Study on dehydrogenation behaviour of molten steel in single snorkel refining furnace (SSRF) by a mathematical model

Abstract

A mathematical model of an industrial Single Snorkel Refining Furnace (SSRF) has been developed to investigate dehydrogenation behaviour in the vacuum refining process. Dehydrogenation reactions were considered to take place at three sites: Ar bubble surface, bath surface, and the bulk steel. The effects of Ar bubble behaviour and H2 heterogeneous nucleation behaviour of the bulk steel on dehydrogenation were taken into account specifically. The simulation results are in good agreement with the measured industrial data. Using the model, the dehydrogenation rate at different sites and its contribution to total dehydrogenation value were evaluated in detail. The results indicated that bulk steel dehydrogenation occurred only the high concentration range, while the other two reactions participated in the whole refining process. In addition, the effect of argon flow rate on the dehydrogenation rate was also obtained and the reasonable range of argon flow rate for SSRF treatment has been recommended.

List of symbols

Cross-sectional area of the bath surface (m²)

A _Ar, A _s

Surface area of Ar bubble, area of bath surface (m²)

Heat capacity of Ar gas (5.22×10³ J kg⁻¹ K⁻¹)

D _H

Diffusion coefficient of dissolved hydrogen in the melt (m²s⁻¹)

d _n

Hole diameter of the porous plug (m)

d _V

Diameter of bath surface in the vacuum vessel (m)

Fr _m

Froude number

f _H

Activity coefficient of hydrogen in molten steel

Argon flow rate (L min⁻¹, m³s⁻¹)

Acceleration of gravity (m s⁻²)

H _melt

Distance from the porous plug to bath surface (m)

H _in

Critical depth of H₂ nucleation in the bulk steel (m)

h _b, h _H2

Distance from the position of Ar bubble, H₂ nucleation site to the bath surface (m)

Elemental component ‘j’ dissolved in molten steel

Dehydrogenation rate at Ar bubble surface of the ladle, the vacuum vessel (kg s⁻¹)

Dehydrogenation rate at Ar bubble surface, bath surface and the bulk steel (kg s⁻¹)

Equilibrium constant of the dehydrogenation reaction

k _H, k _S

Mass transfer coefficient of hydrogen to Ar bubble surface, bath surface (m s⁻¹)

k _in

Process parameter of inner steel dehydrogenation (s⁻¹)

m (1, 2 … N _L)

Serial number of bubble from the bottom of the ladle to bath surface of the ladle

N _L , N _V

Number of Ar bubbles in the ladle, the vacuum vessel

Number of moles of H₂, Ar gas in the Ar bubble (mol)

Number of moles of H₂, H₂+Ar gas escaped from molten steel per time (mol)

n (1, 2 … N _V)

Serial number of bubble from bath surface of the ladle to that of the vacuum vessel

p (1 … N _L + N _V)

Serial number of bubble from the porous plug to bath surface in the vacuum vessel

P ^θ

Standard atmospheric pressure (101,325 Pa)

P ₀, P _b

Pressure at the porous plug, Ar bubble inside (Pa)

P _V, P _S

Pressure of the vacuum vessel, surface tension of H₂ bubble (Pa)

Equilibrium H₂ partial pressure in Ar bubble, the vacuum vessel, the bulk steel (Pa)

Circulation flow rate of molten steel (kg s⁻¹)

Ideal gas constant (J mol⁻¹K⁻¹)

Initial Ar bubble radius, Ar bubble radius in the rising process (m)

The first, second derivative of Ar bubble radius to time (m s⁻¹, m s⁻²)

Nucleation radius of H₂ gas in the bulk steel (m)

Cross-sectional area of the activation area (m²)

T ₀ , T, T _melt

Temperature of Ar gas at the porous plug, Ar gas in the ascending process, molten steel (K)

u _slip

Ascending velocity of Ar bubble (m s⁻¹)

V _g

Volume of a bubble (m³)

W _L, W _V

Weight of molten steel in the ladle, the vacuum vessel (kg)

Greek Letters

Density of molten steel and density of Ar gas (kg m⁻³)

Constant (α = 30)

Surface tension of molten steel (1.8 N m⁻¹)

Viscosity of molten steel (5.8×10³ N m⁻²)

Heat transfer coefficient of Ar gas (J (m² s K)⁻¹)

Activated coefficient of the bath surface

Upward cone angle of the gas plume

Introduction

With the increasing demands for high-purity steel, the vacuum degasser ratio of the converter and electric furnace steel has been improved rapidly, and various vacuum refining technologies are developing continually. Currently, RH degasser has been worldwide applied and shown good degassing performance in producing special steel. However, the refractory materials of immersed snorkels and its vacuum vessel are seriously corroded by liquid steel circulation in RH environment, and their service life is short. In order to cope with the problem, the single snorkel refining furnace (SSRF) has been developed as a new type of vacuum refining equipment since the 1970s through the reformation of RH degasser [1].

