A New Lance Design for BOF Steelmaking

A New Lance Design for BOF Steelmaking

Sambasivam, R

This article summarizes the outcome of research work carried out to improve the performance of the oxygen lance in the LD steelmaking process. It is stated that the lack of control of the foamy slag and the augmentation of interfacial area creation between the slag and metal are the major hindrances in running the process effectively for improved turn-down steel quality. The ineffectiveness of the existing design in producing liquid metal droplets in the presence of slag foam is explained. In order to augment the droplet generation, a new oxygen lance with a central subsonic nozzle through which flow can be controlled has been introduced and its blowing performances studied using numerical and water model studies. The jet characteristics studied in the numerical simulations show no jet coalescence. The interferences of the jets with the bath have further been analyzed by hydrodynamic model studies. It has been found that the droplet generation rate improves significantly due to the presence of the central jet. Further, it has been observed that controlling the flow rate through the central hole can be used as an effective process control tool.

DOI: 10.1007/s11663-006-9004-3

© The Minerals, Metals & Materials Society and ASM International 2007


THE worldwide increase in demand for steel has encouraged countries such as India with a good raw material base to seriously consider significantly augmenting their production capacity. Such an increase in the capacity would address the projected increase of per capita consumption of steel and help India’s projected economic development. TATA Steel, as part of its growth strategy, feels that the proposed capacity increase must be supported by advancements of the present day’s steelmaking processes with emphasis on improving steel quality and shortening of the processing time. Typical of Indian raw materials, the phosphorous content of the hot metal charged into the LD vessel at TATA Steel is as high as 0.22 pet. This level of input phosphorous is very high as compared to the world average given the fact that the demands for the lower phosphorus in the finished products for certain critical grades of steel are severe. Hence, at TATA Steel, efforts have been on studying the entire steelmaking process from a more fundamental point of view using numerical, experimental, and analytical tools in order to consistently produce steel with a low turndown phosphorous (

The formation of a liquid slag with the right composition such as basicity and FeO content is the foremost important step in the LD steelmaking process, because dephosphorization depends on the composition of the slag, the temperature, and the kinetics and mass transfer achieved through mixing of liquid metal and slag. During the refining period, carbon oxidizes to form carbon monoxide gas, which tries to escape through the slag causing it to swell in volume, and this, in general, is called slag foam. The foam, thus formed, occupies a large volume of the vessel and is expected to cover the lance head completely during the peak decarburization period. This foam provides the huge interfacial area between the slag, metal, and gas phases and thereby promotes the interfacial refining reactions such as dephosphorization. The production and maintenance of the slag foam are mainly controlled by the oxygen jets from the lance. The process of steelmaking is highly dynamic due to the continuously changing conditions such as temperature, composition, and basicity during the oxygen blowing period, and the lance height is varied in order to control the intensity of the impingement of the supersonic oxygen jets on the liquid metal surface. This is termed as soft and hard blowing, respectively, for the higher and shorter lance distances from the metal bath (“lance height”). Apart from acting as an oxygen supplier, the lance also plays an important role in controlling the dynamics of the steelmaking process. Thus, it is imperative that proper designing of the lance and controlling the flow rate or the lance height during the blow can be done to improve the efficiency of the LD Steelmaking process and enhance the quality of the steel produced.

As is evident from the preceding explanations, apart from the thermodynamics of the refining reactions, the fluid dynamics inside the vessel also plays a major role in determining the kinetics of the refining reactions. It is a well established fact from the reaction kinetics of the dephosphorization in BOF that the rate limiting step for the removal of phosphorus is the mass transfer occurring at the metal-slag interface. This metal-slag interface is provided by the generation of metal droplets during the blow in the steelmaking vessel, and these droplets enhance the kinetics of the process and help in closely approaching thermodynamic equilibrium. Additionally, the generation of the metal droplets is advantageous during the period when there is slag foam on the top of the liquid metal surface in order to avoid the problem of lance skulling. Hence, it is desirable to have control over the droplet generation during the blow.

This article presents a novel lance design towards improved droplet generation, as and when it is needed, in order to accelerate the refining reactions, especially dephosphorization. A seven-hole lance is proposed with the central nozzle being supplied by a separate flow rate controllable oxygen line and its impact on the generation of droplets studied in detail. The results obtained with numerical simulations and hydrodynamic model experiments have been presented in detail. The strategy to enhance process control during steelmaking by varying the flow rate through the central hole is also discussed.


