A method for predicting a composition change curve in a mixed casting process of different steel grades

Abstract

The present application relates to a kind of prediction method of component change curve in different steel mixed pouring process, intermediate ladle flow field numerical simulation is carried out for mixed pouring process under different typical process conditions and the simulation result is stored in computer, all numerical simulation results are retrieved based on current molten steel mixing ratio, the rate of change when the molten steel mixing ratio under different process conditions is drawn scatter diagram, interpolation calculation is carried out in scatter diagram and the rate of change of mixing ratio under all process conditions is obtained, the predicted value of mixing ratio and molten steel composition at next time is calculated based on the rate of change of mixing ratio under current process condition, repeat the above calculation process, until mixed pouring process is completed.This method effectively obtains the evolution process of molten steel composition in whole time process by numerical simulation means, and the molten steel composition prediction result under different process conditions is obtained by using appropriate interpolation method, which can fully adapt to different continuous casting machine production process and improve reliable prediction result to guide mixed pouring production slice.

Description

Technical Field

[0001] This patent application belongs to the field of continuous casting technology in the metallurgical industry, and more specifically, it relates to a method for predicting the composition change curve during the mixing of different steel grades. Background Technology

[0002] In the continuous casting production of modern steel enterprises, in order to improve production efficiency, reduce production costs, and reduce the number of unplanned ladle changes, the technique of mixing different steel grades is often adopted. That is, in the same casting cycle, when the previous heat of molten steel is about to be finished, the next heat of molten steel with different composition (i.e., different steel grades) is directly and continuously poured through the tundish.

[0003] However, this mixed casting process inevitably produces a "transition billet" with a composition between the two steel grades. The composition of this transition billet is dynamic, and its length and composition distribution directly affect product quality and production costs. If the length of the transition billet is estimated to be too long, a large amount of qualified molten steel will be downgraded or recycled, resulting in huge waste; if it is estimated to be too short, some unqualified products may be released, affecting downstream users.

[0004] Currently, the prediction of the composition of the tundish billet mainly relies on empirical models or simple physical models. Empirical models typically assume that the mixing of molten steel in the tundish is an ideal state such as "piston flow" or "completely mixed flow," but these models ignore the complex flow field structure in the tundish (such as dead zones and short-circuit flows), resulting in poor prediction accuracy. Physical simulations (such as hydraulic simulations) can more realistically reflect the flow field, but they are costly, time-consuming, and difficult to cover the variable process parameters in the production field (such as casting speed and tundish capacity), making it impossible to achieve online real-time prediction.

[0005] Therefore, how to provide a method that can accurately and in real time predict the changes in molten steel composition during the mixing of different steel grades and adapt to different production conditions in order to guide the production of precise billet cutting is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a method for predicting the composition change curve during the mixing of different steel grades. This method can comprehensively consider the complex fluid dynamics within the tundish and adapt to the varying process conditions on site, achieving high-precision, real-time prediction of the steel composition throughout the mixing process. This provides a reliable basis for accurate billet marking and effectively solves the aforementioned problems in the background technology.

[0007] To solve the above problems, the technical solution adopted by the present invention is as follows: A method for predicting the composition change curve during the mixing of different steel grades includes the following steps: S1. Establish a numerical simulation database: For various typical process conditions during continuous casting and mixing, numerical simulations of the flow field are performed on the mixing process of different steel grades in the tundish. The sequence data of the steel mixing ratio changing with time under each process condition are obtained as simulation results, and all simulation results are stored in the numerical simulation database. S2. Real-time data acquisition and retrieval: During the mixed casting production process, the process conditions and steel mixing ratio at the current moment are acquired at regular intervals, and based on the current steel mixing ratio, the mixing ratio change rate data corresponding to the mixing ratio at the current moment under all process conditions are retrieved from the numerical simulation database. S3. Interpolation calculation: Using each process condition as the coordinate axis, the retrieved mixing ratio change rate data is plotted into a three-dimensional scatter plot, and the interpolation algorithm is used to perform interpolation calculation on the scatter plot. The mixing ratio change rate under the current actual process conditions is found in the interpolation calculation results. S4. Predict the steel mixing ratio and composition: Based on the current steel mixing ratio and the mixing ratio change rate obtained in step S3, calculate the predicted steel mixing ratio for the next moment, and calculate the predicted steel composition for the next moment based on the detected composition of the two steel grades. S5. Repeat steps S2 to S4 until the casting process is completed, thereby obtaining the predicted curve of steel composition change during the entire casting process.

