Professional Documents
Culture Documents
814822
A multiphase numerical analysis focused on flow dynamics and particle trajectories during steel tapping
operations was developed. The numerical results indicate that lighter additions than steel (ferrosilicon and
aluminum) are independent from bath level, fall height and flow dynamics of the melt. Neutral buoyant
additions (FeMn) are strongly dependent on fluid dynamics of the melt and bath height. Denser additions
(like FeNb) yields long residence time inside the melt before first emerging to the bath surface. However,
when this ferroalloy is added at high bath levels, close to the end of tapping, the particles remain in the
corner formed by the bottom and the wall of the ladle during long times prolonging their melting rates.
KEY WORDS: tapping steel; air bubbles; additions; ferroalloys.
ing into the melt. Final oxygen levels, assuming thermodynamic equilibrium during steel tapping, depend then on
efficient mixing and melting-dissolution processes of ferroalloys and aluminum. Due to these reasons it is important
to know the trajectories and residence times of particles of
ferroalloys in molten steel. The problem this paper is dealing with had been already analyzed by Guthrie et al.6) who
employed a balance of forces on a particle and evaluated the
importance of drag, buoyancy, added mass and history forces
acting on a particle submerged into a liquid. Their physical
and mathematical models included experiments of wooden
particles with different densities into a tank of still water.
After the mathematical analysis of their experiments these
authors concluded that the history term in the balance of
forces has negligible influence on the particle dynamics
emphasizing the importance of drag, buoyancy and mass
added forces. Maximum depth penetration of particles, for
a given initial entry velocity, depend on the density ratio
between the particle of ferroalloy and steel, higher ratios
mean deeper penetrations. Tanaka et al.7) performed also
physical and mathematical modeling for ferroalloys additions in a 250 ton steel ladle. They established, through
dimensional analysis, modeling criteria for addition sizes
and entry velocities between a model and the actual ladle
linked by the square root of the scale factor of the model.
These authors simulated the effects of steel motion on
spherical particle trajectories assuming a one-way coupling
mechanism between liquid and particles (liquid steel flow
1. Introduction
Tapping is probably the most important operation leading
to clean steel production since it is during this time that
deoxidizers and alloying elements, in form of ferroalloys or
metallic, are added and slag carryover must be avoided to
simplify later ladle furnace operations. Naturally, initial
oxygen content in steel governs the efficiency of those additions but certainly air entrainment by the plunging jet and
the bath surface turbulence contribute to form a multiphase
flow made of liquid steel, air and solid particles of deoxidizer. Under these circumstances excessive air entrainment
works as a cushion dampening the steel motion and hindering the mixing and the melting-dissolution processes of ferroalloys. Ferroalloy additions to molten steel initially freeze
a shell of steel around the particles and this shell melts back
after the ferroalloy or metal13) (like Al and Ni) addition has
melted within this shell. Hereof, the residence time of ferroalloys particles inside the melt during steel tapping is
important to have high alloying and deoxidizing efficiencies. Assuming thermodynamic equilibrium, complete mixing conditions and efficient ferroalloys dissolution, the
group of the authors demonstrated that the amounts and
types of inclusion chemistries depend on the addition
sequence during steel tapping, steel level in the ladle and
oxygen concentration in the melt.4,5) Therefore, those findings underline the importance of an efficient mixing process
assisted by the momentum transfer effects of the jet plung 2012 ISIJ
814
1
q
t q q + ( q q vq ) = S q + p =1 m pq m qp
(1)
q =1 q = 1 ................................ (2)
n
= q =1 q q ............................. (3)
n
= q =1 q q ............................. (4)
n
2. Mathematical Model
(ui u j )
( ui )
p
+
=
+
t
x j
xi x j
u
eff i
xj
+ g ... (5)
2012 ISIJ
du pi
dt
p
18 C D Re p
u j u pi +
gi
2
p d p 24
1 d
ui
ui
ui u pi +
p xi
2 p dt
ui = ui + ui ................................. (9)
u = ui 2 = 2 k / 3 ..................... (10)
where ui is mean fluid phase velocity (m/s); ui is random
velocity fluctuation (m/s); random number; is a normally
distributed random number; k is local level of turbulent
kinetic energy (m2/s2). Therefore, the flow field is calculated
through Eqs. (1)(6) plus equations for the kinetic energy
and its dissipation rates together with the auxiliary equations
of the k- model. Once that the flow field is calculated it is
used to calculate the trajectories of particles in the melt
through Eqs. (7)(10).
........ (7)
CD =
24
(1 + 0.186 Re0p.6529 ) .................. (8)
Re p
The first term in Eq. (7) is the drag force per unit particle
mass, the second term is the gravitational force, the third
term is the virtual mass force accelerating the fluid surrounding the particle,11) and the fourth term is the force
stemming from the pressure gradient in the fluid. To incorporate the stochastic effect of turbulent fluctuations on
the particle motion, the random walk model17,18) was used.