Figure 1 depicts the structure diagram of the SSRF. During its refining process, molten steel is pumped into the vacuum vessel under reduced pressure after the snorkel is immersed into molten steel of the ladle, and Ar gas is eccentrically injected into molten steel at the ladle bottom. Driven of the ascending Ar bubbles, molten steel rises from one side of the immersed snorkel and falls on the other side, thus establishing a circulation system between the ladle and the vacuum vessel. As the birth of the SSRF, various physical experiments and numerical simulations have been carried out. The reaction mechanism of SSRF in decarburization, deoxidation, and desulphurization, as well as the process parameters of SSRF, such as the flow field, mixing characteristics, circulation flow rate, blowing position, and other, was studied theoretically [2–9].

Figure 1.

Schematic illustration of SSRF.

Now the new-type refining equipment has been put into actual industrial production and achieved remarkable results. In Japan, Nippon Steel has independently developed a similar vacuum degasser called revolutionary degassing activator (REDA) since the 1990s [10]. According to industrial tests [11], the final carbon content of 175 ton REDA decreased to 4–6 ppm in 20 min and that of 350 ton REDA decreased to 3 ppm in 30 min. In China, a lot of industrial tests have been conducted using 80 tons SSRF in Shanxi Taigang Stainless Steel Corporation Limited. Based on the decarburization tests [12], the results showed that the final carbon content of 10 ppm has been achieved in 20 min. The desulphurization tests [13] confirmed that the maximum desulphurization ratio of 81.2 pct has been achieved and the sulphur content of molten steel can be reduced to about 10 ppm. Likewise, the maximum dehydrogenation ratio of 83.7 pct has been achieved and the hydrogen content of molten steel can be reduced to about 1.2 ppm [13]. However, up to now, the research results on the dehydrogenation of SSRF are limited to the industrial tests, and there are few detailed reports on its mechanism. Besides, it is well known that the control of dissolved hydrogen in steel is of special importance in the vacuum refining process. Once the hydrogen content at the end of refining exceeds 2 ppm, many defects, such as white spot, shrinkage cavities, and crack, may be exposed in the steel products [14,15].

In order to clarify the dehydrogenation mechanism of SSRF, a dehydrogenation model based on different reaction sites was proposed. The modelling concept based on reaction sites was originally used to simulate the decarburization process and has been widely accepted now [7,16]. Yu and Louhenkilpi [17] made a mathematical simulation of dehydrogenation in VD treatment, and found that the dehydrogenation rate of Ar bubble surface was much higher than that of the bath surface due to the larger gas–liquid interface area of Ar bubbles. The terminal hydrogen level of 1.77 ppm was obtained under a pressure of 100 Pa. Zhu [18] pointed out that dissolved hydrogen removal in RH treatment occurred at three places: Ar bubble surface, bath surface, and the bulk steel in the vacuum vessel. Considering that less H₂ gas was generated, bulk steel dehydrogenation was also neglected in his paper. The dehydrogenation reaction in the bulk steel is not occur theoretically until the vacuum pressure drops a certain value and dissolved hydrogen content exceeds 2.5 ppm [19,20]. However, these researches on the mechanism of dehydrogenation reaction only involved Ar bubble surface and bath surface, and there is no systematic explanation for the heterogeneous nucleation of H₂ in the bulk steel.

In the study, a mathematical model that the dehydrogenation reactions take place at Ar bubble surface, bath surface, and the bulk steel is developed to investigate the dehydrogenation behaviour in SSRF refining process. The calculated results are in good agreement with the industrial test results. Based on the model, the dehydrogenation behaviour of each reaction site is explained in detail, and the reasonable argon flow rate is proposed for the dehydrogenation of SSRF.

