Characterization of Non-metallic Inclusions and Clusters during Production of Low-carbon IF Steel

Abstract

One of the quality criteria for Interstitial Fee (IF) steels is the metal purity with respect to non-metallic inclusions (NMI), which are harmful for the plastic properties of the material. Furthermore, they cause a formation of surface defects in flat rolled products and reduce the rate of steel casting due to nozzle clogging. This article presents the results of a study of the content, composition, size and morphology of non-metallic inclusions and clusters in steel samples taken during ladle treatment, casting as well as from slabs and steel sheets after rolling of IF steel. The characteristics of NMI and clusters were determined by using conventional two-dimensional quantitative metallographic investigations of polished sections of steel samples (2D method), electrolytic extraction (EE method) of samples followed by investigations of inclusions and clusters by using scanning electron microscopy and energy dispersive spectroscopy and fractional gas analysis (FGA method). By using EE method, different types of inclusions and clusters, their formation, growth and behavior during different stages of IF steel production were studied. The results obtained by the EE method agreed well with the results of the quantitative determination of oxide NMI by using the FGA method. The method of fractional gas analysis shows the dynamics of changes in the content of various types of oxide non-metallic inclusions during ladle treatment and casting of steel. The obtained results can be used to analyze the causes of the formation of harmful NMI in the metal and to optimize ladle treatment of IF steel grades.

1. Introduction

The development of modern materials for the automotive industry makes it possible to produce lighter bodies without a loss of structural strength. This became possible after the invention and the subsequent introduction of Interstitial Fee (IF) steel grades, which do not contain of interstitial atoms in the metal matrix. The IF steel production technology was developed in Japan in the early 1990s, making it possible to produce high-quality steels with ultra-low concentrations of carbon (< 30 ppm), nitrogen (< 30 ppm), as well as and oxygen (< 10 ppm) to ensure a high degree of purity from non-metallic inclusions.^1,2,3,4,5,6)

One of the quality criteria for IF class aluminum-killed steels is the metal’s purity with respect to non-metallic inclusions (NMI), which have a negative impact on the plastic properties of the material as well as cause a formation of surface defects in flat rolled products.^7,8) Furthermore, they reduce the manufacturing properties due to a decrease in the steel pouring speed, as they clog the steel pouring nozzles.^9,10,11)

Many researchers^3,8,12) consider that the most harmful non-metallic inclusions are clusters of aluminum oxides with sizes >100 μm, since they have the largest negative impact on the final IF steels products – steel sheet. Other researchers claim that the titanium oxides inclusions, which arise during the melting process of IF steels modified by Ti, can aggregate and form clusters of inclusions.^13,14,15) Moreover non-metallic inclusions in low carbon Al killed and low Si mild steel grades of steel have a significant influence on the submerged entry nozzle clogging during continuous casting processes.^16,17)

Since non-metallic inclusions often negatively affect the quality of steel and the production, it is necessary to understand how non-metallic inclusions in steel are formed, modified and removed during the ladle treatment of steel. Various methods are normally used to study non-metallic inclusions in steels.¹⁸⁾ However, most methods are “autonomous” - not fast, which require a large amount of time spent on the preparation and analysis of samples. Therefore, one important task in steelmaking industry is development and improvement of online analytical methods for evaluation of NMI in steel samples, which allow online correction of technological processes of steel production.

The method of optical emission spectroscopy with pulse discrimination (OES/PDA) developed in the last decade, which was used for determination the content and composition of non-metallic inclusions in ultra-low carbon (ULC) steels,⁸⁾ is a fast method. A significant drawback of the OES/PDA method is the relatively small volume of the analyzed metal and the lack of certified methods and standard samples. The advantages of the fractional gas analysis (FGA) method are that it is fast and gives representative results.^19,20) The method was tested in combination with the method of electrolytic extraction of non-metallic inclusions, followed by analysis of the inclusions using a scanning electron microscopy (SEM). It was previously shown for the model alloy²¹⁾ that the peaks of oxygen in various oxides in steel samples obtained by the FGA method are in good agreement with the different types of inclusions observed in 3D, when using SEM in combination with energy dispersive spectroscopy (EDS) after electrolytic extraction.

One of the reasons for carrying out this study was the fact of frequent clogging of steel casting nozzles in a series of IF-steel melts at a metallurgical plant and the need to find the causes of this phenomenon. The aims of this study were following: i) to develop a new approach for quantitative and qualitative determination of non-metallic inclusions using a combination of EE and FGA methods, ii) to determine the characteristics (such as composition, morphology, size ranges and amount) of non-metallic inclusions in steel samples and iii) to study the dynamics of formation, modification and removing of different non-metallic inclusions in a steel melt at all stages of the production of ULC IF steels.