[0008] In step S1, the process conditions are as follows: the process parameters in continuous casting and mixing production are combined in the form of a rectangular array, and the process parameters include the amount of steel passed through and the amount of steel remaining in the tundish. The numerical simulation of the flow field adopts the computational fluid dynamics (CFD) method, which simulates the flow and mixing process of molten steel by solving the Navier-Stokes equations and the component transport equations.

[0009] In step S2, the current steel mixing ratio is 0 at the start of casting, and its value at subsequent times is the result of repeated iterations from steps S2 to S4. The mixing ratio change rate is calculated as follows: the numerical simulation database contains simulation results, and the result is the difference between the mixing ratio after time Δt and the current mixing ratio. Δt is the time step, measured in seconds (s).

[0010] In step S3, the interpolation method is as follows: The tundish residual steel quantity parameter in the three-dimensional scatter plot is fixed sequentially. A two-dimensional scatter plot is generated with the steel throughput as the x-axis and the mixing ratio change rate as the y-axis. A third-order Bézier curve is used to smooth the two-dimensional scatter plot of the current tundish residual steel quantity. Points are taken at intervals of 0.1 t / min on the smoothed Bézier curve, and the tundish residual steel quantity value, steel throughput value, and mixing ratio change rate value at each point are recorded. This process is repeated to smooth all two-dimensional scatter plots with different tundish residual steel quantity values ​​and record the results. In the three-dimensional scatter plot composed of all records, the steel throughput parameter is fixed sequentially. A two-dimensional scatter plot is generated with the tundish residual steel quantity as the x-axis and the mixing ratio change rate as the y-axis. A third-order Bézier curve is used to smooth the two-dimensional scatter plot of the current steel throughput. Points are taken at intervals of 0.1 t on the smoothed Bézier curve, and the tundish residual steel quantity value, steel throughput value, and mixing ratio change rate value at each point are recorded. Then, continue to smooth all the two-dimensional scatter plots for different steel volume values ​​and record them. This record will be used as the final interpolation result.

[0011] In step S4, the predicted value of the steel-molten steel mixing ratio at the next moment is calculated using the following formula: P(t+Δt)=P(t)+R(t) Where P(t) is the steel-molten steel mixing ratio at the current time t; P(t+Δt) is the predicted steel-molten steel mixing ratio at the next time t+Δt; R(t) is the rate of change of the mixing ratio calculated by interpolation at the current time t; and Δt is the time step, in seconds.

[0012] The predicted composition of the molten steel at the next moment is calculated using the following formula: C(t+Δt)=C0+P(t+Δt)(C1-C0) Where C(t+Δt) is the predicted value of the molten steel composition at the next time t+Δt, in units of %; P(t+Δt) is the predicted value of the molten steel mixing ratio at the next time t+Δt; C0 is the detected value of the molten steel composition of the front furnace, in units of %; and C1 is the detected value of the molten steel composition of the rear furnace, in units of %.

[0013] Due to the adoption of the above technical solution, the beneficial effects achieved by this invention are: 1. High prediction accuracy: This invention is based on CFD numerical simulation, which fully considers the complex fluid dynamics phenomena in the intermediate package, fundamentally overcoming the shortcomings of traditional empirical models that are too simplistic, and making the prediction results closer to physical reality.

[0014] 2. High adaptability: By establishing a numerical simulation database covering a variety of typical working conditions and adopting advanced interpolation algorithms, this method can flexibly adapt to the production needs of different continuous casting machines and different process parameters, without the need for complex online simulations for each working condition.

[0015] 3. Good real-time performance: The core calculation process of this method (retrieval, interpolation, prediction) has a small computational load and a fast response, which can meet the requirements of online real-time prediction and provide timely guidance for production operations.

[0016] 4. Clear guiding significance: This method can output the composition change curve of the entire process, providing reliable data support for accurately determining the start and end points of the transition billet and optimizing the billet cutting scheme. It helps to minimize scrap and downgraded products and significantly improve the economic benefits of enterprises. Attached Figure Description

[0017] Figure 1 This is a flowchart of the present invention.

