fe ions into adjacent Fe3O4 increases their content in the metal oxide; at 600 °C, nucleation and growth of Fe1 − yO occur, and some Fe3O4 precipitates and white α‐Fe are found in the Fe1 − yO layer, whereas the majority of the Fe3O4/iron eutectoid structure is retained; between 700 and 800 °C, the oxide scale forms outer and inner layers of Fe3O4 and Fe1 − yO, respectively. After preheating and reduction, the reduced products are porous iron over 500–600 °C, while porous and dense iron occur at 700 °C; dense iron forms as whiskers or granules between 700 and 800 °C. The surface quality of hot rolled steel strip after reduction annealing is subsequently improved as surface cracks are infilled by the newly formed dense iron.

Prediction and Online Optimization of Strip Shape in Hot Strip Rolling Process Using Sparrow Search Algorithm‐Online Sequential‐Deep Multilayer Extreme Learning Machine Algorithm

In the hot strip process, the prediction of strip shape is a key factor to improve the quality of products. However, strip‐shape prediction models based on traditional machine learning algorithms with relatively simple network structures often rely on data preprocessing and feature selection. Herein, five strip‐shape prediction models are proposed by combining sparrow search algorithm (SSA), particle swarm optimization algorithm (PSO), extreme learning machine (ELM), and deep multilayer extreme learning machine (DELM). The prediction experiments show that the SSA‐DELM model has highest accuracy, and the determination coefficient (R2) of crown prediction and flatness prediction of SSA‐DELM model is 0.971 and 0.979. 100% of the samples have a prediction error of less than ±7 μm and ±7 IU. Furthermore, to solve the problem of concept drift affecting prediction accuracy in industrial processes, online optimization method is proposed. Especially, online sequential deep ELM model (OS‐DELM) optimized by SSA and guided by double window drift detection is proposed to fill the gap. Compared with other online optimization methods, the update time with this method can be reduced by up to 83.05% averagely, which is more suitable for online strip‐shape prediction in hot rolling process.

Influence of the Variation of Mechanical Properties and Thickness in Hot-Rolled Strips of Electrical Anisotropic Steel to Stabilize Cold Rolling

Influence of the Variation of Mechanical Properties and Thickness in Hot-Rolled Strips of Electrical Anisotropic Steel to Stabilize Cold Rolling

Time-Dependent Evolution of Volume Fraction and Stability of Retained Austenite in a Hot-Rolled and Intercritically Annealed Al-Alloyed Medium-Mn Steel

The development of superior mechanical properties in medium-Mn requires the optimization of microstructural parameters such as retained austenite (RA) stability, volume fraction, and morphology. The present work explores the possibility of using a continuous annealing approach instead of conventional batch annealing to perform an intercritical annealing (IA) treatment in a hot-rolled strip of an Al-alloyed 5Mn steel. Dilatometric studies were performed at a temperature of 680 ºC with soaking times ranging from 1 to 300 min to follow the microstructural changes as a function of time. The microstructures thus obtained were thoroughly characterized by means of X-ray diffraction, SEM and TEM, TEM-EDS microanalysis and EBSD phase and orientation maps. It was observed that with increasing soaking times, the volume fraction of retained austenite gradually increases, albeit at the cost of its stability. The comparison of martensite start temperatures (Ms) based on the chemical composition of austenite at 680 ºC with that experimentally obtained at higher process temperature revealed the effect of the grain size on the reduction of RA stability for longer process times. Accordingly, mechanical tests results showed that the yield stress, tensile strength and hardness decrease with an increase in the IA soaking time.

Prediction model of hot strip crown based on industrial data and hybrid the PCA-SDWPSO-ELM approach

The accurate prediction of strip crown is the precondition of the shape preset model in hot strip rolling. In this study, a new hybrid strip crown forecasting model is proposed in combination with extreme learning machine (ELM) and the industrial data. The production data of 1780 mm hot strip rolling are collected by the on-site data acquisition system to form dataset. Principal component analysis (PCA) is used to reduce the dimension of the input data for modeling samples. To improve the prediction accuracy of ELM, an improved PSO based on S-curve decreasing inertia weight (SDWPSO) is proposed to optimize the initial weights and biases of ELM. Finally, the optimal ELM model and simple production dataset are used to establish a strip crown prediction model of hot strip rolling named PCA-SDWPSO-ELM. The comprehensive performance of the proposed hybrid PCA-SDWPSO-ELM prediction model is evaluated by MAE, MAPE and RMSE. The superiority of the proposed model is also proved by comparing the prediction results with those of the other three comparison models. The research shows that the hybrid PCA-SDWPSO-ELM method can solve the problem of nonlinear and strong coupling in traditional engineering. It is suitable for parameter prediction and optimization in the iron and steel manufacturing industry, especially in the process of shape control in hot strip rolling.