2. Experimental

2.1. Materials and Procedures

This study is focused on assessment of non-metallic inclusions in metal samples taken during different stages of production of low-carbon IF steels. A typical composition of IF steel investigated in this study is as follows (in mass%): 0.003–0.006% C, 0.004–0.037% Si, 0.10–0.60% Mn, 0.004–0.010% S, 0.030–0.064% Al, 0.003–0.007% N, 0.013–0.058% Ti.^8,13,22)

The technological process of IF steel production includes the following steps: 1) use of desulfurized hot iron ([S] < 0.005%) as a charge for converter (~380 tons); 2) steel smelting in an oxygen converter; 3) RH vacuum degassing of the steel melt; 4) ladle treatment of the steel in a ladle furnace (LF) including aluminum killing, alloying by addition of titanium and niobium, and heating of the melt up to required temperature in 380-tons ladle (with a temperature range of +/− 5°C from the target casting temperature); 5) continuous casting of steel. A schematic illustration of main technological operations and metal sampling occasions is shown in Fig. 1.

Fig. 1.

Main technological operations and metal sampling during production of low-carbon IF steel.

2.2. Investigation of Non-metallic Inclusions

The selected metal samples were studied by using quantitative metallographic investigations of polished sections with an optical microscope, fractional gas analysis (FGA) and electrolytic extraction (EE) of metal specimens followed by investigations of typical non-metallic inclusions by using scanning electron microscopy and energy dispersive spectroscopy.

The quantitative metallographic analysis of NMI on polished sections of metal samples by using optical microscopy or scanning electron microscopy is the most common method of two dimensional (2D) investigations of non-metallic inclusions in steels. This method can be used to evaluate the number, type, elemental compositions, linear dimensions and the volume fraction of different NMI. However, this 2D method is rather labor-intensive. Moreover, the obtained results are strongly influenced by the quality of the polished section of the metal sample and by some limitations of 2D investigations of sections of inclusions on metal surface.

A combination of the FGA and EE methods for sample investigation allows us to quantify the dynamics of the formation, modification and removal of different NMI at each stage of IF steel production.

2.2.1. Electrolytic Extraction and Scanning Electron Microscopy

Unfortunately, the conventional 2D investigations of NMI on polished cross sections of metal samples do not give accurate data of the inclusion size, morphology and number. Especially it is clear by investigations of inclusions having complicate morphologies and clusters, since only one random cross section of inclusion or cluster can be observed on the polished surface of the metal sample.^23,24,25,26) The use of an electrolytic extraction technique allows to dissolve the metal matrix and investigate the undissolved non-metallic inclusions as three dimensional (3D) objects on filter surfaces after EE and filtration by using a scanning electron microscope (SEM). Therefore, the EE method was applied in this study for 3D investigations of the NMI characteristics (such as morphology, composition, size and number) in different steel samples.

The selected steel samples were extracted followed by investigations of different non-metallic inclusions at the laboratory of KTH Royal Institute of Technology (Stockholm, Sweden). The metal samples were dissolved in a 10% AA electrolyte (10 v/v% acetylacetone-1 w/v% tetramethylammonium chloride-methanol) by using the following electrical parameters: a 50–60 mA current, a 2.5–3.5 V voltage, and 500 or 1000 coulomb electric charges. The metal weight dissolved during electrolytic extraction of steel samples ranged from 0.15 to 0.29 g. After a completed extraction, the electrolyte containing undissolved NMI was filtered by using a polycarbonate membrane film filter with an open pore size 0.4 μm. Then, different non-metallic inclusions were examined using SEM in combination with energy dispersive spectroscopy (EDS). Figure 2 shows a typical SEM image of different NMI and clusters observed on the filter after EE as well as size measurement (L and W are the maximum length and width) of observed clusters. In addition, the equivalent sizes (d_eq = (L + W)/2) and aspect ratios (AR = L/W) were calculated for each measured inclusion and cluster.

Fig. 2.

Typical SEM image of different NMI observed on the filter after EE (a) and size measurement of observed cluster (b).

