2025 Volume 65 Issue 10 Pages 1430-1440
In this paper, the oxygen lance of a steel plant is taken as the research object. The jet behavior under the influence of two factors of oxygen nozzle deflection angle and oxygen flow is studied by combining numerical simulation with industrial experiment. The results show that the coalescence behavior of the oxygen jet is obvious, and the oxygen jet basically completes the coalescence at about 0.9 m at the outlet of the oxygen lance. When the nozzle deflection angle of oxygen lance is 12°, the surface of liquid steel fluctuates greatly. When the oxygen flow is increased, the impact crater area of scheme 3 is effectively expanded and the stability is increased. Under the three experimental schemes, the maximum exposed area of liquid steel are 7.18 m2, 7.08 m2 and 7.42 m2 respectively. The reflection jet angle (θ) changes the most at low position, from 68.0° to 54.8°, 67.2° to 48.9°, 67.6° to 59.8°, respectively. When the oxygen lance position exceeds 1750 mm, the θ changes are less than 50°, the slag diffusion area is the largest, and the liquid steel reaction is dynamically stable. The industrial test shows that the composition of liquid steel and slag fluctuates greatly when 12° oxygen lance is used for blowing. When the nozzle deflection angle of oxygen lance increases to 13° and the oxygen flow increases to 42000 Nm3/h, the average oxygen content of liquid steel can be controlled to 573.28 ppm, the average total iron content in the slag can be reduced to 16.35%.
The modern smelting of long process is carried out according to the four main production units of “ironmaking for blast furnace → steelmaking for converter → secondary refining → continuous casting”.1,2) As one of the main production steps, steelmaking for converter undertakes the important primary refining tasks such as decarbonization, dephosphorization, heating and alloying of hot metal. It is also an important link in the continuous breakthrough of smelting technology, cost reduction and efficiency improvement.3,4,5) In order to enable steel enterprises to achieve low-cost and high-quality production in a highly competitive market environment, many researchers have studied the fluid behavior in the converter on the structural parameters of the top blowing process. The aim is to find the best top blowing smelting parameters to ensure the quality of liquid steel and further reduce the cost of blowing. Protopopov et al., Yang et al.,6,7) studied the effects of parameters such as the number of oxygen lance nozzles, Mach number (object velocity/sound velocity) and deflection angle on jet velocity distribution, coalescence behavior and impact area. The results show that increasing the number of oxygen lance nozzles will accelerate the jet attenuation and increase the jet coalescence. Under the same inclination angle, the change of Mach number has little effect on the jet. Under the same Mach number, the increase of nozzle inclination angle leads to the increase of impact range and the decrease of impact strength. Some scholars have also redesigned the oxygen lance to achieve the purpose of strengthening smelting. Solonenko et al.,8) proposed a new cylindrical oxygen lance nozzle, and the experimental results show that the oxygen lance can stabilize the supersonic oxygen flow, increase the length of the initial and transition segments of the oxygen jet, and achieve the purpose of strengthening smelting. Wang et al.,9) studied the effect of swirl structure oxygen lance on the shape and mixing efficiency of melt pool. The results show that the effect of oxygen lance position on melt pool shape is greater than that of oxygen lance flow, and the effect of oxygen lance nozzle deflection angle on melt pool shape is greater than that of oxygen lance nozzle swirl angle. Compared with conventional oxygen lance, the mixing efficiency of melt pool under swirl oxygen lance is increased by 5%–13.3%. Zhang et al.,10) also carried out numerical simulation of swirl oxygen lance. The results show that the swirl oxygen lance jet has higher radial and tangential velocity and more obvious impact area than the conventional oxygen lance jet. In addition, it shows excellent stirring capacity for the melt pool, increasing the average velocity of slag and reducing the volume of dead zone. Compared with the swirl structure oxygen lance, Hironori et al.,11) designed a new dual-angle, dual-flow, staggered dual-structure six-nozzle oxygen lance, and studied its jet characteristics. The results show that the mixing force of the oxygen jet on the melt pool can be improved by using the staggered dual-structure six-nozzle oxygen lance, and the performance of the oxygen jet is better than that of the traditional oxygen lance. When the large nozzle flow ratio is 55%–65%, the large nozzle angle is 12°–14°, and the small nozzle angle is 16°–18°, the nozzle flow ratio plays a leading role in the size of the oxygen jet velocity, and the larger the flow distribution, the larger the jet velocity. Compared with the structural design of oxygen lance, Wang et al.,12) studied the effect of the oxygen lance nozzle deflection angle on the impact area of the oxygen jet through numerical simulation, and found that the oxygen jet with an deflection of 15° had the largest effective impact area, and the mixing of the melt pool was more uniform. In addition to the influence of oxygen lance structure parameters on the melt pool, Jia et al.,13) studied the influence of the ratio of H (melt pool depth) to D (melt pool diameter) on the reaction rate of slag and metal under top blowing. The results show that the surface reaction rate constant not only increases with the increase of H/D, but also affects the fluctuation behavior of liquid steel level and the formation behavior of slag emulsion, which increases the reaction interface between slag and metal. The above scholars basically judge liquid steel flow in melt pool from the perspective of oxygen lance structural parameters by studying oxygen jet behavior (such as impact area, oxygen jet attenuation degree, etc.) or impact crater and flow field changes (such as dead zone volume proportion, impact crater depth, etc.), which can explain the influence of top blowing on liquid steel flow in the melt pool to a certain extent. However, there is a lack of in-depth description of the melt pool, such as the change of oxygen jet after impact, the diffusion behavior of slag layer, the action form of jet on the melt pool at different oxygen lance positions and other indicators that can deeply describe the flow of liquid steel in the converter have not been deeply studied.
