Modeling of Vortex Flows in Direct Current (DC) Electric Arc

Furnace with Different Bottom Electrode Positions

OLEG KAZAK

This article is devoted to the numerical modeling of electrovortex and convection flows in the
direct current (DC) electric arc furnace with a different position of the bottom electrode. The
electromagnetic, temperature, and hydrodynamic distribution parameters are given. The shear
stress on the fettle area is offered as a criterion for the estimation of vortex flow influence on the
increased wearing of the fettle. It is shown that lifting the bottom electrode above the fettle
surface at the electrode radius leads to the decrease of shear stress on the fettle area by 30 pct.
Putting the bottom electrode lower than the fettle surface by the distance equal to the electrode
radius and its expanding by the same distance reduces the stress by 10 pct.

DOI: 10.1007/s11663-013-9899-4
Ó The Minerals, Metals & Materials Society and ASM International 2013

I. INTRODUCTION transfer is presented. In the modern works, a three-

dimensional model of magnetohydrodynamics processes
THE problem of electrovortex flow control in direct in DC EAF of cylindrical shape is developed.[2] At the
current (DC) electric arc furnaces (EAFs) with the current stage of studying, the processes in liquid metal
bottom electrode is a current important issue in modern (both task-specific software packages and general pur-
metallurgy.[1–5] As compared with alternating current pose ones with special preferences and selection method
(AC) EAFs, these furnaces consume 10 to 15 pct less of solution) are used.[1,2] In several articles, the arc
electric energy, 20 to 30 pct less refractory materials, 1.5 column is studied as well as the effect of arc column on
to 2.0 pct less feedstock, and 20 to 60 pct less alloying liquid metal movement and heat transfer from plasma
addition. Environmental safety is an essential advantage region throughout all the volume of liquid metal.[7,8]
of this furnace type, as the dust production diminishes 8 Some works deal with the possibility of the molten metal
to 10 times (from 52.5 to 6.3 kg/h), and the noise level movement operation. For example, some scientists
goes down from 105 to 85 dB when this furnace type is propose to operate the molten metal movement using
operated. The industrial practice has shown that DC bath lancing.[7,8]
EAFs have higher efficiency, low heat loss, lower However, these problems have been understudied so
components wear, and higher quality of steel pro- far and are important because solving them will make it
duced.[3] This furnace type allows using the steel quality possible to increase the furnace durability, to cut down
control, as well as analyzing and improving the compo- the product unit cost, and to reduce the energy
sition with the help of alloying addition in the metal consumption and environment pollution.
smelting process. The current article describes the physical processes
The exploitation of these furnaces has shown a high and mathematical statement of vortex flow in DC EAFs.
rate of fettle wear near the bottom electrode.[5] The During the liquid period, the impact of different factors
reason for the increased wear is connected with vortex is estimated with the help of the similarity criteria. The
flows of liquid metal caused by spatial homogeneity of algorithm of numerical modeling of proceeding pro-
electromagnetic fields and temperature. However, mix- cesses by standard software packages is developed. The
ing liquid metal in the bath during smelting processes is numerical modeling for the working period of DC EAF
necessary for the homogenization of alloying compo- with different bottom electrode positions is carried out.
nents and temperature throughout the liquid metal As a result, the electromagnetic, heat, and hydrody-
volume. Therefore, the most important objective is to namic parameters are obtained that allowed to estimate
estimate the intensity of vortex flows as an effect of the influence of Lorentz force and convection on the
different factors. character of the liquid metal vortex flow depending on
There are many scientific works that focus on physical the bottom electrode position.
processes in DC EAFs.[1,2,6–8] In the earlier works, all
general magnetohydrodynamics processes in liquid
metal are described and a mathematical model of heat
II. PHYSICAL PROCESSES IN DC EAF WITH
BOTTOM ELECTRODE
OLEG KAZAK, Researcher and Assistant Professor, is with
Donetsk National University, Donetsk, Ukraine. Contact e-mail:
The operation period of DC EAF with the bottom
[email protected] electrode can be divided into the following stages:
Manuscript submitted February 11, 2013. melting of the burden, liquid period when steel is