2.2.2. Fractional Gas Analysis of Oxide Inclusions

The fractional gas analysis method can be applied to determine the total oxygen content in metal samples, the amount of oxygen contained in different types of NMI, and the volume fractions of different types of oxide inclusions. The FGA is a modification of the method of oxygen analysis including a melting of metal sample in a graphite crucible, followed by heating of a sample at a given linear rate and a reduction of different oxide inclusions in a carrier gas current. This analytical method is based on the difference of the thermodynamic strength of various oxides in steel melt depending on the temperature. When the temperature of a sample melt increases continuously, different oxides are reduced by carbon from the graphite crucible at different temperature intervals. As a result, the oxygen is extracted from oxide inclusions of the melt in the form of carbon monoxide and the gas analyzer captures the sample’s gas evolution curve, depending on the temperature of the melt. The volume fractions of various types of oxide inclusions in the analyzed steel sample are calculated by using the OxSeP-Pro software and the sample chemical composition data. The FGA method can be used as an express tool, which remits to evaluate the contents of various types of oxide inclusions in the metal samples during a short time (10–15 minutes).^27,28,29)

In this study, the FGA method was used to examine selected metal samples (Samples 1–4), which are the same as were used for electrolytic extractions. The fractional gas analysis of all samples was carried out on a LECO TC-600 gas analyzer in Baikov Institute of Metallurgy and Material Science (IMET RAS, Moscow, Russia) under the following fixed parameters of heat treatment: preliminary preheating time of sample - 2 minutes, preheating temperature of sample - 1150°C, heating range - 1200–2400°C, heating rate - 2°C/s. For the FGA evaluation, three specimens were cut out from each metal sample and tested. The weight of each metal specimen was varied from 1.2 to 1.6 g.

The volume fraction of NMI in steel samples by using optical microscopy is calculated according to the Cavalieri-Akken principle, as the ratio of the total area of the detected inclusion (S_NMI) to the analyzed area of the polished section (S_PS). Since the FGA can quantitatively determine the oxygen content in inclusions of each type, the volume fraction of oxide inclusions can be calculated with a higher accuracy compared to metallographic methods by using the next equation:   

$$V_{NMI}=\frac{S_{NMI}}{S_{PS}}=\frac{p_{matrix}}{100}\cdot {\sum }_{i=1}^n\left(\frac{O_{oxide}}{p_{oxide}}\cdot \frac{M_{oxide}}{y\cdot M_O}\right)$$(1)

where p_matrix and p_oxide are the densities of the metallic matrix and oxides of the given composition, respectively; O_oxide is the FGA-determined content of oxygen (wt%) bound in the oxide of the n-th type; M_oxide and M_O are the molecular masses of oxide and atomic mass of oxygen, respectively; y is stoichiometric coefficient for oxygen atom in the given oxide.

3. Results and Discussion

The results obtained by metallographic investigations of polished sections of steel samples by using optical microscopy can only show the total contamination of the sample with respect to non-metallic inclusions. Four main types of non-metallic inclusions can be distinguished for IF steel (Fig. 3). However, it is not known for sure, that the “clusters” observed in the Figs. 3(c) and 3(d) represent real clusters of agglomerated oxide and nitride inclusions or accumulations of separate small inclusions, which are only looking as large clusters. Therefore, the detailed characteristics of non-metallic inclusions and clusters were determined by using the EE method.

Fig. 3.

Typical inclusions observed in steel sheet sample: a) oxide inclusions; b) nitride inclusions; c) oxide cluster and d) nitride cluster.

3.1. Characterization of NMI After Electrolytic Extraction

According to results obtained from 3D investigations of NMI on film filter after electrolytic extractions, different non-metallic inclusions in steel samples were found. They were divided in to five main types depending on their morphologies and compositions. Morphology, composition and dimensions of different types of inclusions and clusters in steel samples after electrochemical dissolution are given in Table 1.

Table 1. Morphologies, compositions and typical dimensions of non-metallic inclusions observed on a filter after an electrolytic extraction. (Online version in color.)

Type I inclusions having a spherical morphology were found mostly in Samples 1 to 3. Sample 1 contained spherical NMI of pure Al₂O₃ and Al₂O₃–SiO₂ (up to 5.3% of SiO₂). The size of most spherical inclusions in Sample 1 varied in the range from 1 to 3.5 μm. In Samples 2 and 3, taken after an addition of Ti and Nb, most of the spherical NMI consisted of complex Al₂O₃–TiO_x (25–42% TiO_x) and Al₂O₃–TiO_x–MgO (1–14% TiO_x and 1–11% MgO) inclusions. The content of SiO₂ in these spherical inclusions varied from 0 to 2.6%. Most of the spherical inclusions in Samples 2 and 3 have size in the range from 2 to 8.5 μm. Furthermore, the average size of Type I inclusions tends to increase during ladle treatment and casting, as shown in Fig. 4(a). Overall, it was found that all spherical inclusions are transformed during casting and solidification of steel into regular and irregular inclusions and clusters of Types III and IV, as shown in Table 2.