In this paper, the impact behavior of high intensity and large flow oxygen jet on the melt pool of a 150 t converter in a steel plant is studied under different parameters and oxygen lance positions. Moreover, this research work differentiated and considered the characteristic evolution processes of supersonic oxygen jet velocity changes, impact crater morphology changes, slag layer diffusion behavior, and dead zone volume of the molten pool under different oxygen lance positions, different nozzle deflection angles, and high oxygen intensities supply. Based on this work, a novel idea was innovatively proposed that the stability of converter blowing could be predicted by the variation in the reflection angle of oxygen jet. This idea was verified in industrial production and achieved good practical application results. Considering the synergistic verification of the research objectives by the characterization techniques, it is of great significance to study the flow characteristics and behavior of liquid steel under the impact of top blowing comprehensively and deeply.
The process of converter top blowing involves mass transfer, heat transfer, chemical reaction and so on. In order to facilitate the numerical simulation experiment of converter, some ideal assumptions about liquid steel should be made before the numerical simulation of converter top blowing.14,15,16,17,18)
(1) It is assumed that the liquid steel flows stably, in-compressible and does not participate in chemical reactions;
(2) Physical parameters such as density, viscosity and specific heat capacity are regarded as constants during the flow of liquid steel;
(3) Ignore the secondary oxidation of the liquid steel surface;
(4) The influence of inclusions on liquid steel flow is ignored;
(5) The influence of foaming of slag layer is ignored;
The physical parameters of each fluid phase in the converter are shown in Table 1.
| Phase | Related parameters | Value |
|---|---|---|
| Slag phase | Density (kg/m3) | 3500 |
| Viscosity (Pa·S) | 0.1 | |
| Specific heat volume (J/(kg·K)) | 1200 | |
| Thermal conductivity (W/(m·K)) | 1.7 | |
| Gaseous phase | Density (kg·m3) | Ideal compressible gas |
| Viscosity (Pa·S) | 1.919×10−5 | |
| Specific heat volume (J/(kg·K)) | 919.31 | |
| Thermal conductivity (W/(m·K)) | 0.0246 | |
| Liquid steel phase | Density (kg/m3) | 7020 |
| Viscosity (Pa·S) | 0.0062 | |
| Specific heat volume (J/(kg·K)) | 750 | |
| Thermal conductivity (W/(m·K)) | 41 |
(1) Fluid Flow Control Equation
The fluid flow in the converter mainly considers the continuity equation, momentum conservation equation, energy transfer equation, etc.19,20,21) For details, see Eqs. (1), (2), (3), (4), (5).
1) Continuity equation:
| (1) |
2) Momentum conservation equation:
| (2) |
3) Energy transfer equation:
| (3) |
Where ui and uj is are the velocity of the fluid in the direction i and j respectively; ρ is the density; p is the pressure; μeff is the effective viscosity coefficient; μe is the effective thermal conductivity; E is energy as a mass-averaged variable.
4) k-ε equation:
| (4) |
| (5) |
Where μt=ρCμk2/ε; μt is the turbulent viscosity coefficient. G is the turbulent energy term; Ym represents the effect of pulsating expansion in compressible turbulence on the overall turbulent dissipation rate. In addition, C1=1.43, C2=1.92, σk=1.0, σε=1.3, Cμ=0.09.
(2) Multi-phase Flow Model Governing Equation
Because the simulation involves gas, slag and liquid steel, and there is no miscibility and reaction among the three. Therefore, in order to clearly know the flow trajectory of the fluid phase, the VOF model is used in this study to track the fluid flow behavior.22,23) When using the VOF model for interface tracking between fluid phases, it should satisfy the following equation:24,25,26)
| (6) |
Where αq is the volume fraction of different phases. According to the formula, the sum of the volume fractions of all the fluid phases is 1. The physical properties of each fluid are calculated from the average volume of the phases. Therefore, the density and viscosity of the fluid phase in the region should meet the following formula:27,28)
| (7) |
| (8) |
Where α is the phase volume fraction, and the subscripts g, l and s represent the gas phase, liquid steel phase and slag phase, respectively.
