Hot Rolled Steel Strip Research Articles

Micro-Segregation induced strain inhomogeneity in >900 MPa UTS martensitic hot rolled advanced high strength steel

High strength martensitic strip steels used in the yellow goods industry require good ductility and high strength. This paper examines how local compositional inhomogeneity of alloying elements (Mn, Mo, and Cr), caused by interdendritic microsegregation during casting and subsequent deformation, affect local straining behaviour in a >900 MPa autotempered martensite hot rolled strip steel. The alloy rich and alloy depleted bands have been characterised using micro-XRF and SEM-EDX through thickness for a 10 mm thick strip. Tensile samples were selectively machined from areas with a range of microsegregation severity in the longitudinal and transverse orientations. Digital Image Correlation (DIC) during tensile testing was used to examine local straining behaviour and shear band formation. Differential straining of up to 2.5x between solute enriched (lower strain) and solute depleted (high strain) bands was found in the strongly microsegregated transverse samples due to the lower hardness of the solute depleted bands (reduced solid solution strengthening). The high local strain in solute depleted bands in the strongly segregated sample led to premature failure (at lower global applied strain) compared to the less segregated samples. No significant strain partitioning was observed for the longitudinal oriented samples regardless of segregation level.

Multi-Scale Lightweight Neural Network for Steel Surface Defect Detection

Defect classification is an important aspect of steel surface defect detection. Traditional approaches for steel surface defect classification employ convolutional neural networks (CNNs) to improve accuracy, typically by increasing network depth and parameter count. However, this approach overlooks the significant memory overhead of large models, and the incremental gains in accuracy diminish as the number of parameters increases. To address these issues, a multi-scale lightweight neural network model (MM) is proposed. The MM model, with a fusion encoding module as its core, constructs a multi-scale neural network by utilizing the Gaussian difference pyramid. This approach enhances the network’s ability to capture patterns at different resolutions while achieving superior model accuracy and efficiency. Experimental results on a dataset from a hot-rolled strip steel plant demonstrate that the MM network achieves a classification accuracy of 98.06% in defect classification tasks. Compared to networks such as ResNet-50, ResNet-101, VGG, AlexNet, MobileNetV2, and MobileNetV3, the MM model not only reduces the number of model parameters and compresses model size but also achieves better classification accuracy.

A Novel Anomaly Detection Method for Strip Steel Based on Multi-Scale Knowledge Distillation and Feature Information Banks Network

To address the problem of image imbalance in the surface inspection of strip steel, this study proposes a novel anomaly detection method based on multi-scale knowledge distillation (Ms-KD) and a block domain core information module (BDCI) to quickly screen abnormal images. This method utilizes the multi-scale knowledge distillation technique to enable the student network to learn the ability to extract normal image information under the source network pre-trained on ImageNet. At the same time, the optimal storage of block-level features is used to extract low-level and high-level information from intermediate layers and establish a feature bank, which is searched for core subset libraries using a greedy nearest neighbor selection mechanism. By using the Ms-KD module, the student model can understand the abnormal data more comprehensively so as to better capture the information in the data to solve the imbalance of abnormal data. To verify the validity of the proposed method, a completely new dataset called strip steel anomaly detection for few-shot learning (SSAD-FSL) was constructed, which involved image-level and pixel-level annotations of surface defects on cold-rolled and hot-rolled strip steel. By comparing with other state-of-the-art methods, the proposed method performs well on image-level area under the receiver operating characteristic curve (AUROC), reaching a high level of 0.9868, and for pixel-level per region overlap (PRO) indicators, the method also achieves the best score of 0.9896. Through a large number of experiments, the effectiveness of our proposed method in strip steel defect anomaly detection is fully proven.