As mentioned earlier, as the cleanliness requirements and chemistry tolerances become more stringent due to the high-end applications of steel, the process control during steelmaking assumes greater significance. Because the oxygen lance is one of the key process control tools available to engineer the rate of the steelmaking reactions within the LD vessel, a more in-depth understanding of the role of the lance and the translation of this understanding into the basic design process of the lance are warranted.

A large number of studies[1-6] have been performed on the steelmaking processes and the role of oxygen lance and its design, in particular, over the years. Standish et al.[3,4] have reported the mechanism and the effect of top and bottom blowing parameters on the droplet generation in the steelmaking vessel. The investigation[3] on the droplet generation by a gas jet impinging on a liquid surface indicated that there exists two different mechanisms of droplet generation in the so-called “dropping” and “swarming” regions, which result in different droplet generation rates. Nozzle inclination angles in the steelmaking vessel have evolved with time and, at present, have been fixed between 9 and 18 deg to the vertical axis. This selection of inclination angle of nozzles was mostly based on the detailed experimental work by Lee et al.[5]

Basic studies, giving information on the depth and width of the cavity formed by the impingement of a gas jet in liquids, are numerous. Sharma el al.[6] observed that the depth of the jet cavity is governed solely by the mechanical forces provided by the oxygen jets. Banks and Chandrasekhara[7] worked on the interaction induced by a high speed air jet directed at a stationary water surface and reported on the parameters such as cavity depth, diameter, and periphery lip height and its relationship with the jet momentum. They worked on the cylindrical and planer geometries and deduced from their investigations a correlation for the maximum depth of penetration. Following the studies of Banks and Chandrasekhara, Wakelin et al.[18] carried out experiments including the dimensions of the cavity and the velocity fields in the gas jets. They further developed the one-dimensional model of depth of penetration that matched well with the experimental results. Molloy[9] described the gas jet and liquid bath interaction as having three flow modes. These free surface modes of deformation have been identified to be dimpling, splashing, and penetrating depending on the jet momentum and the liquid properties. Standish et al.[4] concluded that the droplet generation is not a monotonous function of the lance height and that the maximum droplet generation is obtained at a certain lance height. hot model experiments were carried out by Subagyo et al.[10] in order to understand the droplet generation characteristics. He presented an empirical correlation to predict the rate of droplet generation per unit volume. It was found that the cold model results are well in agreement with what was observed on a hot model experiment. The rate of increase in the droplet generation, in their work, was correlated to the Kelvin-Helmholtz instability criterion.

As discussed previously, the augmentation of the interfacial area between slag and metal can help in enhancing the dephosphorization reaction in the LD vessel apart from the other well-known chemical factors governed by the thermodynamics of the reactions. The ability of the oxygen lance to produce liquid metal droplets has to be given due importance in the design, because improving the other parameters impacting dephosphorization such as higher basicity and increased FeO content of the slag involves higher cost and yield loss. Further, tapping the heat at lower temperatures, which aids dephosphorization, may not be possible for smaller heat sizes. Hence, it is expected that improving the mixing and droplet production rate within the vessel hold the key for sustained cost-effective improvement in dephosphorization in the near future.


Earlier droplet generation studies[3,4] were performed in water models where the slag phase had not been simulated properly as foam. In the LD vessel, after 2 to 3 minutes of the start of the blow, a foamy slag is created. It is evident from the lance samples obtained from the plant that the lance head remains completely covered by the slag foam during most of the blow. However, water model studies did not include the effect of slag foam on the jet characteristics and thereby on the droplet generation.

During expansion, the jet engrosses the ambient fluid surrounding it and starts moving the initially stagnant ambient along with it. The loss of momentum from the high velocity jet to the ambient is directly proportional to the density of the ambient medium. Because the density of the slag foam in the LD vessel is expected to be two orders of magnitude higher than the pure gaseous medium, the jet expanding into the slag foam in the steelmaking vessel loses a significant portion of its momentum to the slag foam. In other words, the jet velocity is much lower when it expands into the slag foam as compared to its expansion in a gaseous medium. Hence, the impact velocity of the jet on the liquid metal surface is less in the presence of the slag foam as compared to that of the jet expanding in a gaseous medium. The reduction in the velocity magnitude in the presence of slag foam was suspected by the earlier researchers[11], but the extent of the reduction in velocity has not been quantified so far. It is well known from the water model studies[1-6] that the shear generated on the liquid surface due to the impinging gas jet and the resulting droplet generation rate is proportional to the impact velocity of the gas jet. Because the oxygen jet expanding into the ambient filled with slag foam has a very small impact velocity, it will not be able to generate metal droplets efficiently during most of the blow. So, the augmentation of the production of metal droplets in the presence of slag foam is the key challenge the steelmakers face presently in order to improve the slag metal interfacial reactions such as the removal of phosphorus.