2.1.3. Geometric Model and MeshingIn this study, the simulation software ANSYS FLUENT 2022 was used to establish and solve the geometric model of the converter. Considering that the gas flow in the oxygen lance is supersonic flow, the fluid flow behavior is closely related to the throat area and the outlet area of the nozzle.29,30,31) Therefore, the requirement of mesh density in the oxygen lance region is particularly important for the converter model. In this study, the mesh of the oxygen lance region is firstly encrypted by means of edge size adjustment, and then the region below the outlet of the oxygen lance is encrypted by means of spatial range adjustment. With the oxygen lance position increasing from 1350 mm to 2350 mm, the overall mesh number is controlled between 1.6 million and 2.3 million, and the mesh quality is greater than 0.84. The corresponding relationship between the quantity and numbers of the encrypted grid under different oxygen lance positions are shown in Table 2. The mesh division diagram of the converter model is shown in Fig. 1. The specific furnace body, oxygen lance structure and parameters are shown in Fig. 2.
| Mesh encryption site | Oxygen lance position (mm) | Mesh numbers (ten thousand) | Mesh quality |
|---|---|---|---|
| Oxygen lance zone and oxygen jet zone | 1350 | 167.6 | 0.845 |
| 1550 | 180.9 | 0.846 | |
| 1750 | 190.5 | 0.846 | |
| 1950 | 199.4 | 0.846 | |
| 2150 | 213.5 | 0.846 | |
| 2350 | 222.5 | 0.846 |
When calculating the fluid flow behavior, the specific boundary conditions are set as follows:
(1) Boundary Conditions of Converter Inlet
The pressure inlet is selected as the boundary condition at the inlet of oxygen lance nozzle. Since the gas is a supersonic compressible gas flow, the inlet pressure value can be calculated according to the Bernoulli equation:32,33,34)
| (9) |
Where P0 is the working oxygen pressure, Pa; P is the furnace pressure, approximately atmospheric pressure, 101325 Pa; k is the specific heat ratio, which is 1.4; Ma is the Mach number,
(2) Converter Outlet, Wall Surface and Solving Boundary Conditions
The upper surface of the converter is used as the pressure outlet, and the rest of the walls adopt the static non-slip boundary condition.35,36,37,38) PISO algorithm was used to solve the problem, and the momentum, turbulent kinetic energy and turbulent kinetic energy dissipation rate were all solved in second-order mode. The speed residuals are controlled below 10−3 and the remaining residuals are controlled below 10−2.
(3) Fluid Phase Boundary Conditions
Each fluid phase distribution region is defined using local initialization. In this study, the converter slag is set as a uniform fluid, and the viscosity of the slag system is 0.1 Pa·s according to the different steel grade.39) The actual steel output of converter is 170 t and steel slag per ton is 85 kg. The density of converter slag is 3200 kg/m3. The thickness of slag layer is calculated according to the body size of converter.
| (10) |
Where M is the total steel mass, t; N is the amount of slag in tons, kg; ρslag is the slag density of converter, kg/m3; r is the radius of the converter body, m. Therefore, the calculated thickness of slag layer is 200 mm. The distribution of each fluid phase is shown in Fig. 3.
In the actual process of converter smelting, the oxygen jet has a great change on the shape of the impact crater, and the oxygen reflection jet after the impact has a great influence on the slag layer of the melt pool.40) Therefore, it is necessary to study the oxygen reflection jet. In this study, the definition of oxygen reflection jet angle and two oxygen reflection jet shapes are introduced. The oxygen reflection jet angle is defined as follows: Taking the oxygen reflection jet velocity of 20 m/s after jet impacting the melt pool and the lowest depth of the impact crater as the reference, taking the end of the oxygen reflection jet as the angle characteristic point, and introducing the oxygen reflection jet angle θ. The oxygen reflection jet shape can be divided into outward expansion type and inward contraction type, see Fig. 4 for details. In this study, in order to describe the oxygen reflection jet angle more accurately, the actual value of θ is the average value of the oxygen reflection jet angle under the five oxygen lance sections.