[0018] Figure 2 This is a predicted curve showing the change of steel composition over time in this invention. Detailed Implementation

[0019] The present invention will be further described in detail below with reference to the embodiments.

[0020] A method for predicting the composition change curve during the mixing of different steel grades, such as... Figure 1 The flowchart shown includes the following steps; S1. Establish a numerical simulation database: For various typical process conditions during continuous casting and mixing, numerical simulations of the flow field are performed on the mixing process of different steel grades in the tundish. The sequence data of the steel mixing ratio changing with time under each process condition are obtained as simulation results, and all simulation results are stored in the numerical simulation database. S2. Real-time data acquisition and retrieval: During the mixed casting production process, the process conditions and molten steel mixing ratio at the current moment are acquired at regular intervals, and based on the current molten steel mixing ratio, the mixing ratio change rate data corresponding to the current molten steel mixing ratio at the time of all process conditions are retrieved from the numerical simulation database. S3. Interpolation calculation: Using each process condition as the coordinate axis, the retrieved mixing ratio change rate data is plotted into a three-dimensional scatter plot, and the interpolation algorithm is used to perform interpolation calculation on the three-dimensional scatter plot. The mixing ratio change rate under the current actual process conditions is found in the interpolation calculation results. S4. Predict the steel mixing ratio and composition: Based on the current steel mixing ratio and the mixing ratio change rate obtained in step S3, calculate the predicted steel mixing ratio for the next moment, and calculate the predicted steel composition for the next moment based on the detected composition of the two steel grades. S5. Repeat steps S2 to S4 until the casting process is completed, thereby obtaining the prediction curve of steel composition change during the entire casting process.

[0021] In step S1, the "each process condition" in "obtaining each process condition" refers to the process parameters in continuous casting mixed casting production being combined in the form of a rectangular array. The above process parameters include the amount of steel passed through and the amount of steel remaining in the tundish.

[0022] In step S1, the numerical simulation of the flow field adopts the computational fluid dynamics (CFD) method, which simulates the flow and mixing process of molten steel by solving the Navier-Stokes equations and component transport equations.

[0023] In step S2, the current molten steel mixing ratio is 0 at the beginning of the mixing process, and the value of the current molten steel mixing ratio at subsequent times is the value calculated by repeated iterations from step S2 to S4.

[0024] In step S2, the method for calculating the rate of change of the mixing ratio is as follows: The difference between the mixing ratio after time step Δt and the current mixing ratio in the numerical simulation database (which contains simulation results), where Δt is the time step and the unit is seconds (S).

[0025] In step S3, the interpolation algorithm is as follows: S31. Sequentially fix the tundish remaining steel quantity parameter in the three-dimensional scatter plot, generate a two-dimensional scatter plot with the steel flow rate as the x-axis and the mixing ratio change rate as the y-axis, smooth the two-dimensional scatter plot of the current tundish remaining steel quantity parameter with a third-order Bézier curve, take points at intervals of 0.1t / min on the smoothed Bézier curve and record the tundish remaining steel quantity value, steel flow rate value and mixing ratio change rate value at each point. S32. Then continue to smooth all the two-dimensional scatter plots under different tundish residual steel values ​​and record them. In the three-dimensional scatter plot composed of all records, fix the steel flow parameters in sequence, and generate a two-dimensional scatter plot with the tundish residual steel value as the x-axis and the mixing ratio change rate as the y-axis. Smooth the two-dimensional scatter plot of the current steel flow using a third-order Bézier curve. On the smoothed Bézier curve, take points at intervals of 0.1t of tundish residual steel and record the tundish residual steel value, steel flow value and mixing ratio change rate value at each point. S33. Then continue to smooth all the two-dimensional scatter plots under different steel volume values ​​and record them. This record is used as the final interpolation calculation result.

[0026] In step S4, the predicted value of the steel-molten steel mixing ratio at the next moment is calculated using the following formula: P(t+Δt)=P(t)+R(t) Where P(t) is the steel-molten steel mixing ratio at the current time t; P(t+Δt) is the predicted steel-molten steel mixing ratio at the next time t+Δt; R(t) is the rate of change of the mixing ratio calculated by interpolation at the current time t; and Δt is the time step, in seconds.