Improved YOLOv5l-based Detection of Surface Defects in Hot Rolled Steel Strips

To address the problems of complex background, different sizes and easy to miss and mis-detect in the detection of surface defects in hot-rolled strip, an improved YOLOv5l-based method for detecting surface defects in hot-rolled strip is proposed. Firstly, by adding the SimAM attention mechanism module to the aggregation network, the important information is focused with high weights to improve the recall rate of the original algorithm; secondly, by replacing all C3 modules in the YOLOv5l structure with C2F, a richer gradient of information flow is obtained to improve the accuracy rate of the original algorithm. The experimental results show that the average detection accuracy using the improved YOLOv5l improves by 5.3% and the accuracy rate by 8.3% compared to the original network, resulting in higher detection accuracy and lower error and miss detection rates, meeting the requirements of hot-rolled strip steel inspection in industrial manufacturing.

Strain evolution during hole expansion testing of 800 MPa tensile strength hot-rolled steels

One limitation of the standardized ISO 16630 hole expansion test is that it provides only one result: limiting hole expansion ratio (HER). In practice, steels with similar HER values can have different cut edge forming behavior due to possible differences in strain localization tendencies. Digital image correlation (DIC) strain measurement during formability testing allows more in-detail analysis of strain-state near the cut edge. In this paper, strain evolution during hole expansion testing was investigated for three 800 MPa tensile strength grade hot-rolled strip steels. The steels were selected to have differences in microstructures and anisotropies of mechanical properties. Two different hole expansion test methods with DIC strain measurement were utilized to investigate different edge loading scenarios: in-plane stretching with a flat-top punch and more out-of-plane stretching with a hemispherical punch. Test holes were prepared according to ISO 16630 standard. In order to examine strain evolution and localization during testing, strains were measured with circle-shape sections around the hole in various distances from the cut edge. Strain localization behavior was investigated in different sheet directions and the effect on the hole expansion ratio was evaluated. Results show considerable differences in the cut edge forming behavior between the investigated materials.

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.

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.

Cost-Sensitive YOLOv5 for Detecting Surface Defects of Industrial Products.

Owing to the remarkable development of deep learning algorithms, defect detection techniques based on deep neural networks have been extensively applied in industrial production. Most existing surface defect detection models assign equal costs to the classification errors among different defect categories but do not strictly distinguish them. However, various errors can generate a great discrepancy in decision risk or classification costs and then produce a cost-sensitive issue that is crucial to the manufacturing process. To address this engineering challenge, we propose a novel supervised classification cost-sensitive learning method (SCCS) and apply it to improve YOLOv5 as CS-YOLOv5, where the classification loss function of object detection was reconstructed according to a new cost-sensitive learning criterion explained by a label-cost vector selection method. In this way, the classification risk information from a cost matrix is directly introduced into the detection model and fully exploited in training. As a result, the developed approach can make low-risk classification decisions for defect detection. It is applicable for direct cost-sensitive learning based on a cost matrix to implement detection tasks. Using two datasets of a painting surface and a hot-rolled steel strip surface, our CS-YOLOv5 model outperforms the original version with respect to cost under different positive classes, coefficients, and weight ratios, but also maintains effective detection performance measured by mAP and F1 scores.

Production of a Non-Stoichiometric Nb-Ti HSLA Steel by Thermomechanical Processing on a Steckel Mill

Obtaining high levels of mechanical properties in steels is directly linked to the use of special mechanical forming processes and the addition of alloying elements during their manufacture. This work presents a study of a hot-rolled steel strip produced to achieve a yield strength above 600 MPa, using a niobium microalloyed HSLA steel with non-stoichiometric titanium (titanium/nitrogen ratio above 3.42), and rolled on a Steckel mill. A major challenge imposed by rolling on a Steckel mill is that the process is reversible, resulting in long interpass times, which facilitates recrystallization and grain growth kinetics. Rolling parameters whose aim was to obtain the maximum degree of microstructural refinement were determined by considering microstructural evolution simulations performed in MicroSim-SM® software and studying the alloy through physical simulations to obtain critical temperatures and determine the CCT diagram. Four ranges of coiling temperatures (525–550 °C/550–600 °C/600–650 ° C/650–700 °C) were applied to evaluate their impact on microstructure, precipitation hardening, and mechanical properties, with the results showing a very refined microstructure, with the highest yield strength observed at coiling temperatures of 600–650 °C. This scenario is explained by the maximum precipitation of titanium carbide observed at this temperature, leading to a greater contribution of precipitation hardening provided by the presence of a large volume of small-sized precipitates. This paper shows that the combination of optimized industrial parameters based on metallurgical mechanisms and advanced modeling techniques opens up new possibilities for a robust production of high-strength steels using a Steckel mill. The microstructural base for a stable production of high-strength hot-rolled products relies on a consistent grain size refinement provided mainly by the effect of Nb together with appropriate rolling parameters, and the fine precipitation of TiC during cooling provides the additional increase to reach the requested yield strength values.