In this study, two types of oxygen lance deflection angles and two specifications of oxygen lance flow rates commonly used in a steel plant are used as the basis for the design of the simulation research scheme. Combined with the actual working conditions of the converter smelting operation of the steel plant, the specific experimental scheme of the research is considered as shown in Table 3. The relationship between oxygen lance flow rate and oxygen pressure is shown in formula 2.11.41,42,43)
| (11) |
| Experimental scheme | Oxygen pressure (MPa) | Oxygen flow rate (Nm3/h) | Deflection angle | Oxygen lance position (mm) | |
|---|---|---|---|---|---|
| Scheme 1 | 1-1 | 0.86 | 38000 | 12° | 1350 |
| 1-2 | 1550 | ||||
| 1-3 | 1750 | ||||
| 1-4 | 1950 | ||||
| 1-5 | 2150 | ||||
| 1-6 | 2350 | ||||
| Scheme 2 (Original scheme) | 2-1 | 0.86 | 38000 | 13° | 1350 |
| 2-2 | 1550 | ||||
| 2-3 | 1750 | ||||
| 2-4 | 1950 | ||||
| 2-5 | 2150 | ||||
| 2-6 | 2350 | ||||
| Scheme 3 | 3-1 | 0.95 | 42000 | 13° | 1350 |
| 3-2 | 1550 | ||||
| 3-3 | 1750 | ||||
| 3-4 | 1950 | ||||
| 3-5 | 2150 | ||||
| 3-6 | 2350 | ||||
Where q is the single-hole flow rate, Nm3/min; T0 is the inlet temperature, 300 K; CD is 0.95; P0 is the oxygen pressure, MPa; A0 is the oxygen lance throat area, m2.
Figures 5 and 6 are the oxygen jet variation diagram and the central axis (see Fig. 2) velocity distribution diagram respectively. It can be seen from Fig. 3.1 that under the three oxygen lance process parameters in this study, oxygen lance jets all have a relatively obvious coalescing phenomenon, that is, the jets begin coalescence around 0.3 m at the outlet of the oxygen lance, and basically complete coalescence around 0.9 m. Due to the large outlet area of the oxygen lance nozzle and the close distance between each nozzle, there is a pressure difference between the two sides of the high-speed jet after the action of Laval nozzle. This pressure difference caused by turbulent enfranchisement causes the jet to offset, resulting in the mutual attraction of each oxygen strands and the coalescence of the jet. When the jets are completely coalescing, the flow velocity at the central axis is the highest, which is 237.8 m/s, 228.9 m/s and 239.8 m/s under various experimental conditions, and the core length of the jets is 1.67 m, 1.32 m and 1.89 m, respectively. Compared with the existing working parameters of the oxygen lance, it is found that reducing the nozzle declination angle and increasing the oxygen blowing flow rate will lengthen the core length of the jet, thus improving the blowing intensity in the center of the oxygen lance. As can be seen from Fig. 3.2, compared with schemes 1, 2, and 3 has obvious velocity variation in low pressure region. This is because the increase of oxygen flow rate will further reduce the pressure of the low pressure area at the outlet of the oxygen lance, thus forming a local speed mutation. The research data show that from the beginning of jet coalescence to the completion of jet coalescence, the central axis jet velocity of scheme 1 is slightly larger than that of scheme 3, and both are larger than that of original scheme 2. When the jet coalescence is completed, the effect of increasing the oxygen flow rate on the central axis velocity of the jet is obviously greater than the change of the nozzle declination angle. Therefore, in the actual production operation process, it is not possible to improve the intensified blowing within the center range of the oxygen lance only by increasing the oxygen supply intensity or reducing the nozzle deflection angle and other single indicators, and various production indicators should be comprehensively evaluated.
Figure 7 shows the velocity distribution of oxygen jet in cross section. It can be seen from the figure that at the oxygen lance position of 0.6 m and 0.4 m, the jet flow velocity of scheme 2 and 3 at the cross section does not change significantly, while the jet flow velocity of scheme 1 at the center of the cross section is greater than that of scheme 2 and 3 due to the influence of small nozzle deflection angle. In addition, by comparison, it is found that the velocity variation rules of the transverse axis velocity (Y=0) and longitudinal axis velocity (X=0) of the cross section of each scheme are consistent, except that the jet velocity of the two axes will suddenly change at the oxygen lance position of 0.4 m, and the jet velocity around the two axes is not much different at the oxygen lance position of 0.6 m, indicating that the oxygen jet has coalesced at this time. This conclusion is consistent with the results discussed in the previous section.
