Asymmetric flow in multi-mode continuous casting and rolling mold under electromagnetic braking

A critical review of the challenges of developing continuous casting mold fluxes for high-Ti steels

The low hardness of copper alloys, which are the substrate material used for continuous casting molds, makes them prone to plastic deformation, wear, and high-temperature oxidation, leading to premature failure and the formation of surface defects on billets. In this work, the microstructure, phase composition, mechanical, and tribological properties of Cr3C2–NiCr coatings deposited by high-velocity oxy-fuel (HVOF) spraying onto copper substrates used in molds were investigated. This research was driven by the need to extend the service life of copper molds in continuous steel casting processes. It was established that spraying parameters have a decisive influence on porosity, coating thickness, microhardness, and friction behavior under conditions simulating billet contact with the working surface of the mold. Among the investigated regimes, the coating deposited at a powder feed rate of 11.39 m/s exhibited a dense lamellar structure and the highest level of microhardness. Tribological tests confirmed that this coating exhibited the lowest coefficient of friction, whereas the other coatings were characterized by higher porosity and poorer wear resistance. Thus, the results emphasize the necessity of optimizing spraying parameters to develop highly effective HVOF protective coatings for copper molds operating under extreme thermomechanical loads during steel casting.

This study employs a numerical simulation approach to investigate argon bubble flow behavior within a steel continuous casting mold, with a focus on the impact of bubble swarm correction models. Three scenarios are compared: one without any correction and two incorporating drag coefficient corrections, specifically designed for bubble swarm effects. The results demonstrate that incorporating these correction models significantly improves the predictive accuracy of simulations. In particular, the inclusion of a bubble swarm correction model reduces the error in predicted bubble trajectories by 51.7% and 23.0%, respectively, when measured by Hausdorff distances against experimental trajectory data, compared to the scenario without corrections. These findings underline the importance of selecting an appropriate drag correction model for accurate simulations of bubble dynamics and their interaction with the liquid steel in continuous casting molds. This study highlights that drag correction models tailored to the specific conditions of the continuous casting process are essential for achieving realistic predictions.

Background Thermal and fluid-mechanical conditions in continuous casting (CC) moulds are only roughly known although highly relevant for the product quality. Manual process control is difficult due to the big number of influencing factors. During continuous casting, manual top-freezing controls must be carried out. Every manual performed mould control can affect the strand quality and even increase the risk of failure. Therefore, regular top-freezing controls are performed after a certain casting duration. However, top-freezing events between the regular controls cannot be detected and are a major risk for plant safety. Methods In the RFCS project RealTimeCastSupport, the aim of the research was the digitalisation and optimised control of continuous casting machines. A real-time support system was developed to predict quality-relevant top-freezing events and thus achieve improved control. This was reached by offline material tracking, synchronisation of data streams and statistical analysis by application of Big Data technologies, the development of a digital twin and the exploitation of various CC data and surface inspection to predict reliability of steel production. Results The following results were achieved: Identification of defect promoting scenarios by correlation of statistical results and surface defect detection. Realisation of an offline 3D digital twin of the mould considering heat transfer, inert gas feeding and solidification. Offline reproduction of the identified defect promoting scenarios with the 3D digital twin to find thermal and fluid mechanical reasons for the detected behaviour. Conclusions The application of a real-time support system enables the prediction of top-freezing events during the whole casting process. Subsequently, this significantly increases the plant safety and offers to carry out top-freezing inspections in a more targeted manner in the future. This publication is part of a series of papers in the frame of the dissemination project METACAST.

The dispersed bubbly flow dramatically changes the shear-induced turbulence (SIT) in gas–liquid systems. Modifications to the liquid turbulence model by adding source terms or a bubble-induced turbulence (BIT) viscosity term can improve the accuracy of numerical simulation. However, there is no consensus on the definition of the characteristic time scale of turbulent eddies induced by bubbles, which directly relates to the modeling work of the BIT source. We compared the effects of different character time scale models on multi-scale bubbly flow and developed a novel BIT model based on the gas-phase intrinsic turbulence parameters. Then, the coupling mechanism between SIT and BIT models was revealed. The results show that the population balance and Sato models show better consistency than other models for simulating bubble diameter and flow regimes. The bubble response time significantly underestimates the characteristic time, thereby enhancing the collision frequency between bubbles. A novel mixture BIT model is proposed based on the turbulent properties of gas and liquid phases. Compared with the numerical simulation of the without BIT model, the mean relative error predicted by the novel BIT is reduced by 45.28%. Finally, introducing the BIT viscosity model and the smallest eddy dissipation time model forms a positive feedback loop between BIT and liquid-phase SIT, thereby enhancing turbulent dispersion. However, an anomalous negative feedback loop mechanism was formed between BIT and SIT when using the bubble response time model.

