Inclusions in Steel and Secondary Steelmaking

Inclusions in Steel and Secondary Steelmaking

Non-metallic inclusions, (hereinafter referred to as ‘inclusions’), are chemical compounds consisting of at least one non-metallic component, such as sulphur or oxygen. In steels, inclusions are an unwanted but mainly an unavoidable phase caused by the content of oxidizing agents in the liquid steel due to the raw material or introduced during steelmaking. Inclusions are formed into various types depending on their favourable thermodynamic conditions in almost all treatment practices involving liquid steels.

Harmful effects of inclusions are highly dependent on their chemical compositions, volume fractions, dispersions, and morphologies. Normally large and unbreakable inclusions with high melting points are the most unwanted ones. However compared to these inclusions, small and breakable ones or those with lower melting points are more preferred. The reason for these preferences is that inclusions which have lower melting points or are breakable are likely to be deformed, crushed to smaller inclusions, or disappeared in following hot or cold forming processes (effect of forming process and reduction ratio) or heat treatments which the steels can have after the continuous casting and the solidification process.

The main inclusions in steelmaking are the oxides and the sulphides. They are formed as a result of the steelmaking process and hence form integral part of the steel. Whereas oxides are normally regarded harmful, sulphides plays a tricky role in determining steel properties such as machinability.

Apart from some applications where inclusions are supposed to be demanded, like sulphides for improving machinability, they normally deteriorate mechanical properties and surface quality of steel products and can cause nozzle clogging and disruption of steelmaking and forming processes. It is widely believed that due to the presence of sulphide and oxide inclusions some of the mechanical properties of steels like ductility, toughness, anisotropy, and formability can be negatively affected. The remaining of inclusions inside steel matrix damages the mechanical properties and result in the failure of the material.

Quality requirements demanded for continuously cast steel products are increasingly becoming stringent in recent years to meet customer demand based on needs of enhancing productivity and efficiency through production of flawless products and improvement of processing performance.

Existence of inclusions in liquid steel is in general one of the major factors which causes steelmaking-caused quality defects, and they are not only directly causing product defects but also clogging nozzles, thereby causing operating abnormalities such as drift of liquid steel in a mould and entrapment of powder, thus indirectly causing product defects as well.

‘Steel with high cleanliness’ cannot be expressed in a simple way as requirements of quality characteristics vary depending on their uses, and hence, acceptable sizes of inclusions, compositions, and amounts also vary depending on such steel uses. The term ‘clean steel’ is used with caution. This is because of (i) varying cleanliness demands for steels for different applications, (ii) varying cleanliness in steels produced in different operations, and (iii) the normal understanding of the word ‘clean steel’, which some literally interpret as meaning the absence of inclusions in the steel.

The definition of ‘clean’ is not absolute, but depends on the individual steel production process and its in-service use of the final product. The term ‘clean steel’ is hence variable depending on the steel producer and steel application. Due to the variable nature of the term ‘clean steel’, it is sometimes proposed to talk more accurately of (i) high purity steel as steel in the case of low levels of solutes (sulphur, phosphorus, nitrogen, oxygen, and hydrogen) and (ii) low residual steel as steel with low level of impurities (copper, lead, zinc, nickel, and chromium to name just a few) mostly originated from scrap.

Steel cleanliness has implications both from operational and product performance points of view. The term ‘clean steel’ is normally used to describe steel which has (i) low level of solute elements, (ii) controlled level of residual elements, and (iii) low frequency of oxides created during steel making, ladle metallurgy, casting, and rolling.

Clean steels are steels with a low frequency of product defects which can be correlated to oxide inclusions. In addition, clean steel is increasingly understood as steel for which the composition is under tight control of alloying elements to improve product properties and property consistency. There is one constant in producing high purity, low residual and clean stee , which is the continual drive to reduce solute elements and residuals in the steel and control frequency, distribution, and size of inclusions.

The cleanliness assessment in steel is having a problem. There are only very few large (macro) inclusions, which are difficult to detect for the reason that their number is small. In contrast, the number of very small (micro) inclusions is almost infinitesimal and their size makes them nearly undetectable. It appears that 5 micrometers (0.005 mm) represents the borderline between tolerable micro inclusions and potentially harmful macro inclusions. These sporadic large inclusions represent the foremost quality problem for steel plants in producing clean steel.