In this model, particle velocity fluctuations are based on a
Gaussian-distributed random number chosen according to
the local turbulence kinetic energy. The random number is
changed, thus producing a new instantaneous velocity fluctuation, at a frequency equal to the characteristic lifetime
a)
Fig. 1.
b)
Density (Kg/m )
Apparent density
(dimensionless)
Fe Mn (50 mass%)
6 200
Fe Nb (65 mass%)
Fe Si (75 mass%)
Al
2012 ISIJ
Table 2.
0,950
Steel mass
in ladle
(ton)
Center
(m)
Medium
(m)
Side
(m)
Bath level
(m)
Velocity in y
direction (m/s)
7 980
1,140
30
0,11
0,593
1,1
0,94
6,61
3 354
0,485
50
0,11
0,593
1,1
1,58
5,58
2 594
0,390
70
0,11
0,593
1,1
2,23
4,29
816
3. Results
3.1. Velocity Fields by Plunging Jet
Figures 3(a), 3(b), 3(c) and 3(d) show the velocity fields
during steel tapping for steel filling levels of 10, 30, 50 and
75 tons, respectively. The first figure shows essentially the
effects of the plunging steel jet on the surrounding air which
receives momentum to form long recirculating flows at each
side of the jet. Steel, in the ladle bottom, Fig. 3(a), observes
a horizontal flow without well defined recirculations. At 30
tons, Fig. 3(b), air forms four recirculating flows, two smaller ones in contact with the melt and other two, long ones,
in the upper side of the ladle. Steel forms recirculating flows
at both sides of the jet. Besides, along the jet there is the formation of a boundary layer of air with larger velocities at
the metal-gas interface. At 50 tons, Fig. 3(c), the flow pattern described by the precedent figure remains, however, the
two recirculating air flows in contact with the melt are now
smaller. The steel flow forms two longer vertical recirculating flows just as it does when the steel level is 30 tons. At
75 tons, Fig. 3(d), the flow patterns suffer radical changes;
the two recirculating flows of air remain. However, flow
pattern of steel now indicates vertical-downwards displacements toward the ladle bottom without forming recirculating
flows; rather three smaller recirculation flows are formed
along the jet length and one in the ladle corners as is indi-
a)
Fig. 2.
a)
c)
b)
Fig. 3.
817
b)
d)
2012 ISIJ
a)
Fig. 4.
b)
b)
a)
Fig. 5.
b)
Fig. 6.
2012 ISIJ
818
a)
b)
Fig. 8.
c)
Fig. 7.
a)
b)
c)
d)
d)
a)
Addition of ferroalloys with different size a) FeMn diameter of 1 cm MRT 5.6 s. b) FeMn diameter of 5 cm MRT
22.2 s. c) FeSi diameter of 1 cm MRT 0.5 s. d) FeSi
diameter of 5 cm MRT 0.8 s.
Fig. 9.
819
b)
c)
2012 ISIJ
Fig. 10.
a)
Fig. 11.
b)
c)
2012 ISIJ
Fig. 12.
820
plunging jet, yet, the high turbulence of that location induces relatively large standard deviations. Therefore, availability of buoyant or light additions can sometimes be optimum,
i.e., large MRT and some other times are not optimum, i.e.
short MRT.
Fig. 13.
Fig. 14.
3.4. Discussion
Addition of buoyant additions whenever possible should
be carried out close or just in the plunging jet to induce long
MRTs and large particle trajectories and this effect will be
favored with high steel levels in the ladle. Other falling locations will lead, generally, to very short MRTs and fast
emerging particles with minor effects of particle sizes and
resulting in low alloying efficiencies. Therefore, particle
dynamics of ferrosilicon and aluminum are very similar.
From a practical standpoint it can be said that MRTs of
buoyant additions to steel baths are independent from fluid
flow patterns in the ladle with the exception of the plunging
jet neighborhood. Even, if the melt in the ladle would be
stagnant buoyant additions will show the similar behaviors
to those described here whenever the addition are made outside the influence of the plunging jet. Naturally, with denser
additions fluid flow plays more important roles on alloys
availability as is discussed below.
For neutral floating particles such as ferromanganese the
best falling position is the mid point between the jet and the
ladle wall at a tonnage of 50 tons and this is due to the dragging effects on these particles exerted by the recirculating
flow of the melt. When the falling position is located close
to the ladle wall the MRT decreases considerable and the
standard deviation grows meaning that sometimes these particles will go deep into the melt and at some other times they
will emerge relatively fast all depending on the instantaneous fluid flow structure. Addition of ferromanganese
reaches the highest MRT at steel levels of 50 tons with a
falling position in the middle of the plunging jet and the
ladle wall. At a steel level of 30 tons the MRT of ferromanganese is longer in a falling position close to the jet and
improves when the steel level is 70 tons. The same trend is
observed when the falling position is close to the ladle wall.
MRTs of neutral particles are then dependent on the instantaneous fluid flow pattern and the best steel level for efficient alloying is at level of approximately 70% of the total
steel tonnage in the ladle.
Heavier additions than steel like FeNb, have the largest
MRTs and alloying efficiency is independent from steel
tonnage in the ladle. However, close to a level of complete
ladle filling heavy additions sink to stagnant melt regions
being subjected to low melting rates. The most recommendable steel level for these additions would be about 60% of
the total steel tonnage when the melt has been deoxidized
and melt streams do not carry them to stagnant regions.