METALLURGICAL AND MATERIALS TRANSACTIONS B

produced, and tapping. The time of liquid period ranges Gr ¼ 0:5<1 (with the low power of the arc) to
from 15 pct to 60 pct of all operation period depending Gr ¼ bDTgL u20  18:5>1(with the full power of the
on the steel type that is produced and on the quality of arc), which corresponds to the insignificant or essential
starting raw material.[1–4] It is essential that the pro- contribution of the convection to the general vortex
cesses in DC EAF during the liquid period should be flow.[9] According to the preliminary estimation during
estimated. the full arc power period, it is necessary to take
In this type of furnace, the vortex flow of liquid metal convection into account, but convection can be
is the result of spatial unevenness of the current with the neglected at the period of low arc power.
absence of outer magnetic field. The current in the liquid The relative power of Joule heating as compared with
creates a magnetic field of its own, which causes vortex another heat source (heat from the arc) is low
movement of the liquid. j2 L
Q ¼ rqcu0 0 DT  103 <<1. This means the arc heat is
Convection flows make their own contribution to the
vortex flow and appear under uneven distribution of the more intense than joule heating.[9] The Peclet heat
temperature throughout the liquid volume. It is shown number that defines the ratio of the free heat convection
in the Reference 9 that heat convection in electrovortex transfer to the molecule heat conduction equals
flow with axial symmetry appears when the radial Pe ¼ u0 L=v  105 <<1, which indicates the domination
gradient exists (@T=@r 6¼ 0). The direction of convection of molecule heat conduction over free heat convection.[9]
depends on an increase or decrease of temperature value The magnetic Reynolds number is a part of the
with the increase of the distance from the axis of magnetic induction equation. The magnetic Reynolds
symmetry. number is low in this problem (Rem ¼ l0 ru0
To build a mathematical model of the processes in L  0:4<1), meaning that the movement of liquid con-
DC EAF, the following assumptions can be made: ductor does not change the magnetic field and the
calculation can be carried out in noninduction approxi-
(a) The medium is considered nonmagnetic (the Curie mation.[9]
point for steel is 1033 K [760 °C]). The processes in the DC EAF during metal smelting
(b) The medium is a good conductor and its permit- are not steady. However, they are rather slow and can
tivity can be neglected. be described in quasi-steady or just steady formulation.
(c) The processes are steady. For steady statement, the molten metal movement in
(d) Convective current, caused by the medium move- the furnace can be described by the system of equa-
ments compared to the current of conductance, can tions for magnetic, heat transfer, and hydrodynamic
be neglected. processes.
(e) Physical characteristics of the medium (conduc- The electromagnetic processes in liquid metal can be
tance, viscosity and heat-conduction indices, etc.) described by Maxwell’s equations
are assumed to be homogeneous and isotropic, and
! ~¼ 0
they depend on temperature. r  B ¼ l0~
j; r  B ½1
(f) Medium heating caused by viscosity (viscous dissi-
pation of energy) can be ignored as compared to the
Joule heating. rE ~ ¼ qe
~ ¼ 0; r  E ½2
e0
(g) Chemical reactions are not taken into account.
(h) The arc column is in local thermodynamic equilib-
rium, fixed temperature and the interface between Ohm’s law for fluid in motion
the plasma arc and the molten pool is fixed as a flat  
surface. The last one is made for simplicity of the ~
j¼r E ~þ ~
uB ~ ½3
model because in a real situation, the plasma jet
impinges on the surface and leads to boiling of the and charge conservation law
melt at the location of arc root attachment.[1]
r ~
j¼0 ½4
During the liquid period, the temperature difference
throughout the metal volume can range depending on where ~j is current density, qe is charge density, B ~ is the
the mode of furnace operation. Thus, when the arc magnetic induction intensity vector, E ~ is electrical field
works at full power, the temperature ranges from intensity, r is specific conductance, l0 is permeability of
3773 K (3500 °C) in the arc area at the cathode to free space, e0 is permittivity of free space, and ~u is liquid
1923 K (1650 °C) in the bottom electrode area and velocity.
along the fettle surface. At the low power of the arc, the The heat parameters are calculated by heat transfer
temperature difference throughout the metal volume equation
does not exceed 50 K (50 °C). It should be noted that 
the metal at this period is liquid. qCp u  rT ¼ r  ðða þ aT ÞrTÞ þ j2 r ½5
The velocity of the liquid that appears under
electromagnetic
pffiffiffiffiffiffiffiffiffiffi force can be estimated as where q is density, Cp is specific heat, T is temperature, a
u0 ¼ j0 L l0 =q  0:3 m/s.[9] The Grashof number that is heat conduction coefficient,  aT is turbulent heat
defines the ratio of relative intensity of convection conduction coefficient, and j2 r is the Joule heat source.
depending on the temperature range and electrovortex The hydrodynamic processes in the liquid can be
flow in the furnace is ranged in different periods from described by the Navier-Stokes equation