Clean steel, in addition to lowering the oxides and sulphides inclusions, and controlling their composition and morphology, needs lowering of other residual impurities such as phosphorus, hydrogen, and nitrogen content and other trace elements in steel. Sometimes the concept of clean steel is argued as a debatable concept. It is since the term clean steel is relative, and it is true to say that the ‘the cleanliness of steel, like beauty, is very much in the eye of the beholder’. Further, as per the argument, the concept leads to the impression that steel with fewer numbers of inclusions are superior in performance which is not always the case.

There are different steel grades produced for different purposes. The cleanliness level of the steel for each purpose depends on the inclusion number, morphology, composition and size distribution of each steel grade. For example, in free machining or resulphurized steel, the design is not to completely remove the inclusions but to modify them to improve machinability. Hence, a balanced opinion regarding permissible level of inclusion or cleanliness for each steel grade is really of great technical and economic importance both to the steelmaker and the user. To a large extent, clean steel is required to meet the customer’s specifications and requirements for an application with regard to the inclusion characteristics.

The presence of inclusions in steel is inseparable from the steelmaking processes. Their presence is frequently regarded as harmful, but sometimes equally advantageous. Whichever the effect, their presence in steel is part of the steelmaking process and is to be exploited to the advantage of the final steel properties.

The chemical composition of the inclusions and their volume fraction are determined by the management of the different steps involved in the production process of steel namely (i) primary steelmaking, (ii) secondary steelmaking, and (iii) continuous casting operation. Hence, the population of the inclusions depends on the relation existing between the applied operative parameters and the features of the steel grades being produced.

Based on the origin, inclusions can be oxidation particle, refractory fragment, top slag entrainment, and reoxidation product etc. A variety of methods are applied to remove the inclusions, such as ladle stirring, slag refining, tundish operation, and continuous casting mould. The removal of inclusions in the continuous casting mould is difficult since the liquid steel becomes solid and inclusions have less opportunity to float out. The removal of inclusions and the final distribution of inclusions in the steel product highly depend on the properties of inclusions, transport of inclusions in the liquid steel and the interaction between inclusions and solidifying shell. Hence, the understanding of the entrapment of inclusions and their final distribution in the steel product are important for the control of cleanliness and the quality of the steel product.

One of the functions of the secondary steelmaking is to remove the inclusions produced through deoxidation of the steel bath during the process. Basically, once incorporated into liquid steel, inclusions undergo three stages in order to be removed. These stages are (i) flotation, (ii) separation, and (iii) dissolution. The first step involves the transportation of the inclusion to the steel / slag interface and during the second, separation, the surface tension of the steel ruptures, allowing the inclusion to stabilize at the steel / slag interface. In the final stage, dissolution, the inclusion’s return to the steel bath is eliminated when it is fully incorporated into the slag. In case the final two steps fail to take place, the inclusion is subject to re-entrainment into the liquid steel bath through entrapment, depending on the flow patterns in the ladle or tundish.

Studies show that the separation phase for solid inclusions takes place very quickly. It was calculated that, when submitted to standard slags in secondary steelmaking, solid inclusions take less than 0.0007 seconds to break the surface tension of steel after flotation. However, it takes a considerable time of upto 7 seconds for the inclusions to reach the steel / slag interface. This time difference is caused by the smaller contact angle between liquid particles and the liquid steel. By contrast, the dissolution time for liquid inclusions can be disregarded since they are miscible in the slag. Hence, the third step in inclusion removal is the most pronounced for solid inclusions. These inclusions have limited solubility in slags and hence are sensitive to the physical and chemical characteristics, temperature gradients, and volume of the slag in question. As such, the removal behaviour of solid inclusions is subject to control by mass transfer, reaction kinetics and chemical interactions with the slag. Dissolution hence becomes the controlling step for the removal of solid inclusions and analyzing the factors in play during this phase is essential to explain the inclusion absorption capacity of slags.