Fig. 4.

Average sizes of Type I spherical inclusions (a) and Type III regular and irregular inclusions (b) in different steel samples.

Table 2. Transformation of spherical inclusions of Type I during casting and solidification of IF steels into regular and irregular inclusions and clusters of Types III and IV. (Online version in color.)

Plate-like (or flake-like) inclusions of Type II, which are almost pure Al₂O₃, were found in Samples 1 to 3. However, they were not observed in the slab and steel sheet samples (Samples 4 and 5). Moreover, it was found that the size and number of these inclusions decrease during ladle treatment. For instance, the average size of plate-like inclusions decreases considerable from 11.4 ± 2.3 μm in Sample 1 to 8.6 ± 2.6 μm in Sample 3.

Regular and irregular inclusions of Type III consisting of pure Al₂O₃ were observed in all metal samples. The size of these non-metallic inclusions increases during ladle treatment from 1.3–4.3 μm in Sample 1 to 3.4–10.0 μm in Sample 3. However, the average size of Type III inclusions in Samples 4 and 5 after casting and solidification of steel are decreases slightly or are similar to those in Sample 3, as shown in Fig. 4(b).

Clusters of oxide NMI (Type IV) were observed in all five metal samples. As illustrated in Fig. 5, the chemical compositions of agglomerated inclusions in these clusters showed a significant variation from sample to sample. While Sample 1 contains only clusters with pure Al₂O₃ inclusions, the chemical composition of NMI in clusters grew more complex in all subsequent samples. More specifically, they contained also Al₂O₃–TiO_x, Al₂O₃–TiO_x–MgO and (Al₂O₃–MgO–TiO_x) + (Ti,Nb)N,C + MnS. Moreover, some inclusions in clusters contained up to 3% of SiO₂, as can be seen in Fig. 5(c). Sample 3 had the largest number of clusters having a high content of MgO (6–14%), while the number of such clusters in Samples 4 and 5 decreased. It should be pointed out that the average length of measured clusters decreases significantly from 14.3 and 14.7 μm in Samples 1 and 2 to 10.8 and 6.5 μm in Samples 4 and 5 respectively, as shown in Fig. 6(a). The largest clusters (up to 51 μm) were found in Sample 2 after a completed ladle treatment. It can be explained by intensive formation and growth of clusters by agglomeration of oxide inclusions after an Al addition (Sample 1) and during ladle treatment (Sample 2). The following decrease of the size and number of clusters is due to flotation of large size clusters from the melt during casting of liquid steel (Samples 3 and 4) and due to the following destruction of clusters during rolling (Sample 5).

Fig. 5.

Contents of TiO_x (a), MgO (b) and SiO₂ (c) in inclusions of Type IV clusters observed in different steel samples.

Fig. 6.

Average size of Type IV clusters (a) and Type V inclusions and clusters (b) observed in different steel samples.

The morphology and compositions of typical clusters observed in different steel samples after electrolytic extraction are shown in Table 3. It can be seen that the pure Al₂O₃ inclusions in clusters grow and obtain a more regular shape (with flat faces) during ladle treatment (Samples 2 and 3), which is followed by a precipitation of (Ti,Nb)N,C layer on Al₂O₃ inclusions. However, although the complex oxides in clusters of Samples 2 and 3 (such as Al₂O₃–TiO_x and Al₂O₃–TiO_x–MgO) have smooth surfaces, they obtain more regular/irregular shapes after the precipitation of TiN or (Ti,Nb)N,C inclusions (Samples 4 and 5).

Table 3. Morphologies and compositions of typical Type IV clusters observed in different steel samples after a completed electrolytic extraction. (Online version in color.)

The nitride inclusions and clusters of Type V contained mostly Ti and small amounts of Nb and C were found in Samples 2–5. However, it should be pointed out that that the oxide inclusions were also often present as a core in these NMI. The average size of nitride inclusions in Samples 2 and 3 taken from the liquid steel is about 1 μm due to fast solidification of these samples, as shown in Fig. 6(b). An increase of the nitride sizes up to 3–4.5 μm in Sample 5 can be explained by a heterogeneous precipitation on oxides and a significant growth during solidification of casted steel in slabs and during heat treatment before rolling.

Overall, a precise determination of volume fractions for each type of inclusions and clusters in this steel by using the electrolytic extraction method is limited due to large variations of the compositions of inclusions and due to limitations with respect to the volume of analyzed metal sample. Therefore, the FGA technique was used also in this study as a complimentary method to study inclusions.