Isothermal Structure Transformation and Kinetics Behavior of Oxide Scale on Low Carbon Steel

The phase transformation of the oxide scale that forms on hot-rolled steel strips during cooling is one of the most important phenomena, which determines the surface quality of the strips, because the properties of the oxide scale, such as spallation resistance, crack initiation, and propagation, and pickling behavior, are strongly dependent on the oxide scale microstructure. To obtain steel strips with high surface quality, it is crucial to understand the microstructural development of the oxide scale. In this paper, a thermogravimetric analyzer was used to study the isothermal structure transformation behavior of FeO formed at high temperature in an inert gas (Ar) at 300–550°C for 1000–15000 s. Meanwhile, the isothermal dynamic model of FeO eutectoid transformation was established based on the JMAK equation combined with experimental data. The experimental results show that the oxide scale structure of the sample after pre-oxidation is composed of Fe2O3, Fe3O4, and FeO. At 300°C and 550°C, only the pro-eutectoid structure is formed in the oxide scale, but no eutectoid structure is formed. When the isothermal temperature is in the range of 350–500°C, both pro-eutectoid structure and eutectoid structure are formed in the oxide scale. With the extension of isothermal time, the eutectoid reaction continues until the complete decomposition of FeO. In addition, the FeO eutectoid transformation isothermal dynamics model was used to predict the eutectoid structure content of FeO eutectoid decomposition under different isothermal conditions, and the experimental data were compared with the predicted results, which verified the high prediction accuracy of the model.

Deep ensemble transfer learning-based approach for classifying hot-rolled steel strips surface defects

Over the last few years, advanced deep learning-based computer vision algorithms are revolutionizing the manufacturing field. Thus, several industry-related hard problems can be solved by training these algorithms, including flaw detection in various materials. Therefore, identifying steel surface defects is considered one of the most important tasks in the steel industry. In this paper, we propose a deep learning-based model to classify six of the most common steel strip surface defects using the NEU-CLS dataset. We investigate the effectiveness of two state-of-the-art CNN architectures (MobileNet-V2 and Xception) combined with the transfer learning approach. The proposed approach uses an ensemble of two pre-trained state-of-the-art Convolutional Neural Networks, which are MobileNet-V2 and Xception. To perform a comparative analysis of the proposed architectures, several evaluation metrics are adopted, including loss, accuracy, precision, recall, F1-score, and execution time. The experimental results show that the proposed deep ensemble learning approach provides higher performance achieving an accuracy of 99.72% compared to MobileNet-V2 (98.61%) and Xception (99.17%) while preserving fast execution time and small models’ size.

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.

Rain-Like Layer Removal From Hot-Rolled Steel Strip Based on Attentive Dual Residual Generative Adversarial Network