Bringing on now the metallurgical aspects it can be said
that a recommendable and ideal tapping procedure for low
carbon aluminum killed steels, even for ultra low carbon
steels, would be the following:
Compact plunging jets by ensuring good surface conditions of the EBT nozzle and avoid slag carry-over.
Start the tapping with and addition of coke for predeoxidation purposes forming a steel cushion. Add,
simultaneously any synthetic slag if that is indicated by
2012 ISIJ
q
q
vq
EBT
CAS
VOF
eff
0
t
pi
4. Conclusions
A numerical-multiphase fluid flow model to study
dynamics of alloy additions during steel tapping operations
has been developed and from its results the following conclusions can be drawn:
(1) Flow patterns of steel during tapping operations suffer radical changes with steel level in the ladle. These
changes of flow patterns have strong effects on neutral and
heavy additions to steel.
(2) At low steel levels, beginning of tapping, entrainment of air induces small bubbles being dragged into the
melt homogeneously distributed providing large interfacial
melt-air areas leading, eventually, to higher nitrogen and
oxygen absorption. At high steel level, close to the end of
tapping, entrained air bubbles are larger, concentrated in the
jet and with smaller interfacial areas. Expecting, under these
circumstances, less oxygen and nitrogen absorption rates.
(3) Minimum residence times in the bath of buoyant
additions to steel bath are practically independent from tapping conditions since buoyancy forces are so large that they
will emerge fast from the bath. Nevertheless, the most recommendable position for addition is as close as possible to
the plunging jet.
(4) Neutral additions are very dependent on the instantaneous flow turbulence in the bath. Most appropriate conditions indicate a bath level of approximately 60% of the
maximum level at the middle position between the jet and
the ladle wall.
(5) Heavy additions require relatively low steel levels in
the ladle. However, at a steel level of approximately 60% of
the total bath level the minimum residence time becomes
independent from the fall position.
t
CD
Rep
dp
p
i
ui
g
h
REFERENCES
1) S. A. Argyropoulos and R. I. L. Guthrie: Metall. Mater. Trans. B,
15B (1984), 47.
2) S. A. Argyropoulos, A. C. Mikrovas and D. A. Doutre: Metall. Mater.
Trans. B, 32B (2001), 239.
3) S. A. Argyropoulos, D. Mazumdar, A. C. Mikrovas and D. A. Doutre:
Metall. Mater. Trans. B, 32B (2001), 247.
4) V. Carreo-Galindo, R. D. Morales and A. Serrano: Proc. 57th Electric
Arc Furnace Conf., Iron and Steel Society, Pittsburgh PA, (1999),
523.
5) V. Carreo-Galindo, R. D. Morales, A. Serrano and M. Toledo: Steel
Res., 71 (2001), 107.
6) R. I. L. Guthrie, R. Clift and H. Henein: Metall. Mater. Trans. B, 6B
(1975), 321.
7) M. Tanaka, D. Mazumdar and R. I. L. Guthrie: Metall. Mater. Trans.
B, 24B (1993), 639.
8) D. Mazumdar and R. I. L. Guthrie: Metall. Mater. Trans. B, 24B
(1993), 649.
9) K. J. Hammad: Proc. ASME 3rd Joint US-European Fluids Eng. And
8th Int. Conf. on Nanochannels, Microchannels and Minichannels,
ASME, New York, (2010), 1.
10) M. Iguchi, K. Okita and F. Yamamoto: Int. J. Multiphase Flows, 24
(1998), 523.
11) Fluent Documentation: Fluent Inc., Lebanon NH, USA, (2007).
12) J. Liow, L. Rudman and P. Liovic: ISIJ Int., 41 (2001), 225.
13) M. R. R. I. Shamshsi and S. K. Ajmani: ISIJ Int., 47 (2007), 433.
14) R. C. Morgans, B. B. Dally, G. J. Nathan, P. V. Lanspeary and F.
Fletcher: Proc. 2nd Int. Conf. on CFD in The Minerals and Process
Industries, CSIRO, Canberra, Australia, (1999), 479.
15) G. S. Daz, R. D. Morales and A. R. Banderas: Int. J. Heat and Mass
Transfer, 48 (2005), 3574.
16) O. Davila, R. D. Morales and L. G. Demedices: Metall. Mater. Trans.
B, 37B (2006), 71.
17) J. E. Olsen and S. Cloete: 7th Int. Conf. of CFD in Minerals and Process Industries CSIRO, Melbourne Australia, (2009), 1.
18) J. F. Domgin, P. Gardin and M. Brunet: 2nd Int. Conf. on CFD in
Minerals and Process Industries CSIRO, Canberra, Australia, (1999),
181.
Acknowledgments
The authors thank the institutions IPN and CoNaCyT for
the continuous support to Group of Metallurgical Process
Engineering at ESIQIE-IPN.
Nomenclature
m pq Mass transfer rate from phase q to phase p.
m qp Mass transfer from phase p to phase q.
S q Constant or user define mass source for each
phase.
2012 ISIJ
822