3.2. Characterization of NMI by using FGA

Depending on the temperature (T_m) of the start of reduction of different oxides in a carbon-saturated metal melt, the peaks on the obtained gas evolution curves were divided into four main groups. These correspond to specific compositions of oxide inclusions, as given in Table 4.

Table 4. Classification of oxide inclusions in metal samples depending on the reduction temperature.

Group	Oxide composition	Tm, K	Tm, °C
1	Sample 1: FeO–SiO2–Al2O3 Samples 2–4: SiO2–TiO2–Al2O3	1700–1785	1427–1512
2	TiOx–Al2O3	1785–1900	1512–1627
3	Al2O3	1900–2010	1627–1737
4	Al2O3–MgO–TiOx	2010–2100	1737–1827

It was found that the results for oxide inclusions and clusters in metal samples obtained from the EE method agreed well with those from the FGA method. Figure 7 shows the typical time chart of peaks on gas evolution curves during FGA determinations of metal samples and the corresponding oxide inclusions and clusters observed after electrolytic extraction. However, it should be noted that the inclusions and clusters of different types, which were classified based on morphology observed NMI after EE, can be present in one peak group of oxides according to composition obtained by FGA analysis and opposite. It can be seen that the Groups 1 and 2 contain mostly globular inclusions of Type I (SiO₂–Al₂O₃, SiO₂–TiO₂–Al₂O₃ and TiO_x–Al₂O₃) having lower melting temperatures and lower reduction temperatures. The pure Al₂O₃ inclusions and clusters of Types I-IV correspond to one peak of Group 3. The complex Al₂O₃–TiO_x–MgO inclusions and clusters of Types I, IV and V are consisted in Group 4. According to previous publications, the large size clusters and inclusions of Groups 3 and 4 are more harmful for the final properties of IF steel compared to those of Groups 1 and 2. Therefore, the formation, growth and behavior of clusters and inclusions of Groups 3 and 4 are most interesting and are the main focus to study more in depth.

Fig. 7.

Typical time chart of peaks on gas evolution curves during FGA determinations of metal samples and the corresponding oxide inclusions and clusters observed after electrolytic extraction.

The FGA peaks, which correspond to the oxygen contents in various types of NMI, make it possible to evaluate the volume fractions of corresponding oxide inclusions and clusters in steel samples on different stages of steelmaking. Figure 8 shows the FGA results of the oxygen contents in various oxide inclusions in the steel samples obtained during IF steel production. The oxygen content in the samples in this figure corresponds to the volume fractions of the corresponding groups of oxide inclusions in each of the samples being studied. Based on the FGA data, the changes of compositions and volume fractions of different oxide NMI can be evaluated in steel samples taken during different stages of steelmaking. The obtained results can be used for optimization of the ladle treatment process during production of IF-steels, as has been shown for pipeline steels.³⁰⁾

Fig. 8.

Oxygen contents in various oxide inclusions in steel samples obtained by using FGA analysis.

