Research And Applications Of Molten Steel Temperature Prediction Model For Ladle Furnace Refining

The temperature of molten steel is one of the most important technical parameters during steelmaking process and reasonable control of molten steel temperature during refining process is an effective means of ensuring continuous casting performance, reducing production costs and energy consumption and improving product quality. It is important to control the molten steel temperature of LF refining process, since the temperature control precision has a great influence on the operation of continuous casting. However, it is impossible to measure the molten steel temperature continuously during the production because of expensive testing costs and limited conditions. Therefore, it is very necessary to predict the molten steel temperature accurately and continuously in order to improve production efficiency and reduce costs.In the present work,135ton ladle furnace was chosen as the research object. An online prediction model for molten steel temperature during LF refining was established based on production data statistics and numerical simulation. With the function of automatic correction the model has been applied to the plant.. The main research work and results are as follows:(1) Numerical simulation was carried out by using the CFD software Fluent to investigate the flow field of LF refining process, and the results show that the increase of gas content just increases the flow velocity without changing the flow pattern and the mixing time becomes shorter. Better in order to molten bath stirring is achieved when the gas flow rate reaches450NL/min, which is helpful for the rapid reaction and the homogenization of composition and temperature, and when the gas flow rate is beyond that, the mixing time tends to be stable. The critical argon flow rate for exposing of100mm thick slag layer is from150NL/min to200NL/min, and when the argon flow rate increases from200NL/min to600NL/min, the corresponding diameter of slag hole increases from240mm to500mm.(2) A numerical model that coupling the flow and heat transfer of LF refining process was established to investigate the effect of steel ladle with different grades of ladle lining to the cooling of molten steel in the refining process, the result shows that:different thermal states of ladle before tapping have a significant effect on the temperature drop of molten steel, the heat loss of D grade ladle is the lowest during the whole ladle flow process, and it has the smallest effect on the temperature drop of molten steel. The molten steel has an average temperature drop rate of0.58℃/min caused by heat loss of ladle lining under this thermal state. Compared A grade ladle with D grade ladle, there is about a35℃difference of molten steel temperature drop within43min and a0.55℃/min difference of steel average temperature drop rate. The steel temperature drop rate decreases with the processing time goes by, and to a certain time, the effect of ladle lining heat loss on molten steel temperature tends to be stable.(3) By analyzing the electrical characteristics value and the dynamic electric parameters on site, the author got two important parameters of LF Furnace:re=0.69mΩ, xe=3.6mΩ. The author made an electrical characteristics curve on this basis, and determined the power factor and electricity efficiency under certain working voltage and electric flow.(4) Combining the research of flow and heat transfer of the molten steel during LF refining process with the analysis of production condition and the results of data statistic, the real-time prediction model for molten steel temperature was established during LF refining process, and the online operation of the model was also achieved. Through field tracing and verification to the model, the result shows that:the fuzzy judgment method can precisely tell the lining grade of the online ladle, and improve the prediction accuracy; when the model correction function was not used, the molten steel temperature predicated by the model is almost close to the real temperature and the hit rate has achieved80%and92%when the relative error within±5℃and±10℃; when the model correction function was used, the hit rate were90%and95%when the relative error within±5℃and±10℃. The prediction value by the model has a good agreement with the actual value, and has a high stability. It can meet the requirements of production control, and realize the high efficiency and automation of temperature control during the LF refining process.