In order to study the change of fluid velocity after oxygen jet impinges on slag and liquid steel at different oxygen lance positions, the velocity distribution on the transverse axis of the melt pool slag-steel interface (see Fig. 2) was calculated and analyzed, and the results were shown in Fig. 8. It can be seen from the figure that for schemes 1 to 3, the jet velocity at different oxygen lance positions reaches its maximum near X=0 (the center zero position), and the lower the oxygen lance position, the higher the velocity. In the range X=0.5 m–2.66 m, the lower the oxygen lance position, the lower the overall speed. The data show that the central section directly below the oxygen lance (near X=0) is affected by the direct impact of the jet, so the transverse axis velocity distribution in this area is the largest. As the jet diffuses around, the impact kinetic energy weakens, so the transverse axis velocity also decreases accordingly. At X=0, the jet velocities of scheme 1–3 at the lowest and highest oxygen lance positions are 205.1 m/s and 162.3 m/s, 203.8 m/s and 160.6 m/s, 209.5 and 170.1 m/s, respectively. It shows that the impact of jet on the melt pool will be strengthened if the nozzle declination angle decreases or the oxygen lance flow rate increases.
Supersonic jet under different process parameters will produce impact craters with different shape characteristics on the slag-steel interface. The shape of the impact crater indirectly reflects the effective utilization rate of supersonic oxygen jet. Controlling various technical parameters during the blowing process (such as the nozzle deflection angle, oxygen supply intensity, oxygen lance position, etc.) can form a more ideal “funnel” impact crater to avoid the formation of “mantou” foam slag cover. This can improve the slag formation rate of the blowing process, accelerate carbon-oxygen and carbon-phosphorus reaction between the oxygen flow and the impact interface, shorten the blowing time, and strengthen the smelting process reaction. Therefore, in order to accurately describe the impact crater shape formed by the top blowing of the converter at the slag-steel interface of the melt pool, this study conducted three-dimensional visualization characterization of the impact crater shape, as shown in Fig. 9. It can be seen from the figure that under different oxygen lance flow rate and nozzle deflection angle, the depth and fluctuation height of impact crater decrease with the increase of oxygen lance position, and there is an obvious disturbed flow at low oxygen lance position (dashed square in the figure). Therefore, if the low oxygen lance position blowing is carried out for a long time, it is easy to produce liquid steel spatter and even burn the oxygen lance. Compared with experimental schemes 2 and 3, it is found that after increasing the oxygen flow rate, the impact crater depth of scheme 3 is deeper and the interface fluctuation height is larger, which indicates that increasing the oxygen supply intensity can effectively expand the impact crater area, increase the chemical reaction interface, and obtain more favorable dynamic and thermodynamic conditions for the full reaction of liquid steel. Compared with experimental schemes 1 and 2, it can be seen that under the same oxygen flow rate, when the nozzle deflection angle is reduced to 12°, the impact crater shape instability is enhanced under different oxygen lance positions. This is because the reduction of the oxygen lance nozzle deflection angle makes the jet flow flow at the impact crater more concentrated, the impact area is smaller, and the impact energy obtained per unit area is higher, thus making the impact interface volatility and stability worse, which is not conducive to the smooth progress of blowing.
Figures 10 and 11 are the slag-steel distribution diagram and the exposed area of liquid steel on the surface of slag layer respectively, in Fig. 10. It can be seen from the figure that under each scheme, the exposed area of liquid steel shows a trend of first increasing and then decreasing. For scheme 1 to 3, the oxygen lance positions for the maximum exposed area of liquid steel are 1950 mm, 1750 mm and 2150 mm respectively, and the corresponding exposed area of liquid steel is 7.18 m2, 7.08 m2, and 7.42 m2 respectively. According to the analysis, this is because when the oxygen lance position is low, the oxygen jet is more concentrated, and the oxygen reflection jet is difficult to effectively diffuse to the slag layer, resulting in the higher the oxygen lance position, the larger the exposed area of the liquid steel in the slag layer. When the oxygen lance position is raised to a certain height, the oxygen kinetic energy on the surface of the slag layer is not enough to support the continuous outward diffusion of the slag layer, so the exposed area of the liquid steel on the surface of the melt pool will gradually decrease with the continuous increase of the oxygen lance position. In addition, with the increase of oxygen flow rate, oxygen can be diffused to the slag layer better, and the exposed area of liquid steel under each oxygen lance positions is also the largest. When the oxygen lance position is 1350 mm and 2350 mm, the exposed area of liquid steel in scheme 1 to 3 is 6.52 m2, 6.51 m2, 6.92 m2 and 6.45 m2, 6.68 m2, 7.19 m2, respectively.