Abstract A full range overall heat flux density distribution curve of the hot surface of a copper plate along the casting direction has been built. Most researchers use local heat flux density approximation function expressions. It cannot truly reflect the heat flux density of copper plates in slab continuous casting mold in current production sites. By embedding thermocouples in the copper plate and using the reverse heat transfer method verified by measured temperature data, the original local heat flux density distribution along the casting direction was corrected to obtain numerical simulation data that was closer to real industrial production. A finite element model of the temperature field of the chamfered copper plate and a shrinkage model of the chamfered slab shell were established using ANSYS finite element software. The temperature distribution of the silicon steel slab chamfered mold copper plate and shell shrinkage distribution of the mold slab in a certain factory was obtained. From the research results, the chamfering design and sink setting of the factory’s chamfered copper plate are reasonable, and the temperature distribution of the copper plate near the chamfering is uniform, with relatively low-temperature values. The shrinkage of the silicon steel slab’s chamfered area is greater than that of the narrow surface center, and the area with the largest shrinkage is at the junction of the chamfer and the narrow surface, which the width-wise shrinkage at the exit of the mold reaches 5.61 mm. According to the numerical simulation results, the current structure of the multi-taper chamfering mold for silicon steel is reasonable, which helps to transfer heat evenly at the chamfering area and avoid corner defects.

In this study, the effect of the slab continuous casting mold thickness on the mold flow field, the fluctuation of the mold flux and molten steel interface (MF‐MSI), the solidification of molten steel, and the removal and capture of bubbles and inclusions is investigated by numerical simulation and high‐temperature quantitative measurement of the mold surface flow velocity (MSFV). With an increase in the thickness from 180 to 250 mm and 320 mm, the measurement results decrease from 0.2148 m s −1 to 0.2074 m s −1 and 0.1875 m s −1 , respectively. The numerical simulation results present excellent alignment with high‐temperature measurement results of MSFV. The occurrence ratios of the mold flux, bubble, and inclusion defects are all reduced. The values of Δ H decrease from 12.91 mm to 11.79 mm and 9.43 mm. The ratios of the bubbles captured by the solidified shell decrease from 1.25% to 0.86% and 0.58%, the ratios of inclusions removed by the mold flux layer increase from 29.01% to 29.80% and 33.90%, and the ratios of inclusions captured by the solidified shell decrease from 33.69% to 28.76% and 23.54%, respectively.

Subsurface hooks formation during initial solidification in the continuous casting mould degrades the quality of steel slabs owing to the associated entrapment of non-metallic inclusions. To understand the mechanism for hook-capturing inclusions, the characteristics of hooks and the inclusion distribution around hooks were investigated by etching experiment and electron microscopic observation. The results reveal that compared with the general shell, hooks had a stronger ability to capture inclusions of 20–300 μm, especially for large inclusions above 250 μm, and the capture area was within their maximum depth. By analysing the forces on inclusions of different sizes at the solidification front, it is found that the Marangoni force has an obvious effect on the entrapment of inclusions. Based on a comparison of the inclusion floating inclination angle, the initial shell and hook inclination angle, the mechanism for the entrapment of inclusions by hooks and shell were proposed. According to this mechanism, the number density of inclusions smaller than 240 μm is greater in the hook zone, while inclusions larger than 240 μm are captured only by hooks. In addition, by studying the characteristics of hooks in slabs produced with different casting parameters, several suggestions are put forward for controlling hooks in industrial production: superheat degree can be set greater than 25 °C and molten steel flow rate can be set greater than 0.35 m 3 /min, and slab surface cleaning technique should be adopted.

Abstract Fiber optic sensor with Bragg Gratings (FBG) was employed as temperature sensors in continuous casting mold of molten steel. It is superior to conventional thermocouple that temperature profile along the direction of casting can be measured in more finely manner, depending on the number of gratings and interval between each grating. By installing two fiber optic sensors along the direction of mold thickness, heat flux across the copper plate could be obtained. The characteristics of the heat flux were examined in various ways, and the influence of magnetic field imposed by Electro-Magnetic device was discussed. An impinging point could be detected by FBG, and its presence and location were independently validated by Computational Fluid Dynamic (CFD) simulation. Effect of magnetic field strength, width of the casting mold, and depth of the SEN immersion were systematically analyzed. FBG can also be applied to detect possibility of solidified shell remelting by analyzing the heat transfer near the impinging point. The proposed technique can be further used in visualizing the molten steel flow in the casting mold.