Description of the dehydrogenation model

Thermodynamics of dehydrogenation reaction

The dehydrogenation reaction in SSRF can be expressed as follows: (1)

According to Sievert’s law [21], the relationship between the equilibrium concentration of dissolved hydrogen and the partial pressure of hydrogen gas can be expressed by the below equation (2)

In the above reaction, the activity coefficient is calculated by the below equation (3)

where

and [pct j] represent the interaction coefficient and mass per cent content of element ‘j’ in molten steel, respectively. Table 1 lists the composition of alloy elements in liquid steel before vacuum treatment, including the corresponding interaction coefficients for dissolved hydrogen [15,21]. The dimensions of the 80 ton SSRF system are shown in Table 2.

Table 1.

Compositions of molten steel and interaction coefficients for dissolved hydrogen.

Element	C	Mn	Si	P	S	Cr	Al	Ti
pct j	0.3	0.87	0.47	0.012	0.005	1.85	0.03	0.02
	0.06	−0.001	0.027	0.011	0.008	−0.002	0.013	0.08

Table 2.

Dimensions of the 80-ton SSRF system.

Parameters	Values/mm
Upper diameter of ladle	2965
Lower diameter of ladle	2305
Height of ladle	3600
Diameter of snorkel	1000
Height of snorkel	880
Immersion depth of snorkel	400
Diameter of vacuum vessel	1300
Height of vacuum vessel	8700

Reaction model at each dehydrogenation site

In the present study, it is assumed that dehydrogenation reactions take place at three sites in SSRF: Ar bubble surface, bath surface, and the bulk steel. The schematic diagram of each dehydrogenation site is illustrated in Figure 2. Once Ar gas is injected into the molten steel, Ar bubble surface dehydrogenation occurs when the dissolved hydrogen atoms diffuse to the bubble interface, and the generated H₂ gas enters into Ar bubble. The dehydrogenation process proceeds continuously as the bubbles rise from the porous plug to the bath surface. Similarly, bath surface dehydrogenation takes place when dissolved hydrogen atoms combines with each other to generate H₂ gas at bath surface, and escapes from the vacuum vessel with Ar bubbles soon. The bath surface is constantly updated due to the gas stirring and vacuum pumping, which is conducive to the dehydrogenation reaction here. In addition, the dehydrogenation reaction in the bulk steel is also considered at low-pressure level, in which a large number of H₂ bubbles will be formed near the bath surface.

Figure 2.

Schematic diagram of three reaction sites for dehydrogenation of SSRF.

Noticeably, both bath surface dehydrogenation and bulk steel dehydrogenation only occur in the vacuum vessel, while Ar bubble surface dehydrogenation takes place not only in the vacuum vessel but also in the ladle. For the convenience of calculation, the dehydrogenation rates of each reaction site were calculated separately by their respective reaction models, and their sum was taken as the total dehydrogenation rate of SSRF.

Dehydrogenation at Ar bubble surface

Ar bubble surface is an important reaction site for dehydrogenation in SSRF vacuum refining. Ar bubbles injected from porous plug exist in the form of bubble plume in molten steel and provide a large superficial area for dehydrogenation reaction. Also, low H₂ partial pressure in these bubbles signiﬁcantly can effectively accelerate the dehydrogenation reaction at Ar bubble surface. Here, dehydrogenation process mainly includes the following steps: first, mass transfer of dissolved hydrogen to gas–liquid reaction interface; second, chemical reaction occurs at bubble surface; third, generated H₂ gas gets into the gas phase. As pointed out by Ootsuka et al. [15], the mass transfer of dissolved hydrogen (the first step) was regarded as the rate-determining step of Ar bubble surface dehydrogenation. Moreover, Ar bubbles are assumed to spherical and do not coalescence or break when rising in the model.