In the secondary steelmaking processes, slag, steel and ladle are involved. Hence the composition of slag, steel, and ladle are very crucial factors for reaching the desired final steel properties, and they have crucial effects on chemical compositions, volume fractions, dispersions, and morphologies of the inclusions. Some of the most important factors which can affect the physico-chemical properties of inclusions are shown in Fig 1.

Fig 1 Factors affecting various characteristics of inclusions during steelmaking

The evolution of the inclusion population over time during ladle treatment can be complex. Given that the effectiveness of inclusion removal by coagulation, bubble attachment, and interface capture is highly dependent upon the composition, phase, and morphology of the inclusion population developed during the ladle treatment practice, it is evident that there are time periods during ladle treatment where rinsing and flotation treatments are more effective than others. The ability to develop ladle treatment strategies to take advantage of these preferred treatment times where inclusion coarsening and removal rates are the most rapid is ultimately dependent upon the cleanliness of post-flotation-treatment additions, control of reoxidation, and slag entrainment during stirring and the effectiveness of reoxidation protection during steel transfer.

The ladle treatment process is a key process for the production of steels with low inclusion content and an inclusion composition and morphology which is not detrimental to the product or its manufacturing process. However, the ladle treatment process also serves several other functions which sometimes can be in conflict with clean steel production. The most common ladle treatment steps include deoxidation of the steel and the slag, desulphurization of the steel, alloying the steel, adjusting the steel temperature for transfer to the continuous casting machine, control of dissolved gases such as nitrogen and hydrogen, inclusion modification, and inclusion removal, as shown in Fig 2.

Fig 2 Steps in ladle treatment of liquid steel

The type of deoxidation practice used in the steelmaking process plays an important role in the type of the inclusions formed during ladle treatment. For many bar and structural products, the steel is deoxidized with silicon and manganese (Si-Mn killed) while for sheet, plate and special bar quality products, the steel is deoxidized with aluminum (Al killed). Some aluminum killed steels are also treated with calcium to modify the alumina inclusions and sulphur containing inclusions to improve the castability of the steel and mechanical properties of the product. The inclusions which evolve during ladle treatment can be fully solid, fully liquid or a mixture of solid and liquid.

Secondary steelmaking technologies such as ladle metallurgy and vacuum degassing considerably control inclusions in steel. The use of ladle metallurgy for controlled slag-metal reactions, micro alloying, and inclusion shape control has resulted into a great improvement in the steelmaking process. Other enhancement achieved in present day clean steelmaking includes improved practice of deoxidation, stirring, vacuum degassing, and electro slag remelting. Also, a more effective and stringent teeming procedure and the use of up-hill teeming with effective teeming shrouds has greatly contributed to the reduction in reoxidation, and hence considerably reduced oxide inclusion in steel.

Normally after finishing the secondary steelmaking processes due to lack of enough slag-steel stirring and with little possibility of adding any extra material, it is not easy to make any changes to the system and hence morphology, volume fraction, composition, and dispersions of non-metallic inclusions at the end of the secondary steelmaking processes cannot be easily changed. However, there are possible small effects of casting powder in the moulds and the tundish lining during the continuous casting and the solidification process. The deoxidation processes used in the secondary steelmaking practices are described below.

Si-Mn deoxidation process – Combination of deoxidizers (silico-manganese, ferro-silicon, and ferro-manganese) is normally used for improved result. Partial deoxidation is typically utilized during the melting process and a final deoxidation process is achieved in the ladle. This practice promotes the formation of low melting-point deoxidation products which is easily removed from the liquid steel. Si-Mn-Si (silicon-manganese) deoxidation frequently form inclusions of solid silica and liquid manganese silicate. There exists critical ratio [%Si] / [%Mn] at a given temperature which influence the type of deoxidation products formed as shown in Fig 3. For a composition to the left side of the curves in Fig 3, solid silica is likely to be formed whereas liquid manganese silicates are likely to be formed to the right of the curve where manganese content is higher. The equilibrium reaction of Si-Mn deoxidation reaction is given by the equation [Si] + 2MnO = 2[Mn] + SiO2.