In order to enhance the liquid surface activity in ultra-wide slab mould with a cross-sectional dimension of 2920 mm × 150 mm (with a width-to-thickness ratio of nearly 20), this study established a two-phase flow model for steel and liquid slag, comparing and analysing the effect of submerged entry nozzle (SEN) inclination angle on both the flow field and temperature field within the mould. The primary findings indicated that at a nozzle angle of 15°, the upward reflux was weak, resulting in a liquid surface velocity below 0.23 m/s, slow renewal of molten steel, which were not conducive to the melting of the mould flux. To improve both liquid surface activity and the melting and flow of the mould flux, the nozzle angle was reduced from 15° to 10°. This adjustment allowed the maximum liquid surface flow velocity to reach 0.35 m/s, thereby enhancing liquid surface activity and facilitating greater energy transfer to the liquid surface. The uniform distribution of heat-flux density around the mould further demonstrated that reducing the nozzle angle significantly enhances liquid surface activity. However, when the nozzle angle was further decreased to 5°, strong upward reflux was observed, resulting in the adjacent slag layer being washed away to one-quarter of the mould's width, which increased the risk of slag entrapment and exposure at the corners. Given the large width-to-thickness ratio of the ultra-wide slab mould, the flow jet from the submerged nozzle was prone to early collided with the wide surface during its movement towards the narrow surface. This interaction leaded to a reduction in kinetic energy, weaker upward reflux, lower heat flux near the liquid surface, and larger fluctuations in heat flow distribution along the mould's width. This study illustrates that appropriately reducing the nozzle angle can mitigate these adverse conditions and provides a theoretical foundation for selecting SENs for ultra-wide slab moulds.

Influence of substituting CaO with SrO on the crystallization behavior of CaO-SiO2 based continuous casting mold fluxes

During the continuous casting process, the submerged entry nozzle (SEN) should be maintained at the geometric center of the mold. However, in actual production, factors such as deformation of the tundish bottom and inaccurate positioning of the traversing car occasionally cause SEN offset. SEN offset can make the molten steel flow field in the mold asymmetric, increasing the risks of slag entrainment on the surface of the casting blank and breakout accidents. To evaluate the influence of different SEN offsets on the mold flow field, this study uses a slab continuous casting mold with a cross-section of 920 mm × 200 mm from a specific factory as the research object. Mathematical simulations were used to investigate the influence of SEN offsets (including width-direction and thickness-direction offsets) on the flow behavior of molten steel in the mold. A physical water model at a 1:1 scale was established for verification. Two parameters, the symmetry index (S) and the bias flow index (N), were introduced to quantitatively evaluate the symmetry of the flow field, and the rationality of the liquid-level fluctuation under this flow field was verified using the F-number (proposed by Japanese experts for mold level fluctuation control) from the index model. The results show the following: when the SEN offset in the thickness direction increases from 0 to 50 mm, the longitudinal symmetry index (Sy) of the molten steel flow field in the mold decreases from 0.969 to 0.704—a reduction of 27.4%; the longitudinal bias flow index (Ny) of molten steel level fluctuation increases from 0.007 to 0.186, representing a 25.6-fold increase, and the F-number rises from 4.297 to 8.482; when the SEN offset in the width direction increases from 0 to 20 mm, the transverse-axis symmetry index (Sx) of the flow field decreases gradually from 0.969 to 0.753 at a 20 mm offset, which is a reduction of approximately 22.29%; the transverse-axis bias flow index (Nx) increases from 0.015 to 0.174 at a 20 mm offset—an increase of 10.6 times; and the F-number increases from 4.297 to 5.548. Considering the comprehensive evaluation of horizontal/vertical symmetry indices, bias flow indices, and F-numbers under the two working conditions, the width-direction SEN offset has the most significant impact on the symmetry of the molten steel flow field.

Digital estimation of the solidifying corner shell in continuous casting mold based on a mathematical model of heat flux

Effect of two-phase turbulent modalities and bubble-induced turbulence on polydispersed bubbly flow in continuous casting mold

The Cr-Zr-Cu copper alloy is widely used to manufacture mold for continuous casting. The proper coating can avoid the mold's failure under the condition of a high-temperature and friction environment. Three kinds of Inconel 718/WC-12Co coatings (micro-, nano-, and micro-nano-WC particles) were prepared on the surface of the copper alloy by laser. The phases, microstructure, and elemental distribution of L1–L3 coatings were observed and analyzed by x-ray diffractometer, scanning electron microscope, and energy dispersive spectrometer. The properties of L1–L3 coatings were tested by a Vickers hardness tester and friction and wear testing machine. The results show that the main phases are γ-Ni, M3W3C, MC, W0.15Ni0.85, and W2C in the L1–L3 coatings. In the L1 coating, there is an obvious phenomenon of WC agglomeration. In the L2 coating, there are many columnar, dendritic compounds, and “fish-bone” crystals with small size due to the addition of nanoparticles. In the L3 coating, the structure is scattered due to the uneven distribution of compounds. The average values of L1–L3 coatings' microhardness are 709.6, 851.9, and 600.1 HV0.5, respectively. The L2 coating has the maximum average microhardness. When the experimental temperature is 400 °C during the friction and wear test, the wear rates of L1–L3 coatings are 3.29, 1.89, and 2.83 × 10−4 mm3 N−1 mm−1, respectively. The L2 coating has the minimum wear rate due to the smaller grain size, denier microstructure, and “fish-bone” structure.