[0027] The predicted composition of the molten steel at the next moment is calculated using the following formula: C(t+Δt)=C0+P(t+Δt)(C1-C0) Where C(t+Δt) is the predicted value of the molten steel composition at the next time t+Δt, in units of %; P(t+Δt) is the predicted value of the molten steel mixing ratio at the next time t+Δt; C0 is the detected value of the molten steel composition of the front furnace, in units of %; and C1 is the detected value of the molten steel composition of the rear furnace, in units of %.

[0028] The following detailed description is provided in conjunction with specific examples.

[0029] This embodiment takes a slab continuous casting machine in a large steel plant as the application object, and uses the composition change curve prediction method of the present invention during the mixing of different steel grades to predict the composition of the mixing process from steel grade Q355B to steel grade Q235B.

[0030] Step S1: Establish a numerical simulation database First, the key process parameters affecting molten steel mixing and their typical value ranges are determined. In this embodiment, the steel throughput and the amount of steel remaining in the tundish are selected as variables and combined according to a rectangular array: • Steel throughput (t / min): 2, 2.5, 3, 3.5 • Remaining steel in the tundish (t): 10, 15, 20, 25 This results in 4×4=16 typical combinations of process conditions.

[0031] Next, for the aforementioned 16 process conditions, numerical simulations of the flow field were performed using computational fluid dynamics (CFD) software (such as AnsysFluent). The specific process is as follows: 1. Establish a three-dimensional geometric model that is 1:1 scale with the actual intermediate package.

[0032] 2. The model is meshed using an unstructured mesh.

[0033] 3. Set boundary conditions: the inlet is the outlet of the large package long nozzle, the outlet is the outlet of the crystallizer nozzle, and the wall is a no-slip boundary.

[0034] 4. The standard k-ε turbulence model is selected to solve the Navier-Stokes equations to simulate the turbulent flow of molten steel.

[0035] 5. Activate the component transport model, defining the molten steel in the front furnace as component 1 and the molten steel in the rear furnace as component 2. At the start of the simulation, the entire tundish contains component 1; from t=0, the molten steel flowing in at the inlet becomes component 2.

[0036] 6. Set the time step Δt_sim=0.2s and perform transient solution until the mass fraction of component 2 (i.e. the steel mixing ratio P) at the outlet of the tundish reaches 0.99.

[0037] The results of 16 simulations—that is, the sequence data of the steel-to-molten steel mixing ratio P changing with time t under each working condition—are stored in an SQL Server database to form a numerical simulation database.

[0038] Step S2: Real-time data acquisition and retrieval This prediction system is activated when the continuous casting machine begins mixed casting of steel grades Q355B to Q235B. The time step for system prediction is set to Δt = 2s.

[0039] At the initial moment t=0 when the mixing begins, the current steel mixing ratio P(0)=0. The system obtains the current process conditions in real time from the production process PLC (L1 system) via the OPC protocol: the steel throughput is 2.3t / min, and the remaining steel in the tundish is 13.5t.

[0040] Based on the current mixing ratio P(0)=0, the system retrieves the mixing ratio values ​​P_sim(2s) for all 16 operating conditions after a Δt=2s interval from 0 from the numerical simulation database. Then, it calculates the mixing ratio change rate R_sim for each operating condition: R_sim=P_sim(2s)-P(0) For example, the 16 sets of data retrieved might look like this (illustrated):

[0041] Step S3: Interpolation Calculation Based on the 16 data points (through steel volume, surplus steel volume, R_sim) retrieved above, the system performs two third-order Bézier curve interpolations.

[0042] 1. First interpolation (fixed surplus steel): • With a fixed surplus steel amount of 10t, generate a two-dimensional scatter plot of the steel throughput (2, 2.5, 3, 3.5) and the corresponding R_sim values.

[0043] • Smooth the scatter plot with a third-order Bézier curve to obtain a continuous R-stainless steel throughput curve.

[0044] • On this curve, take points at intervals of 0.1t / min (2, 2.1, 2.2, 2.3...) and record the (throughput, surplus steel = 10t, R) value for each point.