The reaction rate (mass per cent s⁻¹) through Ar bubbles in the ladle is expressed by Equation (4). And the equilibrium hydrogen content at Ar bubble surface in the ladle is calculated by Equation (5) (4) (5) The reaction rate (mass per cent s⁻¹) through Ar bubbles in the vacuum vessel is expressed by Equation (6), in which its equilibrium hydrogen content at Ar bubble surface is calculated Equation (7) (6) (7) Based on Higbie's theory [22], the mass transfer coefficient of dissolved hydrogen to the reaction interface is expressed by the below equation (8) The diffusion coefficient of dissolved hydrogen in liquid steel [23] can be estimated by the below equation (9) According to Levich's researches [24], it is known that the rising velocity of Ar bubble (u _slip) is not a function of the diameter due to the shape deformation from sphere to disc once the bubble radius more than 0.001 m. However, it is assumed that the bubble shape is spherical for simplicity in the model, but u _slip is still estimated by the below equation from his discussion (10) The partial pressure of H₂ in Ar bubble is calculated by the below equation (11) where is a constant value and is calculated by the total dehydrogenation amount that takes place at Ar bubble surface during its rising process. The pressure in Ar bubble is determined by the static pressure of molten steel, pressure in the vacuum vessel and additional pressure, which can be obtained by the below equation (12) Dehydrogenation reaction area of a single bubble is obtained by the below equation (13) On the basis of Szekely’s study [25], the size of Ar bubbles under thermal expansion and pressure drop is estimated by the below equation (14) To solve the above equation, the temperature of the Ar bubble (T) [26] was estimated as follows: (15) where is the heat transfer coefﬁcient (J m⁻² s⁻¹ K⁻¹) derived from the penetration theory with the assumption of gas-phase control [27].

The initial radius of Ar bubble is assumed to be 0.006 m in the model. Figure 3 depicts the variation of Ar bubble radius with height from the porous plus at t=360 s. It can be seen that the Ar bubble size gradually increases from 0.006 to 0.034 m as the bubble ascends from the plug to bath surface. Moreover, the growth rate of Ar bubble in the vacuum vessel is higher than that in the ladle.

Figure 3.

Variation of r _b with height from the porous plus.

Figures 4 and 5 represent the variation of , , and the mole fraction of H₂ in Ar bubble with height from the porous plug at t=360 s, respectively. When Ar bubbles was just injected into molten steel, the dehydrogenation reaction at Ar bubble surface proceed rapidly, leading to the increase in and immediately. Instead, the change of appeared a linear decreasing trend in the whole ﬂoating process. Subsequently, increased to ∼1400 Pa and remained steady until the bubbles escaped from the ladle to the snorkel, forecasting drop rate was almost equal to the increase rate of at the stage. However, once Ar bubbles entered the vacuum vessel, the dehydrogenation rate at the Ar bubble surface increased rapidly due to the sharp expansion of Ar bubble radius, which caused a rapid increase in and a slow decrease in .

Figure 4.

Variations of P _H2,b, P _b, with height from the plug.

Figure 5.

Variations of H2 mole fraction in Ar bubble with height from the plug.

Based on the above calculations, the dehydrogenation rate at Ar bubble surface in the ladle and the vacuum vessel can be obtained. Therefore, the total dehydrogenation rate (kg s⁻¹) at Ar bubble surface is also calculated by the below equation (16)

Dehydrogenation at bath surface

Bath surface represents the contact interface between molten steel and its upper gas phase in the vacuum vessel. It is another important dehydrogenation reaction site. As the dehydrogenation principle at the bath surface is similar to that of Ar bubble surface, the mass transfer of dissolved hydrogen to the reaction interface is considered to be the rate-determining step of dehydrogenation at bath surface. Here, the reaction rate (mass per cent s⁻¹) through the bath surface in the vacuum vessel is expressed by the below equation (17) The equilibrium hydrogen content at the bath surface is calculated by the below equation (18) where the mass transfer coefficient (k _S) of dissolved hydrogen is taken as 9.7×10⁻⁴ m s⁻¹ [28].

The effective surface area (A _S) can be divided into two sections in the bubble plume region: non-activated zone and activated zone, as shown in Figure 6, which is expressed by the below equation (19) where (A−S) represents the area of the non-activated area and its value can be calculated by geometry, and ξ represents the activity coefficient of the activated area . However, there is no agreement on the activity coefficient in many studies. Takahashi et al. [29] studied the rate of nitrogen desorption from molten steel by Ar injection together with blowing in a crucible and reported the effective reaction area was 2.1 times as large as the cross-sectional area. Kitamura et al. [30] pointed out that the activated coefficient was 10.51 based on the gas adsorption and desorption experimental in liquid steel-N₂ system. Hence, the average value is determined as the activated coefficient of the activated area in the present model.

Figure 6.

Activated zone and non-activated zone at bath surface of SSRF.

According to Krishna Murthy’s study [31], the angle of gas plume zone (θ) can be estimated by Equation (20), and the modified Froude number is calculated by Equation (21) (20) (21) The partial pressure of H₂ at bath surface can be calculated as follows: (22) where is the mole of H₂ gas escaped per unit time and its value is determined by the reaction amounts of H₂ gas at three reaction sites. n _total is the sum of number of moles of H₂ and Ar gas escaped per unit time at the bath surface.