Figure 12 shows the distribution diagram of gas-slag-steel three-phase interface generated when oxygen jet impinges on the melt pool. It can be seen from the figure that under each experimental scheme, with the increase of oxygen lance position, the covered area of oxygen reflection jet tends to increase significantly, so the exposed area of liquid steel on the surface of the slag layer will also increase with the increase of oxygen lance position, and then with the attenuation of oxygen impact kinetic energy, the oxygen reflection jet energy will also weaken correspondingly, resulting in a decrease in the exposed area of liquid steel. From the shape of oxygen reflection jet, under each scheme, the reflection jet shows inward contraction at low oxygen lance position, the oxygen impact kinetic energy is stronger, and the jet is more concentrated, which is conducive to the rapid melting of slag and the rapid increase of melt pool temperature in the early stage of smelting. With the smelting process, when the oxygen lance position gradually increased, and the reflection jet showed outward expansion, which was conducive to increasing the contact area between the oxygen jet and the liquid steel, and promoting the uniform distribution of the jet velocity in the radial direction of the contact interface. At this time, it is the middle stage of smelting, and the stable operation of the oxygen lance in the 1750–1950 mm interval can quickly carry out the gas-solid-liquid mass transfer, and promote the rapid dephosphorization, decarbonization and heating of the liquid steel. In the later stage of smelting, the impact kinetic energy of jet on the reaction interface can be effectively reduced by raising the oxygen lance position, and the generation of foam slag can be inhibited.
Figure 13 shows the changing characteristics of oxygen reflection jet angle with the change of oxygen lance position. It can be seen from the data in the figure that no matter what kind of oxygen lance blowing method, the variation gradient of oxygen reflection jet angle at low oxygen lance position is large, which indicates that unstable liquid steel fluctuations are very easy to exist at low oxygen lance position (consistent with the phenomenon in Fig. 9). Under each scheme, θ changes from 1350 mm to 1550 mm are the largest, which are 68.0°→54.8°, 67.2°→48.9°, 67.6°→56.8°, respectively. Therefore, in actual production, in order to ensure the smoothness of the converter smelting, it is recommended that the minimum oxygen lance position is not less than 1550 mm, and the operation time under the low oxygen lance position should not be too long. In addition, the oxygen lance position of the above θ changes greatly, then the reflection jet type is very likely to change from inward contraction type to outward expansion type, so when the oxygen lance is adjusted in the low oxygen lance position segment, it is very likely to cause abnormal fluctuations on the impact interface, so frequent operation of the low oxygen lance position should be avoided in actual smelting production. In addition, when the oxygen lance position exceeds 1750 mm, θ values are less than 50° and the change is small, the reflection jet diffusion area is the large, and the dynamic stability of the liquid steel is stable. Combined with the actual production, that is, taking into account the middle and late stage of blowing, at this time, the composition of molten steel in the molten pool, the temperature has been nearly uniform, and the reaction has been quite sufficient, and the upward range of the oxygen lance position at the end of smelting can be referred to 1950 mm, while 1750 mm–1950 mm can be used as the reference oxygen lance position in the middle stage of blowing.
In order to accurately understand the distribution of flow field in the melt pool, the volume quantization of fluids in the low flow zone (0–0.06 m/s) and the active zone (0.06–0.12 m/s) in the melt pool was carried out in this study, as shown in Fig. 14. As far as the influence factors of the oxygen lance nozzle deflection angle are concerned, compared with schemes 1 and 2, it can be seen that when the oxygen lance position is lower than 2150 mm, the volume of the active zone in the melt pool under the action of the oxygen lance jet with large nozzle deflection angle, and the flow field of the melt pool is better. This is because the larger the nozzle deflection angle, the larger the impact coverage area of the melt pool under the same jet strength, the stronger the ability of diffusion mass transfer from the active zone to the low flow zone, so as to accelerate the overall mixing efficiency of the melt pool and shorten the homogenization time of temperature and composition. When the oxygen lance position is greater than 2150 mm, the impact of jet on the kinetic energy attenuation of the impact surface of the melt pool is much greater than the impact caused by changing the nozzle deflection angle, so the melt pool fluid is almost in a low flow zone and the fluid is in a relatively balanced static state, at this time, it can be considered that the oxygen blowing enters the final stage and the smelting is about to end. Under the condition of the same nozzle deflection angle and increasing oxygen supply intensity, it can be found that the volume of the active zone of experimental scheme 3 at any oxygen lance positions is larger than that of scheme 1 and 2, while the volume of the low flow zone on is correspondingly smaller than that of scheme 1 and 2. In the above section, the recommended reference oxygen lance position during the blowing middle stage is between 1750 mm and 1950 mm, the average active zone volume of scheme 3 is about 25% larger than scheme 1 and about 7.15% larger than scheme 2. Therefore, it can be seen that in order to increase the rapid stirring efficiency of the melt pool in a short time, strengthening the chemical reaction conditions of the smelting process, increasing the oxygen supply intensity is more practical than increasing the nozzle deflection angle, and can play a better metallurgical reaction effect.