Fig 3 Equilibrium relation of silicon-manganese deoxidation of steel

Hence, decreasing of the Si / Mn ratio yields liquid manganese silicate inclusion. Liquid manganese silicates inclusions are glassier and plastic in deformation when hot rolled. However, the plasticity index of these type of inclusions decreases and becomes more brittle when cold rolled. When combination of silicon-manganese, and aluminum are utilized for deoxidation, manganese and silicon are charged in a form of ferroalloys during the melting process and aluminum utilized in the ladle for final deoxidation. Inclusions in this system mainly consist of alumina, silica and mullite when the manganese content is low. However, liquid manganese silicates are formed with increasing temperature when the manganese content of the steel is high. One of the studies indicate that, to form liquid inclusions at steelmaking temperatures, the composition area (for lines L1 and L2, and a third line representing 40 % maximum alumina) is limited by SiO2 / MnO ratio of 17-1.9 with respective Si / Mn ratio of 0.47-1.7 in the inclusion content respectively (Fig 4).

Fig 4 MnO-SiO2-Al2O3 ternary phase diagram

Aluminum deoxidation – Aluminum is one of the most effective deoxidizers used in steelmaking. Steel deoxidized with aluminum contains normally alumina inclusions such as corundum. Solid alumina inclusions are reported to cause nozzle clogging during continuous casting of liquid steel. They are more detrimental and undesirable in steel. Alumina inclusions have high melting temperature and remain solid at steelmaking temperatures. In addition, alumina inclusions are hard and are either undeformed or deformed in brittle manner when rolled. The equilibrium reaction of aluminum deoxidation is 2[Al] + 3[O] = Al2O3. Aluminum bars are normally added to the liquid steel for deoxidation.

Calcium treatment of steel – Calcium has a strong affinity for oxygen and hence can be used as deoxidizers. However, the use of calcium as deoxidizer is challenged by its low boiling point of 1,439 deg C, limited solubility of calcium (0.032 %) in steel at 1,600 deg C, and a high vapour pressure of 0.183 MPa at 1,600 deg C. These properties make it difficult and non-economical to use calcium as deoxidizers. However, combinations of Ca and Al or Si / Mn deoxidation form modified primary inclusions with lower activity and melting temperatures. For this reason, in the steelmaking practice normally calcium is added to steel more as an inclusion modifier rather than deoxidizer. Majority of the steel grades are treated with calcium using either Ca-Si alloy or Ca-Fe(Ni) mixture depending on the alloy specification. Normally this treatment is effectively done after trim additions and argon rinsing. Ca-Si alloy wire is normally injected into the liquid steel after the aluminum deoxidation process. Fig 5 shows schematic of inclusion modification by calcium with calcium treatment of steel.

Fig 5 Schematic of inclusion modification with calcium treatment of steel

After effective calcium treatment, all oxide inclusions normally contain some amount of calcium. Effective modification of oxide inclusions in steel depends on the dissolved aluminum and oxygen content of the liquid steel before calcium treatment. For an essential inclusion modification, a calcium lower limit of 15 ppm (parts per million) to 20 ppm is needed. Calcium aluminate inclusions are formed with a CaO-Al2O3 ratio of 12:7, and low melting points of 1,455 deg C of calcium aluminate.These inclusions exist in the liquid state at steelmaking temperatures. Calcium aluminates 12CaO.7Al2O3, 3CaO.Al2O3, and CaO.Al2O3 exist in the liquid state, whereas CaO.2Al2O3 and CaO.6Al2O3 are solid at steelmaking temperatures.

The general effect of calcium treatment on inclusions modifications are summarized as (i) manganese sulphides are reduced in number and size, and they are transformed to Ca-Mn sulphides with varying properties, (ii) aluminum oxides, which are normally hard, angular and frequently appears in clusters are reduced in number or completely eliminated and replaced with complex CaO-Al2O3 or CaO-Al2O3-SiO2 inclusions, (iii) silicates are eliminated and replaced by CaO-Al2O3-SiO2 inclusions, and (iv) complex globular CaO-Al2O3-SiO2 inclusions are formed frequently surrounded by sulphide rim.