Based on the above calculations, the dehydrogenation rate (kg s⁻¹) at the bath surface in the vacuum vessel is obtained by the below equation (23)

Dehydrogenation in the bulk steel

Bulk steel dehydrogenation's first concern is to how H₂ gas bubbles nucleate and grow up, because there is no initial reaction surface similar to the bath surface or the Ar bubble surface. According to Ref. [20], the dehydrogenation reaction in the bulk steel is unlikely to occur through the homogeneous nucleation of H₂ bubbles in liquid steel system, in which a surprisingly large gas super-saturation pressure of about 10⁴ atm is required. Instead, the H₂ bubbles formed on the surface of refractories or granular slag are feasible in energy at a very small super-saturated pressure. Furthermore, it is assumed that H₂ bubbles have been generated in the bulk steel, but the minimum pressure in the bubbles should be at least 0.01 atm if the generated H₂ gas is to be kept intact in the light of Winkler's study [20]. Therefore, the heterogeneous nucleation of H₂ bubbles can only be formed near the bath surface in the vacuum vessel, and the vacuum pressure is regarded as the rate-determining step of dehydrogenation in the bulk steel.

The necessary conditions for H₂ nucleation in the bulk steel can be expressed by as follows: (24) The reaction rate (mass percentage s⁻¹) through forming of H₂ bubbles in the bulk steel is expressed by Equation (25), and the equilibrium content of hydrogen in the bulk steel is calculated by Equation (26) (25) (26) where the process parameter (k _in) is taken as 1.0 after fitted.

In Equation (25), the integral formula on the right indicates that bulk steel dehydrogenation is a volume reaction within the critical depth (H _in). As the upper limit of the integral formula, H _in should be greater than zero and can be calculated as follows: (27) Based on the above calculation, the dehydrogenation rate (kg s⁻¹) in the bulk steel can be obtained by the below equation (28)

Dehydrogenation process in SSRF

In the SSRF system, the concept of the dehydrogenation process can be described as: the dissolved hydrogen content of the vacuum vessel is different from that of the ladle during the refining period because of the different removal rate between these two regions. However, it is assumed that the distribution of dissolved hydrogen is uniform at the initial stage. With continuous circulation and mixing of molten steel, the difference of hydrogen content between the ladle and the vacuum vessel will not be expanded but remain relatively stable.

The circulation flow rate of molten steel (Q) is one of the key links for the dehydrogenation process, which can be used to reflect the fluidity and uniformity of molten steel. Increasing Q can effectively promote the circulating flow of molten steel and the ability of vacuum degassing. According to an empirical formula [7], Q (kg s⁻¹) is estimated by the below equation (29) Figure 7 represents the material balance in the ladle and the vacuum vessel during SSRF reﬁning process. The mass balance relationship of dissolved hydrogen can be expressed by the below equations (30) (31) where the dehydrogenation rates at the Ar bubble surface, bath surface, and the bulk steel have been obtained from Equations (16), (23), and (28), respectively.

Figure 7.

Material balance in the dehydrogenation process of SSRF.

Program and computation process

After establishing the sub-models for dehydrogenation reactions that take place at Ar bubble surface, at the bath surface, and in the bulk steel, the mathematical simulation of the dehydrogenation process in 80 ton SSRF is performed in MATLAB 2017b on a personal computer. The calculation program chart of dehydrogenation process during SSRF treatment is shown in Figure 8.

Figure 8.

Calculation program chart for the dehydrogenation process in SSRF.

Result and discussion

Model validation

The dehydrogenation reaction was calculated for the 80 ton SSSF process using this mathematical model. The operating conditions of the model are consistent with those of the plant (Shanxi Taigang Stainless Steel Corporation Limited). Figure 9 depicts the variation of the relevant operating parameters (P _V and G _Ar) during the actual industrial conditions. The pressure drops rapidly from atmospheric pressure and then reaches a stable deep vacuum state (usually under 100 Pa). And argon flow rate increased rapidly before the beginning of the deep vacuum pressure and stabilized at 320 NL min⁻¹ after 2 min, which lasted for the remaining period. The hydrogen content in molten steel was sampled by a Hydrogen Reading Immersion System (HYDRIS) probe before and after 18 min degassing. The sampling position was 30 cm below the slag of the ladle. Besides, the temperature of molten steel varies from 1893 to 1833K during 18 min degassing, and the effect of slag on dehydrogenation is ignored.