It can be seen in Fig. 8 that the pure aluminates (Group 3, ~55%) and Al₂O₃–SiO₂ inclusions (Group 1, ~26%) are the main types of NMI found in Sample 1 (aluminum was added into the liquid steel in ladle before sampling). Then, after a Ti addition during ladle treatment, the amount of pure Al₂O₃ and Al₂O₃–SiO₂ inclusions in Sample 2 decreases significantly up to 35% and 6%, respectively. At the same time, the amount of TiO_x–Al₂O₃ inclusions (Group 2) increases up to ~56%. However, after all additions of Al, Ti and other ferroalloys during ladle treatment, the content of O in all oxide inclusions increased from 65 ppm in Sample 1 up to 76 ppm in Sample 2. The total content of oxide NMI in Sample 3, taken from the tundish, decreases to 33 ppm (by more than 30 ppm) and the majority of the NMI corresponds to TiO_x–Al₂O₃ (~44%) and pure Al₂O₃ (27%) inclusions. It can be seen in Fig. 8 that non-metallic TiO_x–Al₂O₃ inclusions (Group 2) are intensively removed from the melt during the stage of metal transfer from the ladle to the tundish (a 66% decrease from Sample 2 to Sample 3) as well as at the continuous casting stage (a 52% decrease from Sample 3 to Sample 4). Moreover, it is interesting to point out that the removal of the TiO_x–Al₂O₃ inclusions after ladle treatment is significantly faster compared to the removal of Al₂O₃ inclusions (Group 3). For instance, the O content in the TiO_x–Al₂O₃ inclusions decreases from ~44 ppm in Sample 2 to 15 and 7 ppm in Samples 3 and 4, whereas the corresponding values for the Al₂O₃ inclusions and clusters decrease from ~27 ppm (Sample 1) to 9 (Sample 3) and ~5 ppm (Sample 4). At the same time, the number of inclusions of Groups 1 (SiO₂–TiO_x–Al₂O₃) and 4 (Al₂O₃–MgO–TiO_x) did not show clear tendency to decrease. It was found that the amount of complex inclusions and clusters of Al₂O₃–MgO–TiO_x and MgO–Al₂O₃ (Group 4) in the analyzed samples was small. Specifically, it varied from 3 up to 10% depending on the production stage. Thus, the modification and removing of inclusions in the liquid steel during ladle treatment and casting promoted a decrease of the O contents of the most harmful NMI and clusters of Groups 3 and 4 from ~38 ppm in Sample 1 up to 6 ppm in Sample 4 (taken from steel slab). Herewith the total oxygen content in all oxide inclusions in steel slab was 18 ppm. The decreased oxygen content in non-metallic inclusions indicates that the metal was significantly refined from NMI and clusters during ladle refining. Moreover, the most significant purification from non-metallic inclusions occurred in the tundish and mold of the continuous casting machine. One of the reasons for the intensive removal of inclusions in the casting device compared to the steel ladle is the difference in the height of the metal level in the casting device (1.2 m) and in the steel ladle (4.6 m). It was also assumed that a part of non-metallic inclusions caused clogging of steel casting nozzles, and as a result, a difference was observed between the content of inclusions in Samples 2 and 3 (at the ladle – furnace – intermediate ladle stage) and Samples 3 and 4 (intermediate ladle – ingot). The analysis shows the ratios of non-metallic inclusions to 4 main groups, which gives a complete picture of their modification and removal during refining and continuous casting.

3.3. Comparison of Different Methods for Characterization of NMI

Based on previous publications^{23,24,25,26,31)} and on results obtained in this study, it was found that 3D investigations of the actual morphologies, compositions and sizes of NMI in steel samples after electrolytic extraction by using SEM (EE + SEM method) are more preferable and precise compared to conventional 2D investigations (LOM, SEM, automated SEM+INCA Feature and SEM + ASPEX methods) of NMI on polished metal surface. This is especially true for investigations of clusters and non-metallic inclusions having complicate morphologies. Moreover, the EE method can be used for detailed investigations of mechanisms of formation, growth, agglomeration and modifications of NMI during different stages of steel production.

However, as was reported in previous studies,^32,33,34) the non-metallic inclusions in the metal samples (such as lollipop samples), which were taken from the liquid steel, cannot be distributed homogeneously due to the fluid flow in the sampler mold during sampling and solidification of steel. Therefore, in order to obtain representative estimation of actual NMI in metal samples, it is necessary to increase the analyzed steel volume and to investigate the NMI in different zones of the metal samples. A comparison of the analyzed volume (V_anal.) and weight (W_anal.) of metal sample by using different analytical techniques is given in Table 5. It can be seen that due to some limitation of analyzed volume of metal sample when using the EE method, other analytical techniques (such as the FGA method) should be used for precise evaluations of the amount of different NMI (such as the mass percentage and volume fraction) in steel samples. In this study the V_anal. and W_anal. values of one metal specimen analyzed by using the FGA method is about 2000–10000 larger than that by using the EE+SEM method. As a result, it is believed that the amount of different oxide inclusions in steel samples can be determined more precisely and faster by using the FGA method. However, the morphology, size and compositions of separate inclusions and clusters cannot be determined by using the FGA method.

Table 5. Comparison of the analyzed volume (V_anal.) and weight (W_anal.) of metal samples by using different analytical techniques.

Method	Aobs. (mm2)	Wanal. (g)	Vanal. (mm3)	Comments	Ref.
LOM and SEM (2D)	0.2–1.5	~(0.1–0.2) ×10−4	(4–30) ×10−4		23
SEM + ASPEX (2D)	106–156	(25–61) ×10−4	0.32–0.78	~50 mm2/h	31
OES/PDA (3D)	~7–20	~2×10−4	~0.025	at 4000 sparks, ~5×10−8 g/spark	34
EE + SEM (3D)	1–3	1–7×10−4	0.016–0.093	Wdis.~ 0.15–0.29 g	Ps*
FGA (3D)		1.2–1.6	154–205		Ps*

*:  Ps – present study.