Furthermore, any complicated process such as steelmaking requires tools that will help better control of the process. The critical process control tool available with the steelmakers during the blow in the LD vessel is the variation of the lance height. It is believed that by varying the lance height, the distribution of oxygen between the slag and metal phases is varied and thereby the chemical reactions influenced. A blow profile in terms of varying lance height is designed according to the operating conditions inside the vessel, and this remains the only process control parameter during the blow. Changing the flow rate through the supersonic nozzles is not possible because the nozzle has been designed for a particular pressure ratio and flow rate. If one more tool is provided to effectively control the dynamics of the process, it is believed that it would help the steelmakers to effectively control the process. In this study, while designing the seven-hole lance, the aforementioned needs of generating more liquid metal droplets and providing a versatile tool to control the kinetics of the steelmaking reactions in the vessel have been considered.


In the present lance design being used in the LD shops of TATA Steel’s Jamshedpur works, the lance has six equally spaced supersonic nozzles with an appropriate angle of inclination. In Figure 1, the typical recirculatory flow of gases and slag foam expected with the present lance design inside the LD vessel is shown schematically. The expanding jets entrain the surrounding slag foam and thereby cause the formation of a recirculation regime in the vessel. It is logical to arrive at the conclusion that the slag will more or less follow the path of this recirculation. From Figure 1, it can be clearly seen that the space between the outer surfaces of the jets and the vessel wall will have more slag concentration as compared to the space in between the multiple supersonic jets. As the supersonic jets act as a protective cover to prevent slag entrainment in this central space just below the lance head, the amount of slag in this space is expected to be much less.

As mentioned in the earlier section, the jet expanding into the slag foam loses its momentum to the slag foam, and this slowly moving jet results in poor droplet generation. During most of the blow duration, the peripheral supersonic jets expand into the surroundings filled with slag foam, and so their impact velocities on the liquid metal surface are less and the droplet generation capability lower as compared to the initial stages when there is no slag foam. As compared to this, the space in between the six jets has lesser or no slag concentration due to the absence of slag foam in this space. If a jet expands in this central space in between the peripheral jets, it will not lose all its momentum to the surroundings unlike the peripheral jets. It is expected that this central jet will reach the metal surface with higher impact velocity, which will result in increased tangential forces on the liquid metal surface and produce metal droplets thrown into the slag foam creating more slag metal interfacial area. The schematic representation of the concept of the new seven-hole lance design is shown in Figure 2. This basic observation remains the base of the new lance design proposed in this article.

It is a well-known fact to the steelmakers that increased droplet production in the initial stage of the blow is not preferred due to skull formation and spitting problems from the vessel mouth. In the initial stage of the blow, there is no slag present in the vessel. The presence of the slag, especially a foamy slag, acts as an energy absorber to the high velocity metal droplets produced in the process and slows the velocity of the droplets. This reduces the skulling and spitting problems to a great extent. It takes 2 to 3 minutes after the start of the blow, depending on the Si content of the liquid metal, to form a reasonable amount of slag in the LD vessel. The increased metal droplet production is preferred only after this stage of the process in order to avoid the negative effects of droplet production such as skull formation on the lance and vessel mouth. In the present design, the metal droplet production is controlled in the initial stage of the blow by operating at a higher lance height. However, with the central jet in place, the droplet production is expected to be high even with the higher lance height in the operating range of lance heights in the LD vessel. This implies that the central hole must come into operation at the end of the initial slag formation stage. This can be achieved only by giving a separate supply line for the central hole, whereas the other six peripheral nozzles will have a common supply line.