Rain-like layer removal from hot-rolled steel strip surface is proved to be a workable measure for suppressing the false alarms frequently triggered in automated visual inspection (AVI) instrument. This paper extends the scope of “rain-like layer” from dispersed waterdrops to splashing water streaks and tiny white droplets. And a targeted method with both channel-wise and spatial-wise attention, namely attentive dual residual generative adversarial network (ADRGAN), is proposed. Meanwhile, a newly updated steel surface image dataset with typical natures of “rain-like layer” gathered from actual hot-rolling line, Steel_Rain, is opened for the first time. The comparison experimental results between our proposed network and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">eleven</i> prestigious networks show that our ADRGAN-restored images are the closest to the ground-truth images on <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">six</i> public datasets, especially on the newly-opened industrial dataset Steel_Rain, it yeilds the best scores of 56.8627 peak signal to noise ratio (PSNR), 0.9980 structural similarity index (SSIM), 0.134 mean-square error (MSE) and 0.006 learned perceptual image patch similarity (LPIPS). In the final verification test, the concept of rain-like layer removal has been proved to perform best in defect inspection, where <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">four</i> traditional defect detection algorithms are involved. And as expected, defect detection methods assisted by ADRGAN yield the minimum false-alarms <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup> .

Effect of copper on formation and adhesion of low carbon hot-rolled steel strips

The understanding of scales adhesively bonded requires several adhesion test method. Adhesion behaviour focuses on adhesive between scales-steel interface associated with scales formation and mechanical adhesion. This information on interface is typically obtained from tensile and pickling tests. Such information can give an indication of the adhesive strength of the interfaces. The hot-rolled low-carbon steel containing copper is focused on scales adhesion studies accomplished in laboratory. The key principle for good result is conception of an experimental method. Typical tensile testing machine equipped with a CCD camera has been adapted to investigate the scale failure. Classical gravimetric weight loss method has been conducted to assess corrosion rate. The scales morphology and cross-section of the samples are observed by scanning electron microscope (SEM) equipped with an energy dispersive spectrometer (EDS). Phase identification is determined by X-ray diffraction (XRD). Much effort is needed to carefully assess scales adhesion. The test exposed the effect of copper composition selected to achieve hot-rolled steel properties on the formation and adhesion of oxide scales. Adhesive strength involves possible contributions from physical, chemical and mechanical. Characterisation of the scales-steel interface after the test would be much useful for exact investigation and still as a significant challenge. The experiment shows that element copper is already obvious at interface of relatively high levels of Cu of 0.225%. This effect is even decreased the adherence of oxide scales on the steel substrate by forming copper-rich phase at interface. Steel industry should more apply techniques such as tensile test for scales adhesion determination. This research is continuously looking for new method to assess scale adhesion. Consideration with results indicates that the element copper should more reasonably estimate for its effect to industrial descaling process.

Study and Prediction of Plasticization of Hot-Rolled Steel Strip During Straightening Under Conditions of Alternating Deformation

Study and Prediction of Plasticization of Hot-Rolled Steel Strip During Straightening Under Conditions of Alternating Deformation

Small samples data augmentation and improved MobileNet for surface defects classification of hot-rolled steel strips

Surface defects in hot-rolled steel strips are one of the common product problems for the steel industry, which harm the product appearance, affect the corrosion and wear resistance, and shorten the product service life. The natural defect samples are sparse, category imbalanced, and expensive manual annotations. Therefore, it is crucial to study the data augmentation and classification methods for small sample surface defects. To solve the above problems and improve the accuracy and real-time performance of defect classification, we propose a random offline data augmentation algorithm (Random-CutMix) and an improved MobileNet architecture (SP-MobileNet). The Random-CutMix algorithm expands the dataset by random sampling to balance the number of each defect class. The SP-MobileNet combines the inverse residual module with the channel shuffle mechanism (CSIn-Module) and pyramid split attention (PSA) module, which facilitates cross-group information flow and improves model representation capability and generalization performance with low computational cost. The accuracy, recall, F1 score, parameter, computational complexity, and frame rate of SP-MobileNet with Random-CutMix on the X-SDD dataset were 95.97%, 95.22%, 95.46%, 6.5 M, and 0.54 G, 72 FPS, respectively. The experiment results indicate that our method outperforms the state-of-the-art methods and provides an effective trade-off between accuracy and instantaneity in actual industrial production.