Residual stress control of hot-rolled strips during run-out table cooling

The strips with good flatness after finish rolling will have flatness defects such as edge wave after run-out table cooling. These defects are caused by the residual stress due to inhomogeneous thermal and microstructural behaviors during the cooling stage. The traditional residual stress control methods rely too much on experience, resulting in poor control effect and poor universality. Therefore, this study proposed two residual stress control methods based on model calculation: water flow central crown and center wave compensation control. The former is mainly based on controlling the uniformity of strip transverse temperature and microstructure. The latter is mainly controlled by mutual compensation of internal stresses. In this paper, the mathematical models of water flow central crown cooling and critical center wave compensation stress were established, respectively. Meanwhile, a post-buckling model of hot-rolled strip was established to calculate the aim flatness value at exit of finishing mill under critical compensation stress. Furthermore, a fully coupled finite element model based on temperature, phase transitions, and stress was developed and it was verified using experimental data. Finally, the above two residual stress control methods were realized in the finite element model. The results showed that the two methods can reduce the residual stress and improve the flatness, and the application of center wave compensation control is more effective.

Effect of strip profile of hot-rolled silicon steel on transverse thickness difference of cold-rolled strip

ABSTRACT The cross-sectional profile of hot-rolled silicon steel strips has a great impact on the transverse thickness difference of cold-rolled strips. The genetic algorithm - Levenberg Marquardt - backpropagation neural network model was developed for the analysis of heredity effect of the cross-sectional profile of hot-rolled strip. The feature importance was analysed based on random forest and extreme gradient boosting methods. Furthermore, the influence law was analysed by data visualization. The coupling term of the strip crown and thickness , the edge drop and the coupling term of the strip wedge and thickness are recommended as key indexes of the hot-rolled silicon steel strip. As the transverse thickness difference must be controlled below 7 μm, the corresponding , , should be controlled no more than 60, between -10∼10 μm and no more than 7.5 , respectively. Meanwhile, the requirement for the cross-sectional profile indexes is synchronous.

Applications and Developments of Thermal Spray Coatings for the Iron and Steel Industry.

The steel making processes involves extreme and harsh operating conditions; hence, the production hardware is exposed to degradation mechanisms under high temperature oxidation, erosion, wear, impact, and corrosive environments. These adverse factors affect the product quality and efficiency of the steel making industry, which contributes to production downtime and maintenance costs. Thermal spray technologies that circumvent surface degradation mechanisms are also attractive for their environmental safety, effectiveness and ease of use. The need of thermal spray coatings and advancement in terms of materials and spray processes are reviewed in this article. Application and development of thermal spray coatings for steel making hardware from the molten metal processing stages such as electric arc and basic oxygen furnaces, through to continuous casting, annealing, and the galvanizing line; to the final shaping process such as cold and hot rolling of the steel strips are highlighted. Specifically, thermal spray feedstock materials and processes that have potential to replace hazardous hard chrome plating are discussed. It is projected that novel coating solutions will be incorporated as awareness and acceptance of thermal spray technology grows in the steel making sectors, which will improve the productivity of the industry.

Application of a high-speed infrared thermographic camera to the study of HAZ softening in S960 welded joints

Hot strip rolling of low-carbon advanced high-strength (AHS) sheet steels is difficult because their products require specific chemical compositions to achieve excellent mechanical properties. The formation of a heat-affected zone (HAZ) during welding is therefore a challenge in terms of maintaining high strength values. This fact complicates the formation of the softening area when the hardness in the HAZ rapidly decreases. According to the continuous cooling transformation diagram (CCT), the chemical composition of the base metal is directly related to the fine-grained martensitic microstructure arising from very high cooling rates and microalloying up to 0.2 wt% by Nb, V, and Ti. The gas metal arc welding (GMAW) was used to perform the study, what presents a challenge for thin sheets because of very narrow HAZ formation, suscepible to HAZ softening phenomenon. The width of the HAZ was about 6 mm and determined by applying an infrared thermographic camera to measure the real-time temperature development. The significant reduction in microhardness, known as the softening phenomenon, occurred in the intercritical heat-affected zone or at the boundary between intercritical and subcritical HAZ, according to real-time diagnostics of the temperature development and the microhardness results. The amount of heat produced during welding plays an important role in this context and is directly connected to the heat input of the welding process. The softening area can be reduced and its impact on the mechanical properties can be lowered by appropriately modifying the welding conditions, e.g., optimising the heat input. The purpose of this paper is to point out the change in temperature during the welding process.