Figure 9.

Variation of P _V and G _Ar with time.

To verify the prediction capability of the dehydrogenation model, six heats of initial hydrogen contents varying between 4.1 and 6.2 ppm were analysed, as shown in Table 3.

Table 3.

Variations of hydrogen content between calculated and measured values before and after 18 min degassing.

Initial H content, ppm	Final H content, ppm
Initial H content, ppm	Measured	Calculated
4.1	0.9	1.04
4.2	1.2	1.05
4.3	0.9	1.07
4.7	1.2	1.20
5.4	1.1	1.30
6.2	1.3	1.32

The comparison of calculated and measured hydrogen content before and after 18 min degassing is displayed in Figure 10. The results showed that the measured values lied close to the model predictions, with data points falling within ±0.2 ppm of the 1:1 line, at which the calculated and measured values were exactly equal. Within the allowable error range, the simulation results are in good agreement with actual production data, indicating that dehydrogenation behaviour in SSRF could well explain by the model. Besides, the subtle difference between the two sets of data may be due to the change of weather condition, hydrogen absorption on slag surface, or fluctuation of operating parameters (e.g. gas flow rate) in actual production, which is inevitable.

Figure 10.

Comparison of calculated and measured final hydrogen content before and after 18 min degassing.

Analysis of SSRF dehydrogenation process

Using the mathematical model, the dehydrogenation behaviour of SSRF during vacuum treatment can be analysed. Figure 11 shows for two example heats, the results show that with the increase in refining time, the variation tendency of dissolved hydrogen concentration first accelerates and then slowed down. It is also found that the initial value of hydrogen has no significant effect on the residual content after vacuum treatment, which is basically consistent with previous work of other steelmaking process [32].

Figure 11.

Variations of hydrogen content for heat A and heat B.

Figure 12 depicts the variations of dehydrogenation rate at each reaction site in heat A. From the beginning to t=120 s, the dehydrogenation rate of Ar bubble surface and bath surface increases rapidly with the increase in argon flow rate, and then decreases gradually. However, the dehydrogenation rate of the bulk steel increases rapidly from zero, and catches up with that of the bath surface at t=60 s and peaks at 1.04×10⁻⁴ kg s⁻¹. Thereafter, the dehydrogenation rate in the bulk steel decreases rapidly and decreases to zero at t=340 s, because of the current vacuum pressure (100 Pa) limiting H₂ nucleation in the bulk steel when [pct H]_L<0.000295. From t=340 s to the end of SSRF treatment, the dissolved hydrogen concentration is at a low level, and the dehydrogenation reaction mainly takes place at Ar bubble surface and the bath surface. In addition, Ar bubble surface dehydrogenation can be divided into two regions: the ladle and the vacuum vessel, of which the dehydrogenation rates are plotted in Figure 13. Comparing the two curves, the dehydrogenation rate through Ar bubbles in the vacuum vessel is much higher than that in the ladle because of a rapid expansion of Ar bubble after entering the vacuum vessel.

Figure 12.

Variations of dehydrogenation rate at each site with time.

Figure 13.

Dehydrogenation rates of Ar bubble surface in ladle and vacuum vessel for heat A.

Figure 14 shows the variation of H_in and [H]_in with the decrease in [pct H]_L in heat B. It can be seen that both of H_in and [H]_in increase rapidly from the beginning, reaching to a peak value at t=60 s, and then gradually reduces to zero. This indicates that the variation tendency of H_in and [H]_in is basically the same. Besides, there is a platform for H_in with the decrease in [pct H]_L from 0.00060 to 0.00053. According to Equation (27), H_in is greatly affected by [pct H]_V and P _V. The reason for the transition from the peak to the platform of H_in is that the drop rate of P _V falls behind in that of [pct H]_V. Until [pct H]_L<0.000294, the H₂ nucleation conditions became no longer be satisfied with Equation (24), thus the dehydrogenation reaction in the bulk steel would be too difficult to continue in SSRF.

Figure 14.

Variations of H_in and [H]_in in with time for heat B.