[0045] • Repeat the above operation for data with surplus steel of 15t, 20t and 25t to obtain four sets of dense data points.

[0046] 2. Second interpolation (fixed steel flow): • Using all the dense data points generated in the previous step, fix the steel throughput at 2t / min, and generate a two-dimensional scatter plot of the remaining steel in the tundish (10, 15, 20, 25) and the corresponding R_sim values.

[0047] • Smooth the scatter plot with a third-order Bézier curve to obtain a continuous R-ballast residual steel quantity curve.

[0048] • On this curve, take points at intervals of 0.1t (10, 10.1, 10.2, 10.3...) and record the value of (steel throughput = 2t / min, surplus steel, R) at each point.

[0049] • Repeat the above operation for data with steel throughput of 2.1t, 2.2t, 2.3t…3.5t to obtain 16 sets of dense data points. Record the (steel throughput, surplus steel, R) value for each point.

[0050] • In this record, find the R value corresponding to a steel throughput of 2.3 t / min and a tundish residual steel quantity of 13.5 t. Calculate the mixing ratio change rate R(0) = 0.065 under the current actual process conditions (steel throughput 2.3 t / min, tundish residual steel quantity 13.5 t).

[0051] Step S4: Predict the mixing ratio and composition of molten steel Before making predictions, it is necessary to obtain the compositional analysis values ​​for both steel grades. Assumption: • The manganese content of the molten steel (grade Q355B) was measured at C0 = 1.266%. • The manganese content of the molten steel (Q235B grade) in the later furnace was measured at C1 = 0.161%. Based on R(0) obtained in step S3 and the current mixing ratio P(0), calculate the predicted value for the next time step t+Δt (i.e., 2s): 1. Predict the mixing ratio of molten steel: P(t+Δt)=P(t)+R(t) P(2s) = P(0) + R(0) = 0 + 0.065 = 0.065 2. Predicting the composition of molten steel: C(t+Δt)=C0+P(t+Δt)(C1-C0) C(2s)=1.266%+0.065*(0.161%-1.266%)=1.194% The system uses P(2s)=0.065 and C(2s)=1.194% as the prediction results at time t=2s and plots the point on the monitoring interface.

[0052] Step S5: Repeat steps S2 to S4. The system enters the next 2-second cycle. At t=2s: • The current mixing ratio is updated to P(2s)=0.065.

[0053] • The current process conditions (throughput and remaining steel in the tundish) are retrieved again.

[0054] • The system calculates the mixing ratio P(4s) and component C(4s) at time t=4s by repeatedly executing S2 (retrieval), S3 (interpolation) and S4 (prediction) based on P(2s)=0.065.

[0055] This cyclical process repeats continuously until the predicted steel mixing ratio P(t) reaches 0.99, at which point the system determines the mixing process is complete. Throughout the process, the system displays a real-time predicted curve of the steel composition (e.g., manganese content C) changing over time on the computer screen, providing decision-making support for on-site operators. Figure 2 , is the curve showing the change in Mn element content during this mixing process. The vertical axis represents the Mn element content (%), and the horizontal axis represents the billet position (m) at different time points after the start of mixing.

[0056] Through this embodiment, the method of the present invention can quickly and accurately predict the composition changes during the mixing of different steel grades based on an offline simulation database and real-time process parameters. This solves the problems of traditional methods relying on experience and having low accuracy, and provides strong technical support for achieving precise cutting of transition billets, improving metal yield, and ensuring product quality.

Claims

1. A method for predicting the composition change curve during the mixing of different steel grades, characterized in that: Includes the following steps S1. Establish a numerical simulation database: For various typical process conditions during continuous casting and mixing, numerical simulations of the flow field are performed on the mixing process of different steel grades in the tundish. The sequence data of the steel mixing ratio changing with time under each process condition are obtained as simulation results, and all simulation results are stored in the numerical simulation database. S2. Real-time data acquisition and retrieval: During the mixed casting production process, the process conditions and molten steel mixing ratio at the current moment are acquired at regular intervals, and based on the current molten steel mixing ratio, the mixing ratio change rate data corresponding to the current molten steel mixing ratio at the time of all process conditions are retrieved from the numerical simulation database. S3. Interpolation calculation: Using each process condition as the coordinate axis, the retrieved mixing ratio change rate data is plotted into a three-dimensional scatter plot, and the interpolation algorithm is used to perform interpolation calculation on the three-dimensional scatter plot. The mixing ratio change rate under the current actual process conditions is found in the interpolation calculation results. S4. Predict the steel mixing ratio and composition: Based on the current steel mixing ratio and the mixing ratio change rate obtained in step S3, calculate the predicted steel mixing ratio for the next moment, and calculate the predicted steel composition for the next moment based on the detected composition of the two steel grades. S5. Repeat steps S2 to S4 until the casting process is completed, thereby obtaining the predicted curve of steel composition change during the entire casting process.