The coupling machine learning for microstructural evolution and rolling force during hot strip rolling of steels

Microstructural evolution during hot rolling, which is complex and unperceptive but has direct effects on rolling forces, determines the final microstructure and mechanical properties of steel products. In this paper, a comprehensive set of machine learning (ML) models was developed through rolling forces to reveal the evolutions of recrystallization and grain size of austenite, in which the grain size was estimated by considering the grain shape effect after each rolling pass. By using the backpropagation neural network (BPNN) and genetic algorithm (GA) combined with industrial data, the quantitative connections of parameters in the models for static recrystallization (SRX), meta-dynamic recrystallization (MDRX), and mean flow stress (MFS) were established with steel chemical compositions and deformation conditions. Through the machine learning (ML) approach, softening kinetics and rolling forces can be more accurately predicted than traditional models. Also, the recrystallization behavior and evolution of austenite grain size during hot rolling were analyzed by using the proposed ML models, which are consistent with experimental results.

Bending Force of Hot Rolled Strip Based on Improved Whale Optimization Algorithm and Twinning Support Vector Machine

Bending control is one of the main methods of shape control for the hot rolled plate. However, the existing bending force setting models based on traditional mathematical methods are complex and have low control accuracy, which leads to poor strip exit shapes. Aiming at the problem of complex bending force setting of the traditional algorithm, an improved whale swarm optimization algorithm and twin support vector machine-based bending force model for hot rolled strip steel (LWOA-TSVR) is proposed. Based on the hot rolling field production data of a steel plant, the research group established the bending force prediction model by using the nonlinear approximation ability of the twin support vector machine. The introduction of the Levy flight improvement algorithm improves the generalization ability, prediction accuracy, and convergence speed of the whale swarm optimization algorithm with the help of the convergence of coefficient vectors, solves the problem of a random selection of the parameters of the traditional whale swarm optimization algorithm and optimizes the ability of the whale swarm algorithm to jump out of the local optimum. Based on the actual rolling database, the hit rate of the proposed method reaches 91% (from −5 to 5 KN), which fully meets the requirements of the detection accuracy on the actual production line. The model is not only able to overcome the local search to obtain the global optimal solution, but also has the advantages of fast convergence and higher prediction accuracy. A comparison of the model with twin support vector machines and traditional whale swarm algorithms shows that the prediction accuracy is higher. The experimental results also show that this model has advantages over existing bending force prediction models in terms of improving the accuracy of the strip shape control and providing theoretical guidance for practical bending force settings.

High-resolution image-based surface defect detection method for hot-rolled strip steel

Existing methods for detecting surface defects in strip steel are mainly aimed at the case of low-resolution images, and little research has been done in the case of high-resolution images with a small percentage of valid targets. In response to the practical problems in the hot-rolled flattening stage, we propose a hot-rolled strip surface defect dataset called XLData-CLS, which contains a total of 5000 defective samples and 10,000 normal samples with an image resolution of 1344 × 672 pixels. Based on this dataset, this paper improves the ResNet152 model by adding a feature extraction layer at the front of the model so that it can extract information more effectively for high-resolution images. Experimental results on the XLData-CLS dataset show that the proposed method achieves 99.00% classification accuracy on the test set, which is higher than other comparable deep learning models. Validation results on another datasets also demonstrate the accuracy of our method for high-resolution defective images.

SDDA: a method of surface defect data augmentation of hot-rolled strip steel

The detection of surface defects in hot-rolled strip steel is an important step in the production process of hot-rolled strip steel, and it is closely related to the final quality. At present, many researchers have conducted many in-depth studies on the classification of strip steel surface defects and achieved fruitful results. However, current research has always focused on modifying the model; this approach has limited improvement in classification accuracy and often brings a reduction in model efficiency. A data augmentation method named SDDA, which can perform multiscale transformations on the images input to the deep learning models, is proposed and validated on X-SDD dataset by MixNet model. The test results show that after the SDDA data augmentation, the accuracy of the model reached 96.81%, which is 1.71% higher than the previous best result; meanwhile, since the improvement occurs on the data side, it only leads to an increase in training time and does not increase the inference time of the model. Through comparison experiments, it can be found that our data augmentation approach provides significant improvements in multiple metrics such as accuracy and F1 for a variety of deep learning models.