This study employs computational fluid dynamics (CFD) coupled with the population balance model (PBM) to explore the sensitivity of the drag model for the predictive accuracy of the gas–liquid two-phase flow in continuous casting (CC) mold. Several single bubble drag models have been numerically evaluated for high turbulent intensity and gas flow rate operation parameters. Then, the influence of turbulence effect and bubble swarm mechanisms on bubble dynamic behaviors are investigated. The predicted mean bubble diameter and flow pattern were studied and compared with the experimental data. The results show that all single bubble drag models, except for the Grace model, can predict the bubble size distribution (BSD) well. Meanwhile, all models significantly overestimate the bubble diameter near the nozzle under high gas flow rate conditions. A novel drag correction factor based on local gas holdup and BSD is proposed, which takes into account both the hindrance effect of small bubbles and the accelerating effect of large bubbles. The proposed drag correction factor can accurately predict BSD and flow pattern transition in the CC mold under high gas holdup regions. Compared with the simulation results of the previous single bubble drag model, the mean relative error predicted by the novel drag correction factor is decreased by 69.33%.

A coupling model of the flow field, initial solidification, alternating magnetic field (AM-field) and movement of argon bubble in the mold was established by combining the k-ε turbulence model with solidification model, Discrete Phase Model (DPM), Volume of Fluid (VOF) model and magnetohydrodynamics model. The industrial measured magnetic flux density (MFD) is consistent with the calculated MFD at a 15 mm interval from the wide wall, which verifies the reliability of the magnetic field model calculation. As the electromagnetic stirring (EMS) current intensity (EMSCI) grows, the calculated velocity of molten steel at 1/4 of the width direction of the mold surface decreases, which coincides well with measurement results. In addition, as the EMSCI is increased, the upper roll flow weakens, the horizontal rotational flow increases. The thickness of the solidified shell in the lower part of the narrow wall is increased due to the decreased impact force of the molten steel on the solidified shell of the narrow wall caused by the upward movement of the circulating flow. The volume fraction of argon bubbles around the SEN further decreases, and the distribution of them becomes more dispersed. Based on the research results, the optimal EMSCI under the present operating conditions is recommended as 700 A.

This study examines the behavior of fluid flow, heat transfer, slag entrainment, and solidification within slab continuous casting molds featuring two distinct cross-sectional geometries. A combination of physical modeling and numerical simulation was utilized to assess the performance of the submerged entry nozzle (SEN) in these molds. The findings reveal that increasing the SEN immersion depth (D) significantly mitigated surface level fluctuations at one-quarter of the mold width, with the 250 × 1550 mm2 cross-section demonstrating superior control compared to the 250 × 2000 mm2 counterpart. However, it was found that oil entrainment was more likely to occur when the D value of the SEN was set at 120 mm and 140 mm. Conversely, depths of 160 mm and 180 mm resulted in a more stable water-oil interface. Additionally, increasing the value of D led to a reduction in temperature uniformity across the surface of the mold. The accuracy of simulated shell thickness was confirmed by comparison with empirical formula predictions. Increasing the value of D from 120 mm to 180 mm led to a reduction in shell thickness by 2.2 mm for the 250 × 2000 mm2 cross-section and 0.4 mm for the 250 × 1550 mm2 cross-section at the mold outlet. These results suggest that the current SEN design is appropriate for use in slab continuous casting with both cross sections. It is recommended that the SEN immersion depth is maintained between 140 mm and 160 mm for optimal performance.

To get rid of the restrictions imposed by the components of the original mold flux and minimize fluorides to the largest extent, this article has designed the continuous casting mold flux for high‐aluminum steel with an aluminum content of 1.35% and tested the performance manifestations of the mold flux by integrating equipment like a rotational viscometer, S/DHTT‐TA‐III thermal analyzer, and X‐ray diffraction. The composition of the continuous casting mold flux for high‐aluminum steel is: CaO = 23–43 wt%, SiO2 = 12–25 wt%, Al2O3 = 14–26 wt%, B2O3 = 13–17 wt%, and Li2O = 4–8 wt%. The newly designed mold flux has a lower melting temperature and its performance still satisfies the conditions for continuous casting after the reaction. The crystallization process of the on‐site using mold flux lasts 198 s, with the precipitated phases being Na2CaSiO4 and Ca3Si2O7. The crystallization process of the newly designed mold flux lasts ≈324 s, with the precipitated phase mainly being LiAlO2, and the mold flux can maintain a low viscosity and a relatively thick liquid slag layer for a longer period of time.