Further, it is felt that the flow through the central hole can be used as a tool to control the process dynamics within the LD vessel. Because the central nozzle can result in increased droplet production, the flow rate through this nozzle can be increased or decreased to control the amount of metal droplets. It is obvious that the presence of metal droplets hastens the removal of phosphorous and carbon. So, by controlling the flow rate of gases through the central nozzle, the amount of metal droplets produced in the vessel can be altered and thereby the rates of the reactions effectively engineered. Thus, it is imperative that the flow through the central nozzle should be controllable. Controlling the flow rate through a supersonic nozzle over a wide range is not possible because it is designed for a particular pressure ratio and flow rate. In the actual LD vessel, the flow rate of oxygen in the lance is varied by ± 5 pet from the design flow rate without causing much change to the jet characteristics. However, the wider variations of flow rates, as required to influence the droplet generation and thereby control the process dynamics, are not possible with supersonic nozzles. Increasing the flow rate through the supersonic nozzle or overblowing will result in an expansion wave at the nozzle exit and cause poor jet characteristics. Decreasing the flow rate through the nozzle by operating it with a lesser pressure ratio than the design pressure ratio, or, in other words, if the nozzle is underblown, can cause strong shocks even within the diverging section of the nozzle itself. Such shocks can severely affect the performance of the supersonic nozzle, and the hot metal and slag foam can also be sucked into the nozzle, reducing the life of the nozzle considerably or even leading to a lance failure, i.e., leakage of water into the vessel causing explosions. Thus, it is clear that it is not possible to have a great degree of control of the process by adjusting the flow rate of a supersonic nozzle. Because of the aforementioned needs, it is mandatory to have a subsonic nozzle, i.e., nozzle with only a converging section through which it is easy to obtain a wide range of flow rates by changing the supply pressure of the central nozzle. In order to obtain a higher impact velocity of the central jet at the liquid metal surface when needed, the central subsonic nozzle will be larger than the supersonic nozzles. Thus, the new seven-hole lance design proposed is equipped with one larger central subsonic nozzle that will be controlled separately during the blow with a separate gas supply line and six peripheral supersonic nozzles with high-pressure gas supply line. The suggested new seven-hole lance design[12] is shown schematically in Figure 3.


Numerical and experimental investigations have been performed to analyze the new seven-hole lance design proposed in the earlier section. The jet characteristics and their coalescence pattern have been studied in the numerical studies. The impact velocities of the jets on the liquid metal surface have also been investigated. In the reduced scale hydrodynamic model investigations, the droplet generation mechanism with the seven-hole lance design has been studied. The droplet generation has been quantified and analyzed as the function of flow rate through the central hole. The results obtained from these simulations have been explained in detail.

A. Numerical investigations of Lance Performance

Numerical simulations to study the jet characteristics with the existing six-hole design and the new seven-hole design with a central bigger subsonic nozzle were performed using the commercial computational fluid dynamics software, FLUENT 6.2.1.[13] In the existing design, as mentioned earlier, there are six supersonic nozzles spaced at equal distances along the periphery. In the new design, one more subsonic nozzle has been introduced at the center in between the six peripheral nozzles. The flow rate through the central subsonic nozzle was kept as variable. A flow rate ratio for the central nozzle was defined as the ratio of the flow rate through the central nozzle to that of one of the peripheral supersonic nozzles. In this article, the numerical results obtained with the flow rate ratio of unity, i.e., the flow rate through the central hole equal to that of one of the peripheral nozzles, have been presented. The exit diameter of the central subsonic nozzle was calculated based on the flow rate ratio of 1 by fixing the exit Mach number at this flow rate ratio to unity. The lance height in the model was kept as 2 m, whereas in the LD vessel, it varies from 1.5 to 2.2 m. The liquid metal surface was assumed to be a shear stress free horizontal surface. The penetration and the undulations on the metal surface have not been modeled, because the purpose of this simulation is to estimate the impact velocities of the jets and the jet characteristics with the new seven-hole lance design.

The conservative form of the governing equations was solved to obtain the jet characteristics of the compressible flow. The system of governing equations was closed using the k-ε turbulence model. It is known that the k-ε turbulence model predicts the flow features of the multiple jets with some deviations due to the anisotropy of the turbulence.[14] Nevertheless, this model was used in this simulation because it is easy to get reasonable solutions quickly with the k-ε model with lesser computational efforts. The following boundary conditions were specified. At the gas inlet boundary to the lance pipe mass flow rate, static pressure, stagnation temperature, and turbulence parameters were specified. Standard wall functions were used to model the near wall flow. Static pressure was specified at the vessel opening and the liquid metal surface was assumed to be a shear stress free horizontal surface. A second-order upwind scheme was used to discretize the governing equations. The PISO algorithm was employed for pressure velocity coupling.