Figure 15 depicts the contributions of each dehydrogenation rate to the total dehydrogenation rate. It can be seen that the contribution of dehydrogenation at Ar bubble surface far exceeds those of the bulk steel and bath surface, and accounting for over 60 pct of total removal accumulation. The contribution of bulk steel dehydrogenation increases rapidly from the beginning, and slides into a decreasing after reaching the maximum mass percentage of 16.9 pct at t=88 s. Instead, the contribution of bath surface dehydrogenation is inversely proportional that of the bulk steel. At the end degassing, the dehydrogenation amount of Ar bubbles accounted for the highest proportion (∼65 pct) of the total, and that of bath surface was about 30 pct, while that of the bulk molten steel is the least, only about 5 pct.

Figure 15.

Percentage of accumulated dehydrogenation rate at each site for heat A.

In overall, the reactions of Ar bubble surface and bath surface play a dominant role in the whole dehydrogenation process of SSRF. Compared with their contribution, dehydrogenation through heterogeneous nucleation of H₂ bubbles in the bulk steel contributes little to the dehydrogenation process. It only occurs in the range of high hydrogen content (or early stage of dehydrogenation).

As mentioned before, increasing argon flow rate cannot only expand the total surface area of Ar bubbles and activated area of bath surface (as shown in Figure 6), but also accelerate the mass transfer of dissolved hydrogen to their reaction interface through enhancing the circulating flow rate. Thus, the effect of the argon flow rate on the dehydrogenation of SSRF was investigated. As shown in Figure 16, the hydrogen level before vacuum degassing has 4.7 ppm values which gets reduced to 0.9–2.3 ppm after vacuum degassing during 18 min degassing at constant pressure. And the operating condition to achieve hydrogen level <1.2 ppm was satisfied with the argon flow rate of 320–400 NL min⁻¹. Furthermore, an excessive argon flow rate can improve the dehydrogenation efficiency of SSRF, but it might lead to the corrosion of the snorkel refractory as well as the increase in argon cost.

Figure 16.

Variations of hydrogen content under different argon flow rate.

Conclusion

In the present study, the dehydrogenation model that take place at Ar bubble surface, bath surface, and the bulk steel is developed to investigate the dehydrogenation behaviour in SSRF vacuum treatment. The calculated results were compared with the industrial measured data. And the contribution of hydrogen removal at three sites and the effect of argon flow rate on its removal efficiency were studied. The conclusions are as follows:

The predicted results of the final hydrogen content of six heats are in good agreement with the experimental results (the deviation <0.2 ppm). Hence, the dehydrogenation behaviour in an industrial SSRF can be well explained using the present model.

Both Ar bubble surface dehydrogenation and bath surface dehydrogenation are involved in the whole refining period. However, bulk steel dehydrogenation only occurred in the range of high hydrogen concentration (or early stage), which would stop once the dissolved hydrogen content dropped below 2.95 ppm.

Throughout the dehydrogenation process, the contribution rate of Ar bubble surface dehydrogenation is the largest, about 65%, that of bath surface dehydrogenation is 30%, while the contribution rate of bulk steel dehydrogenation is only 5%.

The dehydrogenation efficiency of SSRF can be significantly improved by increasing argon flow rate. Under the argon flow rate of 320–400 NL min⁻¹, the hydrogen level got reduced to <1.2 ppm after vacuum treatment for 18 min. Additionally, an excessive argon flow rate is not desirable, because it might cause the corrosion of the snorkel refractory as well as the increase in argon cost.

Footnotes

Acknowledgements

The authors would like to thank the National Natural Science Foundation of China (grant number 51674024, 51874034) for the financial support.

Disclosure statement

No potential conflict of interest was reported by the author(s).

References

Zhang

The test results of vacuum argon blowing in single snorkel refining furnace. Dalian Special Steel. 1978;3:5–26.

Cheng

Zhang

Yang

et al. Dynamic model of deoxidation for bearing steel treated with single snorkel refining furnace. Dalian Special Steel. 1994;15:22–25.

Qin

Zhu

Cheng

et al. A study on bubble behavior during vacuum treatment of refining furnace with single snorkel and metallurgical effect. Special Steel. 2010;31:18–21.

Yang

Zhang

Wang

et al. Mathematical simulation of flow field for molten steel in an 80-ton single snorkel vacuum refining furnace. Steel Res Int. 2012;83:55–82.

Rui

Jiang

et al. Effect of elliptical snorkel on the decarburization rate in single snorkel refining furnace. Steel Res Int. 2012;84:192–197.

Rui

Cheng

GG.

A desulphurization mathematical model for single snorkel refining furnace. Adv Mater Res. 2012;476–478:340–345.

