Importance of Slag on the Decarburisation in the AOD Process

Abstract

The following work concerns the decarburisation period in an AOD, based on studies of aproduction AOD reactor. Today, models for prediction and controlling of the decarburisation are used to obtain a stability in the process and to enhance the productivity. One example ofthe importance of stability is that it leads to the possibility of casting longer sequences which enhances higher productivity. Many efforts have been carried out to develop decarburisation models for the control of decarburisation. However, variations in results still prevail and one of the most important variables is the influence from the top slag on the result.

This work focuses on the top slag during decarburisation in an AOD. The purpose is to lay a ground for future possibilities to include the top slag in a more dynamic way in the existing decarburisation prediction and controlling models. The work is made for a bottom blown AOD converter. Samplings of slags, metals and temperatures were made two times during the decarburisation period for 21 heats. The slag samples were evaluated by using a petrography method, where the samples were-grinded, polished and studied in LOM. The LOM study contained 3780 images in total. All were processed using an image analyse software to ensurethe statistical confidence. Some of the slag samples were also studied in SEM, to clarify the presence of different phasesThe work includes the use of statistical methods, PCA and PLS, to evaluate the decarburisation results based on normal production data, data from the extra samplings, and LOM data. In addition, the work included mass balance calculations for comparisons and evaluations of the sample evaluation results. Furthermore, mass balance calculations were used to evaluate and compare theoretical calculated phase fractions and LOM results.The results from the SEM studies showed the presence three major phases: dicalcium silicate(2CS), calcium chromite (CK) and a metal oxide phase containing Ca and Mg (MeOx).The LOM studies showed the presence of the following phases (not including metal dropletsfound in the slag samples) in a decreasing order based on the fraction (in %): i) calciumchromite (sample 1: 29-72.3 and sample 2: 33.1 – 71.3), ii) dicalcium silicate (sample 1: 22.5– 52.3 and sample 2: 11.3 – 43.8), iii) MeOx (sample 1: 0 – 27.7 and sample 2: 0 – 34.8) andiv) an amorphous phase (sample 1: 2.1 – 9.1 and sample 2: 1.8 – 9.8). Metallographic investigations showed that metal droplets are present in micro sizes in the slag samples. The results from the chemical composition determinations showed that there is a linear relation between the NiO content and the FeO content. Based on basic thermodynamic facts, NiO should not be present in these steel/slag systems. Thus, the results show that there are FeO fractions between 0.8 – 8.0% in the slag that do not originate from the slag itself, butthe metal droplets.The results from the statistical evaluation of the process data focusing on an improved decarburization show the following main indications: i) the 2CS/CK ratio should be high, ii)not too large lime additions should be made at the beginning of the process and iii) the liquid part of the slag should be as low as possible.The overall conclusion is that the results from this study can be used for improvements of process models used to control the decarburization processes in AOD converters. 

Link to DiVA