To reduce the computational time of the numerical simulations, only half of the total flow domain was simulated by splitting the entire domain using the vertical midplane of the vessel. The computational domain along with the mesh used for the simulation of the seven-hole lance design is shown in Figure 4. Slightly more than 1.3 million grid cells were generated in the computational domain. The simulations were performed with 12 processors in a Linux cluster. On average, one simulation lasted for 72 to 80 hours.

B. Hydrodynamic Studies of Lance Performance

1. Experimental setup and design of seven-hole lance

A 1:6 scale down model of the LD vessel being used in Tata Steel’s works was designed and made up of Plexiglas to ensure a better visualization of the dynamic nature of the bath. The complete experimental setup is shown in Figure 5. The lance is made up of seven nozzles with separate flow controls to the central and peripheral nozzles. The air supply line to the lance assembly consisted of a pressure regulator and air flow rotameter. The geometrical construction of the sevenhole lance is depicted in Figure 6. The lance tip was thread fitted in order to carry out experiments with different lance tip designs.

The quantification of the droplet generation was studied in order to understand the optimum flow rate through the central nozzle that maximizes the droplet generation rate. The droplet generation rate was measured by a water collecting pan having dimensions of 400 × 100 × 50 mm^sup 3^, and the measurements were carried out for the existing six-hole lance and the new sevenhole lance with a central nozzle. The dimensions of the pan were decided in order to measure the average droplet generation of one nozzle out of the six peripheral nozzles. The rate of droplet generation is expressed in terms of the rate of mass of droplets collected (kg/ min) on the pan.

C. Results and Discussion

1. Experimental

Effect of central nozzle on droplet generation rate. The droplet generation mechanisms were investigated when all seven holes were in operation, and comparisons were made with that of when only the six peripheral nozzles were in operation. It is visually inferred that the extent of droplet generation is much higher when all seven holes were in operation as compared to that when only the six peripheral nozzles were in use. It was observed during the experiments that there is a critical flow rate at which the onset of the improved droplet generation starts. The mechanism for the accelerated rate of droplet generation, due to the presence of the central jet, is explained schematically in Figure 7. The center jet impacts the liquid metal vertically and creates the central strong depression of the liquid surface. The depression thus formed is wavy in nature and provides lips out of the central water paddle, as shown schematically in Figure 7. The water lips thus formed around the paddle are then torn apart by the side jets and yield an improved droplet production. These peripheral jets in the actual vessel are expected to resist the slag foam entering the center space among the peripheral jets in the actual LD vessel. This ensures that the central jet reaches the metal bath surface with high momentum and produces metal droplets similar to the one schematically sketched in Figure 7.

Volumetric estimation of rate of droplet generation with seven-hole lance. The rate of droplet generation was studied over a wide range of air flow rates through the central nozzle in order to obtain the optimum flow rate that maximizes the droplet generation. A flow rate ratio, X, is defined as the ratio of the flow rate through the central hole to that of one of the peripheral nozzles.

The rate of droplet generation is plotted against the flow rate ratio in Figure 8. The flow rate through the central nozzle was varied from a flow rate ratio of as low as 0 pct to as high as 125 pct. However, the total flow rate of air in all the experiments was kept constant, and this value is equal to the flow rate through the existing sixhole design. The optimum flow rate through the central nozzle was obtained by maintaining the balance between an improved droplet generation and control of splashing and spitting out of the mouth of the vessel. It was quite obvious from Figure 8 that as the flow rate through the center hole is increased progressively, the rate of droplet generation is enhanced. It can be observed in Figure 8 that for the flow rate ratio A”, of 1 (100 pet), the droplet generation almost doubles and reaches the maximum value. Beyond this flow rate, there is vigorous splashing and spitting out of the mouth of the model. Thus, from the hydrodynamic model experiments, it is possible to decide the optimum flow rate ratio, as the one that maximizes the droplet generation rate without causing spitting. A seven-hole lance tip with higher inclination angle was also studied, but there was not much improvement in the droplet generation rate for the flow ratio of 1.