Rapid Visualization of Microstructural Evolution during Hot Strip Rolling Using Phenomenological Models and the Laguerre–Voronoi Diagram

Microstructural evolution during hot rolling, which is complex and in a state of the black box, has a significant impact on the microstructure and mechanical properties of steel products. Herein, a rapid methodology is developed to track and visualize the evolution of the microstructural state, and a novel approach coupled with phenomenological recrystallization kinetics is proposed to calculate grain size and its distribution by considering the effects of the size, shape, and mixed fraction of nonrecrystallized grains, and by using the Laguerre–Voronoi diagram, the microstructural morphologies are reconstructed. Based on the proposed approach, the variations of the recrystallization fraction, grain size, and its distribution with time during rolling are analyzed, and the microstructural features in various recrystallization stages are discussed. Additionally, this strategy is demonstrated to be time efficient.

Simulation of mechanical wear of work rolls of a wide-strip hot rolling mill using machine learning methods

Modern methods of modeling a mechanical wear of hot rolling mill work rolls under industrial conditions are presented. The use of machine learning algorithms based on multidimensional linear regression, decision trees, and neural networks made it possible to describe more accurately the features of work roll wear depending on many technological factors. At the same time, the data set from the production database was subjected to additional classification for each technological parameter in a certain range. This classification approach significantly improved the quality of the output model for all major metrics. The input variables for training were the main technological factors measured in the finishing group of stands during the campaign of each work roll. The output parameters – wear curves on a roll grinder measured after roll changes. 93 input parameters were used to calculate each of the 300 output values of mechanical wear along the length of the roll barrel and to train machine learning algorithms. Roll wear forecasting models based on multivariate linear regression, random forest, gradient boosting and neural networks have been developed. A comparative analysis of the developed models with classical approaches for calculating the work roll mechanical wear showed that machine learning methods have a higher accuracy. An approach and recommendations for the introduction of new models to the automated process control system of hot strip rolling mills were proposed.The research was carried out within the framework of the program of strategic academic leadership of the Russian Federation "Priority 2030", aimed at supporting the development programs of educational institutions of higher education, the scientific project PRIOR/SN/NU/22/SP5/26 "Creation of innovative digital tools for the use of applied artificial intelligence and advanced statistical analysis of big data in technological processes of manufacture of metallurgical products", as well as within the framework of scientific and technical cooperation between JSC Vyksa Metallurgical Plant and Bauman MSTU.

Interval prediction of bending force in the hot strip rolling process based on neural network and whale optimization algorithm

In the hot strip rolling process, accurate prediction of bending force is beneficial to improve the accuracy of strip crown and flatness, and further improve the strip shape quality. Due to outliers and noise are commonly present in the data generated in the rolling process, not only the prediction accuracy should be considered, but also the uncertainty of prediction results should be described quantitatively. Therefore, for the first time, the authors establish an interval prediction model for bending force in hot strip rolling process. In this paper, we use Artificial Neural Network (ANN) and whale optimization algorithm (WOA) to produce a prediction interval model (WOA-ANN) for bending force in hot strip rolling. Based on the point prediction by ANN, interval prediction is completed by using lower upper bound estimation (LUBE) and WOA, and three indexes are used to evaluate the performance of the model. This paper uses real world data from steel factory to determine the optimal network structure and parameters of the interval prediction model. Furthermore, the proposed WOA-ANN model is compared with other interval prediction models established by other three optimization algorithms. The experimental results show that the proposed WOA-ANN model has high reliability and narrow interval width, and can well complete the interval prediction of bending force in hot strip rolling. This study provides a more detailed and rigorous basis for setting bending force in hot strip rolling process.

A Novel Named Entity Recognition Algorithm for Hot Strip Rolling Based on BERT-Imseq2seq-CRF Model

Named entity recognition is not only the first step of text information extraction, but also the key process of constructing domain knowledge graphs. In view of the large amount of text data, complex process flow and urgent application needs in the hot strip rolling process, a novel named entity recognition algorithm based on BERT-Imseq2seq-CRF model is proposed in this paper. Firstly, the algorithm uses the BERT preprocessing language model to mine the dependencies in the domain text and obtain the corresponding representation vector. Then, the representation vector is sent to the encoder layer, and the output vector is input to the decoder at the same time, on the premise that the original model only considers the semantic vector. The Teacher-Forcing mechanism is integrated into the decoder layer to randomly modify the labeling results, and error accumulation is avoided to guarantee the sequence recognition effect. Finally, the validity of the labeling results is checked according to the conditional random field constraints, and the overall labeling quality of the algorithm is improved. The experimental results show that this model can efficiently and accurately predict the physical label of hot strip rolling, and the model performance index is better than other models, with the F1-Score reaching 91.47%. This model further provides technical support for information extraction and domain knowledge graph construction of hot strip rolling.