Overall, a combination of these two techniques (EE+SEM and FGA) can provide information on the actual characteristics of different non-metallic inclusions and clusters, which can be used for study of the formation, behavior and removing of harmful NMI and for optimization of parameters of ladle treatment.

4. Conclusions

Non-metallic inclusions and clusters were studied in metal samples taken during ladle treatment, casting and from slabs and rolled sheets of IF low carbon steel. The inclusion characteristics (such as morphology, composition, size, number and volume fractions) were investigated by using conventional two-dimensional metallographic investigations of polished sections of steel samples (2D method), electrolytic extraction (EE method) of samples with followed investigations of NMI by using SEM and EDS, and fractional gas analysis (FGA method).

Based on the results obtained EE method, five main types of non-metallic inclusions and clusters were detected in the metal samples. Moreover, their formation, growth and behavior during different stages of IF steel production were studied. The results obtained by using the EE method agreed well with the results of the quantitative determination of oxide NMI by using the FGA method.

The method of fractional gas analysis shows the dynamics of changes in the content of various types of oxide non-metallic inclusions during ladle treatment and casting of steel. It is shown that the most significant refining of non-metallic inclusions occurred in the tundish and mold of the continuous casting machine. About half of the total number of non-metallic inclusions and clusters in steel melt is removed in the tundish. It was found that TiO_x–Al₂O₃ inclusions are intensively removed from the melt during the metal transfer from the ladle to the tundish as well as at the continuous casting. Moreover, the removal of the TiO_x–Al₂O₃ inclusions after ladle treatment is significantly faster compared to the removal of Al₂O₃ inclusions. Thus, it has been shown that the tundish of a continuous casting machine is a refining unit for secondary metallurgy.

The results obtained by using combination of the EE and FGA methods can be used to analyze the causes of the formation, modification and removing of harmful NMI and clusters in the steel melt, to carry out corrective operations to optimize the ladle treatment of IF steel grades and prevent a nozzle clogging during continuous casting.