Numerical simulations. As mentioned earlier, numerical simulations were performed in order to study the jet characteristics in the presence of the central jet and to analyze the jet coalescence pattern. In Figure 9, the temperature contours depicting the shock structures at the exit of the seven-hole lance design with the flow rate ratio of 1 through the central hole are shown. It can be seen clearly that at the exit of the supersonic nozzles, strong oblique shocks occur due to the mismatch in static temperature between the jet and the surroundings. The temperature at the jet axis approaches the surrounding temperature through a series of shocks. However, the shocks formed at the tip of the subsonic central nozzle are very weak. It can also be observed from Figure 9 that the exit diameter of the subsonic nozzle is larger than that of the supersonic nozzle. This helps in retaining higher impact velocity for the central jet at the liquid metal surface, though the exit Mach number of the central nozzle is less than that of the supersonic nozzle.

The jet coalescence, if it occurs, reduces the impact of the individual jets as the jets engulf a huge mass of ambient medium. Jet coalescence is thus considered detrimental. In order to study the extent of jet coalescence in the new design, the coalescence in the existing six-hole design was analyzed first. In Figure 10, the velocity contours at the middle plane of the domain are shown for the existing six-hole lance design. It can be seen from Figure 10 that the tendency for the jets to coalesce is low and the jets more or less follow the angle of inclination of the nozzles from the vertical axis. It can be observed from Figure 10 that there is a tendency to coalesce at the middle elevations. At the impact position of the jets on the liquid metal surface, the local pressure values increase due to stagnation of the flow. This higher pressure region around the impact positions of the jets pushes the jets away from one another and reduces coalescence.

In Figure 11, the velocity contours at the middle plane of the domain are shown for the newly proposed seven-hole lance design with a central subsonic nozzle with the flow ratio, X, of 1. From Figure 11, it can be observed that the central subsonic nozzle has not at all changed the characteristics of the peripheral supersonic jets. The peripheral jets’ characteristics remain the same as that of the jets shown in Figure 10. The presence of the central jet increases the local pressure in the space in between the peripheral jets and thereby reduces coalescence completely. From Figure 11, it can be seen that the jet mostly follows the geometrical path given by the angle of inclination. It is clear from the preceding explanations that the presence of the central jet does not promote jet coalescence as might be apprehended.

As mentioned in section IV, in order to effectively control the droplet generation during the oxygen blow, the flow rate through the central hole needs to be varied. This can only be possible if the central nozzle is a subsonic one. However, the droplet generation due to the shear at the liquid metal surface can be increased only if the impact velocity of the jet remains high. For the central subsonic nozzle, though the exit velocity at the nozzle is less as compared to that of the supersonic nozzle, the higher impact velocity at the metal surface is maintained by designing the nozzle with a larger exit diameter. It can be seen from Figure 11 that the impact velocities of the central subsonic nozzle and the peripheral supersonic nozzles are almost the same. Further, because the exit diameter of the central nozzle is larger than that of the supersonic jets, the area of the impact is also expected to be larger. This will facilitate the jet to produce higher metal droplet generation when needed during the blow.


As the demand from the market leaves no tolerance in the quality of steel, exercising effective control during the process of steelmaking and improving the effectiveness of the removal of the impurities such as C, P, S, Mn, Si, and P, in particular, in LD vessel without incurring further cost has drawn the attention of the researchers. Because the huge interfacial area between slag and metal generated during the blow is one of the key parameters controlling the process of phosphorus removal, a lot of attention has been paid to the improvement of liquid metal droplet production during the blow. The peripheral oxygen jets when expanding into the slag foam lose their momentum to the slag foam and thereby reach the liquid metal surface with low velocities and lesser ability to produce metal droplets. In this article, as a remedy, the new seven-hole lance design has been introduced, which gives increased droplet production and acts as a better process control tool during the blow. Because it is intended to control and vary the oxygen flow rate through the central nozzle during the blow, the central subsonic nozzle should be fed with a separate gas pipeline.

The proposed seven-hole lance design with the flow rate controllable central nozzle has been studied in detail using computational and experimental tools. Numerical results show that the presence of the central jet did not induce jet coalescence, and the impact velocity of the central subsonic jet was also calculated. Reduced scale hydrodynamic model experiments showed an excellent increase in droplet generation rate with the seven-hole lance design as compared to the existing six-hole design. Based on the hydrodynamic model experiments, the maximum flow ratio through the central hole is fixed to unity. Apart from the increased droplet generation and better process control, the additional benefits expected by having a central controllable nozzle on the steelmaking process are as follows.

