Development And Online Applications Of Deoxidation And Alloying Model For Ladle Furnace Refining

The demand for iron and steel industry automation and high quality steel, has been increasing continuously with the development of industries and the process of technologies. Since the secondary refining is one of the key steps for steel producing, a reasonable control of molten steel components during refining process is an effective meathod of ensuring the qsality of steel product, and reducing production costs and labor intensity. In order to achieve metallurgical automation, it is becoming important to develope deoxidation and alloying model. Therefore, in the present work, the model of deoxidation and alloying was established according to the condition of135t LF in some steel plant. This model has been debugged and validated in the actual production, and shown to achieve automatic control of oxygen and alloying. The main contents and conclusions were drawn as follows:(1) Through the industrial experiment, the average utilization of aluminum wire and aluminum particles were98.36%,61.7%. The utilization of aluminum particles related to the slag-forming practices and oxygen potential. The carbon content increased in electrode heating process of LF refining, and the average raised carbon was0.0291%/(t-h). In LF refining process, sometimes, the manganese could back to steel, and the amount of these manganese related to the initial aluminum content in steel through the equation of△ω[Mn]=0.02327-0.8077·ω[Als]+7.78914·(ω[Als])2.This laid the foundation for developing a reasonable deoxidation and alloying model of the LF.(2) The deoxidizing and alloying model had been established in this paper, that could predict the end aluminum content for desulfurization, then predict the amount of feeding aluminum wire, and this model was established with considering the carbon increasing through electrode heating and the manganese backing to steel in the LF refining process. Considering the price factors, the minimum cost model was established by using simplex method.(3) Through online and on-site inspection, using the experimental data to verified the model, the verification results shows that:the hit rate of the deoxidation model aluminum amount in±10m was80%, the hit rate of aluminum particles in±5kg was86.7%; at alloying model the hit rates of the recovery rates of C, Si, Mn, Ti in±10%were83.2%,71.9%,90.5%and94,2%respectively; in hot metal the hit rate of components of C, Si, Mn in±0,01%were94.2%,95.1%,84.3%respectively, in±0.02%were100%、100%、96.1%respectively; the accuracy of the alloy addition of C, Si, Mn, Tis control error in±5%were59.2%,30.5%,62.8%,52.9%respectively, in±10%were85.7%,67.2%,84.3%,97.3%respectively; the minimum cost model which used FeSi instead of LAlFeSi could reduce the average cost of experimental heats4.6%.(4) The alloy recovery rate was directly related to the accuracy of the model, it need to increase the number of heats in order to improve the accuracy of the recovery rate of alloys.