You

Cheng

Wang

et al. Mathematical model for decarburization of ultra-low carbon steel in single snorkel reﬁning furnace. Metall Trans B. 2015;46:459–472.

Dai

Cheng

et al. Numerical simulation of multiphase flow and mixing behavior in an industrial single snorkel refining furnace (SSRF): the effect of gas injection position and snorkel diameter. ISIJ Int. 2019;59:1214–1223.

Chang

Guo

et al. Industrial investigation of decarburization and desulphurization behaviour of 120 t new single snorkel degasser. Ironmak Steelmak. 2019. doi:10.1080/03019233.2019.1580029.

10.

Sugano

Shinkai

Kato

et al. Steelmaking of stainless steel by new type vacuum degasser (Development of new type vacuum degasser (REDA)−4). CAMP-ISIJ. 1999;12:747.

11.

Kitamura

Aoki

et al. Development of a novel degassing process consisting with single large immersion snorkel and a bottom bubbling ladle. ISIJ Int. 2000;40:455–459.

12.

Duan

Zhang

Yang

et al. Experiment research on decarburization technology of 80t single snorkel vacuum refining equipment. Iron Steel. 2011;46:21–25.

13.

Duan

Zhang

Yang

et al. Desulphurization behavior of an 80t multifunctional single snorkel vacuum refining furnace. Steelmaking. 2011;27:44–48.

14.

Misra

Sohn

Hydrogen and nitrogen control in steelmaking at U.S. steel. Iron Steel Technol. 2009;6:43–52.

15.

Ootsuka

Yamamoto

Nakagawa

et al. The successive hydrogen concentration control in molten steel by direct hydrogen measuring system. ISIJ Int 1996;36:S97–S100.

16.

Huang

Cheng

Wang

et al. Mathematical model for decarburization of ultra-low carbon steel during RH treatment. Ironmak Steelmak. 2020;47:655–664.

17.

Louhenkilpi

Numerical simulation of dehydrogenation of liquid steel in the vacuum tank degasser. Metall Trans B. 2013;44:459–468.

18.

Zhu

BH.

PhD dissertation. Chongqing, People’s Republic of China: Chongqing University; 2017.

19.

Winkler

Bakish

Vacuum metallurgy. Shanghai: Shanghai Science and Technology Press; 1982; 27.

20.

HM.

Refining technology of molten steel in electric furnace. Beijing: Metallurgical Industry Press; 2010; 25.

21.

Weinstein

Elliott

JF.

Solubility of hydrogen in liquid iron alloys. Trans Met Soc AIME. 1963;227:382–393.

22.

Higbie

The rate of absorption of a pure gas into a still liquid during short periods of exposure. Trans AIChE. 1935;31:365–389.

23.

Depuydt

Parlee

NAD.

The diffusion of hydrogen in liquid iron alloys metal. Metall Trans. 1972;3:525–532.

24.

Levich

VG.

Physicochemical hydrodynamics. Englewood Cliffs: Prentice-Hall Inc.; 1962.

25.

Szekely

Martins

GP.

Study in vacuum degassing and fluid mechanics of bubble growth at reduced pressures. Trans Am Soc Met. 1969;345:629–629.

26.

Park

KW.

A new numerical model for predicting carbon concentration during RH degassing treatment. ISIJ Int. 2003;43:1403–1409.

27.

Komarov

Sano

Bubble behavior and heat transfer in preheated gas injection into liquid bath. ISIJ Int. 1998;38:1045–1054.

28.

Fruehan

Martonik

LJ.

The rate of absorption of hydrogen into iron and of nitrogen into Fe-Cr and Fe-Ni-Cr alloys containing sulfur metall. Metall Trans B. 1981;12:379–384.

29.

Takahashi

Sano

Mori

et al. Enhanced nitrogen desorption from molten iron by combined Ar injection and blowing. ISIJ Int. 1987;27:633–637.

30.

Kitamura

Miyamoto

Tsujino

The evaluation of gas-liquid reaction rate at bath surface by the gas adsorption and desorption tests. Tetsu-to-Hagané 1994;80:101–106.

31.

Murthy GG

Ghosh

Mehrotra

SP.

Characterization of two-phase axisymmetric plume in a gas stirred liquid bath a water model study. Metall Trans B. 1998;19:885–892.

32.

Bannenberg

Bergmann

Gaye

Combined decrease of sulphur, nitrogen, hydrogen and total oxygen in only one secondary steelmaking operation. Steel Res Int. 1992;63:431–437.