Sulphides modification by calcium addition – Sulphur has almost unlimited solubility in liquid steel. However, solubility of sulphur in solid steel approaches zero. During solidification, sulphides precipitate in various forms at the grain boundaries producing characteristic steel defects. The chemical affinity of calcium to oxygen is higher than that of sulphur. It is estimated that for calcium to react with sulphur, the sulphur activity is to be around 19 times higher than the oxygen activity in steel, a condition which is difficult to achieve. Hence, the purpose of introducing calcium for sulphur modification is to change the sulphur release mechanism in a way such that the sulphur is bound to or precipitated around calcium containing oxide and do not deposit at grain boundaries as free sulphides during solidification. In calcium free treated steels, sulphur precipitates as small particles of MnS in the last liquid to freeze. The MnS particles are deformed to form stringers when hot rolled. However, sulphide inclusions containing calcium have a globular shape which does not deform during hot rolling.

The extent of inclusion modification in steel is an essential feature in secondary steelmaking by calcium treatment. Portion of the calcium added to the liquid undergoes reaction and remain in the liquid as dissolved calcium in the form of inclusions or go to the slag as slag constituent. The rest escape the system in form of vapour. It is essential that the calcium added is to be consumed by the liquid steel to the maximum extent to make the calcium injection efficient and cost effective. In this regard, study of calcium recovery is an important factor for process optimization.

Steel properties improve with calcium treatment of liquid steel. The improvements in the properties include (i) improvement of mechanical properties especially in transversal and through thickness direction by modifying MnS to undeformed globular (Ca-Mn)S or CaS, (ii) improvement of steel machinability at high cutting by forming protective film on the tool surface which prolongs the life of the carbide tool, (iii) improvement of surface quality and polishability, (iv) minimizing lamellar tearing in large restrained welded structures and the susceptibility of steel to reheat cracking as in the heat affected zones of welds, and (iv) improvement of steel castability by preventing or minimizing nozzle clogging.

The inclusions can also exist as a single phase or as a multiphase inclusion. Normally, liquid inclusions are less likely to induce clogging than the solid inclusions during the continuous casting process. However, the tendency for inclusions to agglomerate and clog is highly dependent upon the interfacial energies and contact angles between the specific inclusion-metal-gas-refractory system.

During the operations in the integrated iron and steel plant, sulphur can be removed by hot metal pretreatment while the oxygen activity is low. As such, there is less need for sulphur removal during ladle treatment in the integrated steelmaking process. In electric furnace steelmaking, however, sulphur removal is to be performed during ladle treatment after the steel and slag are fully deoxidized. This sulphur removal normally needs a more aggressive deoxidation practice and strong stirring to achieve low sulphur levels in the steel. The processing requirements for desulphurization in the ladle can influence the composition and morphology of the inclusions formed in the process. For example, Al2MgO4 spinel inclusions readily form in aluminum killed steels which are subjected to a strong desulphurization treatment, while alumina (Al2O3) inclusions are more common in steels which are not subjected to a strong deoxidation and desulphurization treatment.

The rate of inclusion agglomeration, flotation, capture and retention in the slag layer is a strong function of the composition, morphology and phase (solid or liquid) of the inclusion population in the ladle. While liquid inclusions readily dissolve in the slag once they pass through the slag metal interface, they are also less likely to agglomerate, float and pass to the slag layer than solid inclusions while in the steel bath. This behaviour, in general, leads to a lower coarsening and flotation rate for liquid inclusions than solid inclusions and, hence, a lower overall inclusion removal rate for liquid inclusions compared to solid inclusions.

Inclusion formation during deoxidation

In order for the liquid steel to be cast in the continuous casting machine, dissolved oxygen in the liquid steel is to be sufficiently reduced to avoid CO (carbon mono-oxide) gas evolution during solidification. The most common elements used in steel deoxidation are manganese, silicon, aluminum, and calcium. Complex deoxidation practices can be used which combine these deoxidants. In some cases, titanium, zirconium, and rare earths can also be used for deoxidation.