1. Increased lance tip life: The space in between the peripheral jets is supplied with relatively cold oxygen gas, which otherwise might be filled with hightemperature dust, slag, and hot metal causing wear of the lance tip.

2. Prevention of dry slag formation: During the peak period of the decarburization, there is a tendency of lack of oxygen in the bath; hence, at about 75 pet of the total blowing time, there is a chance of FeO reduction due to the increased need for oxygen. Due to this, often, the emulsion gets disrupted due to the dry slag. The suggested central nozzle can bridge this oxygen gap, which might arise during this part of the blow and, hence, prevent the dryslag formation.

3. Better process control: The metal droplet production influences the removal of phosphorus and carbon. The flow rate controllable central nozzle can influence the droplet production and so can act as one more process control tool.

Based on the preceding discussion, the new seven-hole lance is expected to play a major role in the effectiveness of impurity removal by improving the droplet generation rate. Further, the variable flow rate through the central nozzle can be used to arrive at better dynamic process control strategies during the steelmaking process.


The authors thank the management of Tata Steel for giving permission to publish this work. They are thankful to all who have directly or indirectly contributed toward making the study at the Institute of Fluid Mechanics (LSTM, Erlangen, Germany). One of the authors (SNL) thanks Mr. Buelant Uensal, who helped during the setup of the LD vessel model.


1. A. Chatlerjee, N.O. Lindfors, and J.A. Wester: Ironmaking and Steelmaking. 1976. vol. 3. p. 21.

2. D.J. Price: Process Engineering of Pyrometallurgy, M.J. Jones, ed., Institute of Mining and Metallurgy, London, 1974, p. 8.

3. N. Standish and Q.L. He: ISIJ Int., 1989, vol. 30, p. 455.

4. Q.L. He and N. Standish: ISIJ Int., 1990, vol. 30. p. 305.

5. C.K. Lee, J.H. Neilson, and A. Gilchrist: Ironmaking Steelmaking, 1977, No. 6. p. 329.

6. S.K. Sharma, J.W. Hlinka, and D.W. Kern: Iron Steelmaker, 1977, p. 7.

7. R.B. Banks and D.V. Chandrasekhara: J. Fluid Mech., 1963. vol. 15, p. 13.

8. D.H. Wakelin: Ph.D. Thesis, University of London, London, 1966.

9. N.A. Molloy: J. Iron Steel Inst., 1970. p. 943.

10. Subagyo, G.A. Brooks, K.S. Koley, and G.A. Irons: ISIJ Int., 2003, vol. 43, p. 983.

11. R.D. Pehlke, W.F. Porter, R.F. Urban, and J.M. Gaines: BOF Steelmaking, Introduction, Theory and Design, Part I (ISS, Warrendale, PA, 1982), p. 597.

12. International Patent No. PCT/1N06/00153 filed on 4th May, 2006, based on Indian Patent Application No. 1023/KOL/05.

13. Fluent’s Users Guide, Fluent Inc., Lebanon, 2004.

14. Y. Tago and Y. Higuchi: ISIJ Int., 2003, vol. 43(2), p. 209.

15. P. Nillers: Past Present Future of Steelmaking in Europe BHM, 1992, vol. 137, p. 279.

16. J. Nagai, H. Take, K. Nakanishi, T. Yumamota, R. Tachibana, Y. Iida, H. Yamada, and H. Omari: “Metallurgical Characteristics of Combined Blown Converters,” Kawasaki Steel Technical Report No. 6. Kawasaki Steel, 1982.

R. SAMBASIVAM and S.N. LENKA, Researchers, S. CHANDRA, Chief of Technology (Global Wire Business), and S.K. AJMANI, Head of Research Group, Steelmaking and Casting, are with the TATA Steel, Jamshedpur, Jharkhand State, India – 831 001. Contact e-mail: rsambasivam@tatasteel.com F. DURST, Emeritus Professor and former Head, is with the LSTM-Erlangen, Institute of Fluid Mechanics, Friedrich-Alexander-Universität Erlangen-Nürnberg, D-91058 Erlangen, Germany. M. BOCK, Consultant, is with Saar-Metallwerke GmbH, D-66121 Saarbrucken, Germany.

Manuscript submitted June 23, 2006.

Copyright Minerals, Metals & Materials Society and ASM International Feb 2007

Provided by ProQuest Information and Learning Company. All rights Reserved