Under three different oxygen lance process parameters, the effect of oxygen jet on liquid steel smelting reaction in melt pool was fully studied and discussed from the perspectives of impact behavior of oxygen jet, impact crater shape, oxygen reflection jet angle, etc. Based on the theoretical simulation results, the industrial smelting verification tests were carried out for the above three research schemes respectively. The parameters of the 150 t converter oxygen lance during industrial verification were shown in Table 4. The verification test selected X65MS acid resistant pipe steel (C:0.03%–0.05%, P≤0.015%). This steel is low carbon and low phosphorus, which has good HIC (hydrogen induced cracking) and SCC (stress corrosion cracking) resistance, while maintaining sufficient strength and toughness. Therefore, when smelting the steel, it is required to strictly control the content of carbon, oxygen, phosphorus, sulfur and hydrogen in the liquid steel at the end of smelting, and the blowing time needs to be controlled to reduce the total oxygen content in the liquid steel and the total iron content in the slag, further reduce the oxidation of the slag, and ensure low-cost smelting production. According to the actual blowing situation of the site, the three experimental schemes in this study respectively took 50 smelting data from the converter times. The technical indicators such as the oxygen content of liquid steel at the end of smelting, the blowing time and the total iron content of slag were analyzed to verify the accuracy and operability of the theoretical calculation.
| The structural name of oxygen lance | Parameter |
|---|---|
| Outer diameter of the oxygen lance/mm | Φ299×10 |
| Number of nozzle holes | 5 |
| Throat diameter/mm | 43.8 |
| Outlet diameter/mm | 58.1 |
| Mach number | 2.05 |
| Nozzle deflection angle of oxygen lance | 12° and 13° |
Of course, to ensure the accuracy of the industrial test, the changes in the bottom-blowing conditions and processes, as well as the top-blowing oxygen lance process rules during the verification, were consistent with the original scheme, but the average blowing oxygen content was referred to each experimental scheme. During industrial verification, the oxygen lance position process referred to the top-blowing simulation research results, such as: the lowest gun position is not less than 1550 mm, the lance position is adjusted to 1950 mm in the later stage, and the lance position in the middle stage of blowing should be selected at approximately 1750 mm.
Figures 15 and 16 respectively show the distribution of oxygen content in liquid steel and total iron content in slag at the end of smelting under the industrial verification of each experimental scheme. By comparing the two figures, it can be found that the rules of the two technical indicators are consistent. The production data show that the average oxygen content in liquid steel at the end of the smelting process is 669.32 ppm, while the average total iron content in the slag is 19.44%, when the oxygen lance with 12° nozzle deflection angle is used (scheme 1). The normal distribution of data shows that in the random 50 smelting data of smelting production using scheme 1, the sample range of oxygen content in liquid steel at the end of smelting is 472 ppm, the sample variance is 10674.14, and the discrete fluctuation of oxygen content is the largest. This shows that the small nozzle deviation angle will brings large random fluctuation to smelting production, and it is difficult to control the oxygen lance through the blowing process. In order to restrain the splashing of liquid steel and prevent the occurrence of “back-drying”, the operator has to frequently change the oxygen lance position and adjust the oxygen pressure. This operation mode eventually leads to the maximum fluctuation of oxygen content of liquid steel at the end of smelting, which can not achieve efficient and stable production. The wide fluctuation of actual production data from the impact crater shape change characteristics of the simulation results have also been consistent verification.
The comparison of experimental schemes 1 and 2 shows that when the oxygen lance nozzle deflection angle is appropriately increased to 13°, the average oxygen content in liquid steel at the end of smelting is 646.06 ppm, while the average total iron content in the slag is 18.53%, these two technical indicators are reduced compared with scheme 1, but the decrease is not significant, and the average total iron content is only reduced by 0.91%. This is mainly because there are many factors affecting the total iron content in slag, such as the instability of smelting raw materials, the instability of liquid steel composition, the carbon content of liquid steel at the end of smelting and the comprehensive stirring effect of liquid steel, etc. Therefore, it can be seen that only increasing the oxygen lance nozzle deflection angle can not have a substantial effect on reducing the total iron content in the slag. which can also be effectively verified by the simulation results of the effect of oxygen lance jet on the liquid steel flow field in the melt pool.