Equilibrium between various deoxidant additions and oxygen in steel can also be calculated for a wide range of elements and over a wide range of concentrations and temperatures using thermodynamic software. This software is capable of calculating equilibrium conditions for complex deoxidation and can also account for higher order solute atom interactions in steel which lead to the retrograde oxygen solubility observed with strong deoxidants.  As shown in Fig 6, it can take some time to approach equilibrium concentrations in the bulk steel composition as it takes some time for inclusions to agglomerate and float from the steel. The oxygen content of the steel can vary widely prior to deoxidation. Typical BOF (basic oxygen furnace) and EAF (electric arc furnace) oxygen levels at the furnace tapping can range from 200 ppm to 800 ppm oxygen. In addition, partial deoxidation is sometimes performed during tapping using manganese, silicon and / or aluminum. When deoxidant is added to the liquid steel, the dissolved oxygen in the steel reacts with the deoxidant addition to form an oxide as per equation xM (in steel) + yO (in steel) = MxOy (inclusion.

Fig 6 Dissolved and total oxygen content during ladle processing

Nucleation of oxides can occur by homogeneous nucleation, particularly (i) when the level of super-saturation of the dissolved metal and oxygen is high, or (ii) it can occur by heterogeneous nucleation on inclusions formed earlier during the steelmaking process, or (iii) on argon bubbles used to stir the ladle or (iv) on the refractory walls of the ladle. The rate of nucleation is influenced by the level of super-saturation and the interfacial energy of the deoxidation product in contact with steel. The super-saturation ratio and the interfacial energy between the steel and the oxide both influence the critical radius for nucleation, and hence influence the size and number of inclusions initially formed at the onset of deoxidation.

It is normally accepted that nucleation occurs very quickly upon initial addition of deoxidant and that both homogeneous and heterogeneous nucleation can occur. It has been found that there is a strong relationship between the level of oxygen and the size of the inclusions formed and that the early stages of inclusion formation is controlled by oxygen diffusion in the liquid steel.

It is interesting to note that the time needed to reach a uniform composition in an argon stirred ladle is of the order of minutes, while the time for nucleation is normally of the order of 0.1 second or less. This implies that a wide range of levels of the addition of deoxidant is mixed and dispersed in the ladle. This can lead to the formation of a range of possible inclusion morphologies in the steel bath, depending upon the local deoxidant concentration and oxygen super-saturation levels which exist in different regions within the ladle.

Nucleation can continue to occur until super-saturation is no longer present. At this point, further nucleation can only occur as the temperature of the steel decreases, when alloying elements are added to the steel or when oxygen is re-introduced into the steel during reoxidation events. Under these conditions, the level of super-saturation is normally much lower and heterogeneous nucleation of new oxides onto pre-existing inclusions is normally more favourable.

Inclusion evolution after deoxidation

Coarsening, agglomeration and removal of inclusions – Once a population of inclusions is nucleated, the inclusions can grow by a variety of mechanisms. The average particle size of the inclusion population can increase through Ostwald-ripening, continued heterogeneous nucleation on previously nucleated inclusions, and through particle agglomeration. Particle collisions which lead to agglomeration can occur by Brownian motion, turbulent motion, velocity gradients, and Stokes collisions. Cavity bridges can also promote attraction and agglomeration of non-wetting particles. Fig 7 shows cavity bridge mechanism of particle agglomeration based on laboratory experiments which directly measured the attraction forces between oxides in liquid steel.

Fig 7 Cavity bridge mechanism of particle agglomeration

The size of the inclusions and their density drives the removal of inclusions from the steel by flotation. Stokes’ law provides a rough estimate of the effects of inclusion diameter and particle density on flotation rate. The flotation rate is strongly favoured by a large inclusion diameter and a low inclusion density relative to the liquid steel density.

Rising bubbles from argon stirring also promote the removal of inclusions if wetting between the inclusion and the argon bubble is favourable for attachment. The fluid flows induced by argon stirring or electromagnetic stirring can also promote inclusion transport and removal to the slag layer or to the ladle wall. Fig 8 shows an overview of these mechanisms for inclusion growth and removal from the ladle.

Fig 8 Inclusion growth and removal mechanisms

The effectiveness of argon stirring on inclusion removal depends on several factors which include porous plug location, number of plugs, the size of the ladle, the intensity of the stirring, and the contact angle of the inclusion with the steel and stirring gas. In general, liquid oxide inclusions tend to have much lower contact angles than solid oxide inclusions and normally have less than a 90 degree wetting angle. This has implications for particle agglomeration and bubble attachment mechanisms which influence the rate of inclusion removal from the ladle.