2. The method for predicting the composition change curve during the mixing of different steel grades according to claim 1, characterized in that: In step S1, the "each process condition" in "obtaining each process condition" refers to the process parameters in continuous casting mixed casting production being combined in the form of a rectangular array. The above process parameters include the amount of steel passed through and the amount of steel remaining in the tundish.

3. The method for predicting the composition change curve during the mixing of different steel grades according to claim 1, characterized in that: In step S1, the numerical simulation of the flow field adopts the computational fluid dynamics method, which simulates the flow and mixing process of molten steel by solving the Navier-Stokes equations and component transport equations.

4. The method for predicting the composition change curve during the mixing of different steel grades according to claim 1, characterized in that: In step S2, the current molten steel mixing ratio is 0 at the start of the mixing process, and the value of the current molten steel mixing ratio at subsequent times is the value calculated by repeated iterations from step S2 to S4.

5. The method for predicting the composition change curve during the mixing of different steel grades according to claim 1, characterized in that: In step S2, the method for calculating the rate of change of the mixing ratio is as follows: The difference between the mixing ratio after time Δt and the current mixing ratio in the numerical simulation database, where Δt is the time step and the unit is seconds (S).

6. The method for predicting the composition change curve during the mixing of different steel grades according to any one of claims 1-5, characterized in that: In step S3, the interpolation algorithm is as follows: S31. Sequentially fix the tundish remaining steel quantity parameter in the three-dimensional scatter plot, generate a two-dimensional scatter plot with the steel flow rate as the x-axis and the mixing ratio change rate as the y-axis, smooth the two-dimensional scatter plot of the current tundish remaining steel quantity parameter with a third-order Bézier curve, take points at intervals of 0.1t / min on the smoothed Bézier curve and record the tundish remaining steel quantity value, steel flow rate value and mixing ratio change rate value at each point. S32. Then continue to smooth all the two-dimensional scatter plots under different tundish residual steel values ​​and record them. In the three-dimensional scatter plot composed of all records, fix the steel flow parameters in sequence, and generate a two-dimensional scatter plot with the tundish residual steel value as the x-axis and the mixing ratio change rate as the y-axis. Smooth the two-dimensional scatter plot of the current steel flow using a third-order Bézier curve. On the smoothed Bézier curve, take points at intervals of 0.1t of tundish residual steel and record the tundish residual steel value, steel flow value and mixing ratio change rate value at each point. S33. Then continue to smooth all the two-dimensional scatter plots under different steel volume values ​​and record them. This record is used as the final interpolation calculation result.

7. The method for predicting the composition change curve during the mixing of different steel grades according to claim 1, characterized in that: In step S4, the predicted value of the steel-molten steel mixing ratio at the next moment is calculated using the following formula: P(t+Δt)=P(t)+R(t) Where P(t) is the steel-molten steel mixing ratio at the current time t; P(t+Δt) is the predicted steel-molten steel mixing ratio at the next time t+Δt; R(t) is the rate of change of the mixing ratio calculated by interpolation at the current time t; Δt is the time step, in seconds. The predicted composition of the molten steel at the next moment is calculated using the following formula: C(t+Δt)=C0+P(t+Δt)(C1-C0) Where C(t+Δt) is the predicted value of the molten steel composition at the next time t+Δt, in units of %; P(t+Δt) is the predicted value of the molten steel mixing ratio at the next time t+Δt; C0 is the detected value of the molten steel composition of the front furnace, in units of %; and C1 is the detected value of the molten steel composition of the rear furnace, in units of %.

Images