Comparative analysis of experimental schemes 2 and 3 found that when the oxygen supply flow rate was further increased to 42000 Nm3/h, the average oxygen content of liquid steel at the end of smelting was reduced from 646.06 ppm to 573.28 ppm, while the average total iron content in the slag was reduced from 18.53% to 16.35%, and the technical index was further improved, and the decline rate of the comprehensive index has increased significantly. From the two figures, it can be found that the fluctuation range of technical indicators of scheme 3 has been significantly narrowed, and the sample range of liquid oxygen content of steel in random 50 furnace data is only 253 ppm, and the sample range of total iron content in slag is only 3.25%, which is nearly 50% less than that of experiment scheme 1, and the random 50 smelting data show that the sample range of oxygen content in liquid steel at the end of smelting is only 253 ppm, and the sample range of total iron content in slag is only 3.25%, which is nearly 50% less than that of experiment scheme 1. According to the analysis, this is because increasing the oxygen supply intensity can effectively shorten the oxygen blowing time, strengthen the interaction between the oxygen jet and the liquid steel in the melt pool, further accelerate the decarburization reaction rate of the liquid steel, accelerate the slag formation rate of the melt pool, and rapidly increase the temperature of the melt pool, so as to achieve the goal of strengthening the converter melting.
In summary, the simulation results of this study can form a good verification relationship with the results of industrial production, and the simulation results can effectively guide industrial production. In this study, under the premise of constant bottom blowing and other smelting conditions, by increasing the nozzle deflection angle of the high-strength and high-flow oxygen lance to 13° and increasing the oxygen supply flow to 42000 Nm3/h, the average oxygen content in liquid steel at the end of smelting can be controlled to 573.28 ppm when smelting low carbon and low phosphorus steel, and the average carbon-oxygen concentration product in the liquid steel is 0.0023. This is beneficial to reduce alloy consumption, reduce the formation of inclusions during deoxidation, and improve the purity of liquid steel.
Furthermore, based on the simulation results and industrial verification results, it is not difficult to conclude that even under the same bottom-blowing process conditions, there are certain differences in the smelting effects of the converter under different working conditions, and this is mainly caused by the changes in the oxygen-blowing jet. Therefore, under the same basic conditions, stabilizing the oxygen-blowing jet through reasonable adjustment of the oxygen-blowing volume and the lance position is of great significance for actual smelting.
(1) Under the conditions of this study, the supersonic oxygen lance jet has a serious coalescing phenomenon, the oxygen jet begin coalescence around 0.3 m at the outlet of the oxygen lance, and basically complete coalescence around 0.9 m. When the jet coalescence is completed, the effect of increasing the oxygen flow rate on the central axis velocity of the jet is obviously greater than the change of the nozzle declination angle. In the actual production process, it is not possible to increase the blowing intensity within the center range of the oxygen lance only by increasing the oxygen supply intensity or reducing the nozzle deflection angle and other single indicators, and various production indicators should be comprehensively evaluated.
(2) Compared with the three experimental schemes, it can be seen that when the oxygen lance nozzle deflection angle is 12°, under different oxygen lance positions, the jet flow is more concentrated on the impact crater, the impact impact area is smaller, the impact crater shape instability is enhanced, and the impact interface fluctuation is the largest. When the oxygen flow rate is increased, the impact crater depth of scheme 3 is the deepest and the interface fluctuation height is the largest, which can effectively expand the impact crater area, increase the reaction interface, and obtain more favorable dynamic and thermodynamic conditions for the reaction of liquid steel. Under the three experimental schemes, the oxygen lance positions of the maximum exposed area of liquid steel are 1950 mm, 1750 mm and 2150 mm respectively, and the corresponding exposed areas are 7.18 m2, 7.08 m2 and 7.42 m2 respectively.
(3) Under the three experimental schemes, when the oxygen lance position changed from 1350 mm to 1550 mm, the oxygen reflection jet angle changed the most, which was 68.0°→54.8°, 67.2°→48.9°, 67.6°→56.8°, respectively. When the oxygen lance position exceeds 1750 mm, the reflection jet angle changes are less than 50°, and the reflection jet diffusion area is the large, and the dynamic stability of the liquid steel is stable. Combined with the actual production, 1750 mm–1950 mm can be used as the reference oxygen lance position in the middle stage of the blowing process, at this time, the composition of liquid steel in the melt pool, the temperature has been nearly uniform, and the reaction has been quite sufficient, and the final oxygen lance position can be referred to 1950 mm.
(4) The industrial test shows that the simulation results can guide the actual production well. When using a small nozzle deflection angle oxygen lance, it is difficult to operate the oxygen lance, and the oxygen content in the liquid steel at the end of smelting is fluctuates greatly, so efficient and stable production can not be achieved. In this study, under the premise of constant bottom blowing and other smelting conditions, by increasing the nozzle deflection angle of the high-strength and high-flow oxygen lance to 13° and increasing the oxygen supply flow rate to 42000 Nm3/h, the average oxygen content in liquid steel at the end of smelting can be controlled to 573.28 ppm and the average total iron content in the slag is 16.35% when smelting low carbon and low phosphorus steel.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
The authors would like to express their appreciation to the National Natural Science Foundation of China (No. 52074001); University Natural Science Research Project of Anhui Province (KJ2020ZD25).