References

• 1)   H.  Kimura: Nippon Steel Tech. Rep., 61 (1994), 65, https://www.nipponsteel.com/en/tech/report/nsc/pdf/6110.pdf, (accessed 2020-03-30).
• 2)   S. K.  Paul,  A.  Raj,  P.  Biswas,  G.  Manikandan and  R. K.  Verma: Mater. Des., 57 (2014), 211. https://doi.org/10.1016/j.matdes.2013.12.047
• 3)   T. K.  Roy and  T.  Venugopalan: Assessment of Cleanliness in Interstitial Free Steel, Tata Steel, http://www.iim-delhi.com/upload_events/01AssessmentCleanlinessInterstitialFreeSteel_TataSteel.pdf, (accessed 2020-03-30).
• 4)   J.  Chen,  Y.  Kan,  Y.  Hao,  G.  Liu and  A.  Xiong: J. Iron Steel Res. Int., 16 (2009), 33. https://doi.org/10.1016/S1006-706X(10)60024-6
• 5)   H.  Takechi:  IF  Steels 2003: Int. Forum for the Properties and Application of IF Steels, Iron and Steel Institute of Japan, Tokyo, (2003), 440.
• 6)   T.  Uetani and  N.  Bessho: Kawasaki Steel Tech. Rep., 44 (2001), 43.
• 7)   P.  Ghosh,  C.  Ghosh and  R. K.  Ray: ISIJ Int., 49 (2009), 1080. https://doi.org/10.2355/isijinternational.49.1080
• 8)   M. M.  Pande,  M.  Guo,  R.  Dumarey,  S.  Devisscher and  B.  Blanpain: ISIJ Int., 51 (2011), 1778. https://doi.org/10.2355/isijinternational.51.1778
• 9)   R.  Dekkers: Ph.D. Thesis, Katholieke Universiteit Leuven, (2002), https://www.mtm.kuleuven.be/onderzoek/thermo/publications/doctoral-theses/dekkers, (accessed 2020-03-30).
• 10)   J.  Ikäheimonen,  K.  Leiviskä,  J.  Ruuska and  J.  Matkala: IFAC Proc. Vol., 35 (2002), 143. https://doi.org/10.3182/20020721-6-ES-1901.01175
• 11)   F.  Tehovnik,  J.  Burja,  B.  Arh and  M.  Knap: Metalurgija, 54 (2015), 371.
• 12)   L.  Zhang and  B. G.  Thomas: XXIV National Steelmaking Symp., The Morelia Institute of Technology, Morelia, (2003), 138, http://ccc.illinois.edu/PDF%20Files/Publications/03_Mexico_Nov_Inclusion_review_v5a_updated.pdf, (accessed 2020-03-30).
• 13)   M.  Kiviö and  L.  Holappa: Metall. Mater. Trans. B, 43 (2012), 233. https://doi.org/10.1007/s11663-011-9603-5
• 14)   C.  Wang,  N.  Verma,  Y.  Kwon,  W.  Tiekink,  N.  Kikuchi and  S.  Sridhar: ISIJ Int., 51 (2011), 375. https://doi.org/10.2355/isijinternational.51.375
• 15)   C.  Wang,  N. T.  Nuhfer and  S.  Sridhar: Metall. Mater. Trans. B, 40 (2009), 1022. https://doi.org/10.1007/s11663-009-9290-7
• 16)   B.  Harcsik,  P.  Tardy and  G.  Karoly: Rev. Métall., 109 (2012), 177. https://doi.org/10.1051/metal/2012046
• 17)   N. U.  Girase,  S.  Basu and  S. K.  Choudhary: Ironmaking Steelmaking, 34 (2007), 506. https://doi.org/10.1179/174328107X168075
• 18)   J.  Guo,  S.  Cheng,  H.  Guo and  Y.  Mei: Sci. Rep., 9 (2019), 2929. https://doi.org/10.1038/s41598-018-36766-6
• 19)   K. V.  Grigorovitch,  A. M.  Katsnelson,  A. S.  Krylov and  A. V.  Vvedenskii: Proc. 4th Int. Conf. Analytical Chemistry in the Steel and Metal Industries, Office for Official Publications of the European Communities, Luxembourg, (1996), 527.
• 20)   K. V.  Grigorovitch,  P. V.  Krasovskii and  A. S.  Krylov: Proc. 50th Chemists Conf., British Steel, North Lincolnshire, (1999), 129.
• 21)   A. V.  Karasev,  D.  Gorkusha,  K. V.  Grigorovich and  P. G.  Jönsson: Proc. 10th Int. Conf. on CLEAN STEEL, The Hungarian Mining and Metallurgical Society, Budapest, (2018), USB memory.
• 22)   V. A.  Bigeev and  A. O.  Nikolaev: Aktual’nye Problemy Sovremennoi Nauki, Tekh. I Obraz., 1 (2012), 130 (in Russian).
• 23)   A. V.  Karasev and  H.  Suito: Metall. Mater. Trans. B, 30 (1999), 259.
• 24)   H.  Ohta and  H.  Suito: ISIJ Int., 46 (2006), 14. https://doi.org/10.2355/isijinternational.46.14
• 25)   Y.  Kanbe,  A.  Karasev,  H.  Todoroki and  P. G.  Jönsson: ISIJ Int., 51 (2011), 593.
• 26)   A.  Karasev and  P. G.  Jönsson: Proc. 5th Int. Conf. on Process Development in Iron and Steelmaking (SCANMET-V), Swerea/MEFOS, Luleå, (2016), USB memory.
• 27)   P. V.  Krasovskii and  K. V.  Grigorovich: Russ. Metall. (Metally), 2002 (2002), 108, https://www.researchgate.net/publication/286940719_Thermodynamics_of_nonisothermal_reduction_of_oxide_inclusions_in_carbon-saturated_melts, (accessed 2020-03-30).
• 28)   P. V.  Krasovskii,  W.  Gruner and  K. V.  Grigorovitch: Steel Res. Int., 77 (2006), 64. https://doi.org/10.1002/srin.200606133
• 29)   K.  Grigorovich,  O.  Komolova and  D.  Gorkusha: Proc. 5th Int. Conf. on Process Development in Iron and Steelmaking (SCANMET-V), Swerea/MEFOS, Luleå, (2016), USB memory.
• 30)   K.  Grigorovich,  O.  Komolova and  D.  Terebikina: J. Chem. Technol. Metall., 50 (2015), 574.
• 31)   M.  Nabeel,  M.  Alba,  A.  Karasev,  P. G.  Jönsson and  N.  Dogan: Metall. Mater. Trans. B, 50 (2019), 1674. https://doi.org/10.1007/s11663-019-01605-0
• 32)   O. T.  Ericsson,  M.  Lionet,  A. V.  Karasev,  R.  Inoue and  P. G.  Jönsson: Ironmaking Steelmaking, 39 (2012), 67.
• 33)   Z.  Zhang,  A.  Tilliander,  A.  Karasev and  P. G.  Jönsson: ISIJ Int., 50 (2010), 1746.
• 34)   D.  Janis,  A.  Karasev and  P. G.  Jönsson: ISIJ Int., 55 (2015), 2173.