One of the studies has demonstrated that the rate of removal of particles from a fluid by gas bubbling is very dependent on the wetting angle of the particle-fluid-gas in a water model system, and that the removal rate drops sharply when the wetting angle drops to below 90 degrees. Steel systems also have similar effects. Argon bubbling has also been shown to selectively remove under-modified solid calcium aluminate inclusions from an inclusion population in the continuous casting machine tundish and argon stirring prior to calcium treatment has been seen to be more effective at removing inclusions than the argon stirring after calcium treatment.

Both inclusion agglomeration and bubble attachment mechanisms depend upon the interfacial energy and wetting angle between inclusions and the liquid steel and mechanisms which promote inclusion agglomeration and growth. Bubble attachment and cavity bridge agglomeration are both favoured by a large contact angle, suggesting that several solid inclusions are easier to remove from the ladle than liquid inclusions. The morphology of the solid inclusions can also influence agglomeration and flotation.

Macroscopically, the inclusion removal from the ladle can be observed by monitoring the total oxygen content of the steel or by evaluating the number or area density of the inclusions. For a given stirring intensity, stir configuration and inclusion population, longer rinse stir times normally result in progressively more inclusion removal. This is true both for argon stirring and stirring by other methods such as electromagnetic stirring.

Control of intensity of stirring in the ladle is important. While the increase of the stirring intensity leads to increased rates of inclusion removal, excessively increased stirring intensities can induce reoxidation by exposing the steel to air through the slag ‘eye’ where the slag is pushed back by the rising bubble plume or strong upward recirculating flow during the stirring. This condition can lead to a higher final inclusion content.

Recently, SEM-AFA (scanning electron microscopy – automated feature analysis) analysis has made it possible to analyze a large number of inclusions in a sample in a reasonable time. This has allowed the observation of the size distribution and population number density of inclusion populations at various stages in the process. It has been shown in a study that the shape of the inclusion size distribution changes with time, forming a log normal distribution early in the inclusion population’s life cycle and evolving to a power law or fractal size distribution later in the population’s life cycle when active nucleation is no longer taking place.

Several studies have reported similar trends in the evolution of the shape of the inclusion population in ladle deoxidation using models for nucleation and growth of inclusions during deoxidation in the ladle and by direct experimental observation. Observations of the shape of inclusion size distributions have been used successfully to identify reoxidation events, where oxygen from the air generates a new population of reoxidation inclusions which have a log normal distribution. On a log-log plot, this new population of reoxidation inclusions deviate from linearity for conditions of a tundish first fill and eventually subsiding and re-establishing a linear distribution. Similar observations have been there in the ladle after additions are made.

The type of deoxidant and the conditions of super-saturation employed in the deoxidation process have a strong influence on the rate at which the inclusion population changes over time. It has been observed that there is a rapid increase in diameter and reduction of the number of inclusions with time in aluminum deoxidized steels and little change in diameter and number with time in titanium and complex titanium-aluminum deoxidized steels. Also, there are large differences in the agglomeration rate of different inclusions in steel.

The size distribution of alumina clusters is observed to shift to progressively larger diameters and lower inclusion counts with time. By comparison, the size distribution of the titanium / aluminum complex deoxidant, which is liquid with a low wetting angle, does not increase in diameter with time and the inclusion count drops much more slowly with time because of the smaller overall diameter of the population.

Once the inclusions are transported to the slag layer in the ladle, the inclusions are to contact and become incorporated into the slag layer to be removed from the system. The rate of passage of inclusions through the slag-metal interface and the rate of dissolution of inclusions in a CaO-Al2O3 (lime-alumina) slag have been studied. During the study, it has been observed that alumina inclusions pass through the slag metal interface quickly, but that liquid MnO-SiO2-Al2O3 inclusions take much more time to pass through the interface and, in some cases, are not captured by the slag at all. It has also been found that the rate of inclusion dissolution in the slag is diffusion controlled for alumina inclusions dissolving in the slag.

The rate of dissolution of Al2O3 and MgO (magnesia) inclusions in various slags has also been studied. During the study, it has been found that in some slag systems, intermediate reaction products such as CaAl12O19 and Ca2Al2SiO7 formed on the inclusion surface and slowed the dissolution rate of the inclusion into slag.

The rate of dissolution of Al2MgO4 oxide spinel in slags of various compositions has also been studied. During the study, it has been found that the rate of dissolution is controlled by diffusion in the slag and is inversely related to the viscosity of the slag. If the time for passage through the slag–metal interface or the dissolution rate in the slag is slow, the slag-metal interface can become a site for solid inclusion agglomeration and re-emulsification of inclusions is possible.

Changes in inclusion composition during ladle treatment – The composition and morphology of inclusions can change considerably during the course of ladle treatment. The alloying additions made during ladle treatment can introduce elements which influence the overall composition and number of inclusions which are present. Also, additions such as calcium are used specifically for the purpose of modifying inclusion composition and morphology. Even in the absence of these additions, the inclusions present in a ladle after deoxidation can change composition over time through interactions with the slag and refractories used in the ladle, or with the atmosphere if the ladle stirring is aggressive.

Mini steel plants which produce aluminum killed steels with restricted sulphur levels frequently experience a shift in the inclusion chemistry during desulphurization. Under low oxygen conditions after the aluminum killing, argon stirring can promote the transfer of magnesium from the slag and the ladle refractory lining to the steel. During this process, the alumina inclusions gradually pick up magnesium and the inclusion population gradually shifts from Al2O3 to Al2MgO4 spinel. This exchange is influenced by the oxygen potential of the system and the composition of the slag. It has been found that MgO in slag and in the refractory both contribute to magnesium pickup and spinel formation, but that slag has a higher contribution to the pickup than the refractory.

While calcium treatment is used normally to convert solid alumina inclusions to liquid calcium aluminate inclusions to improve castability, calcium treatment is also quite successful at modifying solid spinel inclusions. Different studies have shown that spinel modification with calcium treatment reduces the magnesium content of the inclusion population and drives magnesium back into solution in the liquid steel. Subsequent exposure to reoxidation can result in the reformation of spinel inclusions through the reoxidation of magnesium in solution in the steel later in the process. These studies have also shown that CaS formed during calcium treatment can serve to suppress the reformation spinels during reoxidation by acting as a reservoir for calcium which can be released during reoxidation to further modify newly formed alumina. Under these conditions, the presence of some CaS in the inclusion population can be beneficial as long as it does not contribute to nozzle clogging during the continuous casting of steel.

It has also been shown that if the spinel inclusion population is high enough before calcium treatment, solid MgO inclusions can also form along with liquid calcium aluminate inclusions after the calcium treatment.  This type of inclusion has been found to be very detrimental in high performance steel such as line-pipe steels. Such conditions can arise when aluminum killed steels are heavily aluminum deoxidized and held in contact with high MgO slags and refractories for extended periods.

Impurities in alloy additives also have a strong influence on the inclusion population and the efficiency of inclusion removal. For example, the effects of calcium content of ferro-silicon on inclusion composition and the tundish clogging sensitivity has been cited. In some steel melting shops, calcium in ferro-silicon is managed to modify inclusions directly without additional calcium treatment while in other operations calcium in ferrosilicon is to be restricted to avoid changes in alumina morphology which promotes clogging. The stage in the ladle furnace processing where inclusions are modified in relation to the application of soft argon stirring to remove inclusions can have a pronounced effect on the inclusion removal efficiency. Argon rinse stirring prior to calcium treatment, when solid inclusions are present, has been shown to be more effective at removing inclusions than argon rinse stirring after the calcium treatment, when the inclusions have been modified to liquid calcium aluminates.

In short, the evolution of the inclusion population over time during ladle treatment can be complex. Given that the effectiveness of inclusion removal by coagulation, bubble attachment and interface capture is highly dependent upon the composition, phase, and morphology of the inclusion population developed during the ladle treatment practice. It is obvious that there are time periods during ladle treatment where rinsing and flotation treatments are more effective than others. The ability to develop ladle treatment strategies to take advantage of these preferred treatment times where inclusion coarsening and removal rates are the most rapid ultimately is dependent upon the cleanliness of post-flotation-treatment additions, control of reoxidation, and slag entrainment during stirring and the effectiveness of reoxidation protection during steel transfer.

Leave a Comment