METALLURGICAL AND MATERIALS TRANSACTIONS B

  
u  r~
q~ u ¼ r  ðpI þ ðg þ gT Þ r~ uÞT
u þ ðr~
 ð2=3Þðr  ~ g þ~
uÞIÞ þ q~ ~
jB ½6
and equation of continuity
r  ðq~
uÞ ¼ 0 ½7
where p is pressure, ~ g is gravitation, t is dynamic-
viscosity coefficient, g ¼ t=q is coefficient of kinematics
viscosity, and ~ I is identity operator for points on the
boundary. The following forces are considered in
Eq. [6)] r  p is pressure force, r  ðg þ gT Þ
ðr~ u þ ðr~uÞT Þ is the force of viscous friction, and
~ ~ Fig. 1—The arrangement of cylindrical DC EAF (1—fettle, 2—liquid
j  B is the Lorentz electromagnetic force. metal, 3—electrodes, and 4—slag).
According to the preliminary estimation, the Rey-
nolds number under the movement in DC EAF
is Re ¼ u0 L=m  106 , which is equivalent to the devel- The no-slip boundary condition was used for hydro-
oped turbulent flow that can be described within the dynamic processes at all boundaries of liquid, both on
k  e turbulence model. The equations of k  e turbu- the boundary of the liquid with the fettle and the
lence model are boundary of liquid with slag. The last approximation is
    based on the fact that slag viscosity is much higher than
g 2qk liquid viscosity, and it can be considered as no-slip
q~u  rk ¼ r  g þ T rk þ gT Pð~ uÞ  r ~
u  qe condition. Mathematically, it means that velocity and
rk 3
turbulent components were all set to zero.
½8
u ¼ 0; v ¼ 0; w ¼ 0; k ¼ 0; e ¼ 0
   
g Ce1 e
u  re ¼ r  g þ T re þ
q~ ½gT Pð~
uÞ For the simulation of flow in the wall region, the
re k
  universal logarithmic law was applied
2qk Ce2 qe2
 r~u  ½9 1=2
3 k qup kconst C1=4
l 1
2 ¼ lnðEyþ Þ
r~
u 2ðr~uÞ qC k2 sw kconst
where Pð~ uÞ ¼ r~uþr~uT  3 ; gT ¼ el ; Ce1 ; Ce2 ; Cl
1=2 1=4
are constants of turbulence model, and k; e is the y k
where yþ ¼ p const lC
; Econst ¼ 9:79, is the empirical con-
t
transported variable of turbulence model. stant, kconst ¼ 0:42 is Von Karman’s constant, sw is
shear stress on the wall, yp is the resultant velocity of the
fluid near the wall, yþ is the dimensionless normal
distance from the resultant velocity, and yp is the
III. THE MODEL OF INDUSTRIAL FURNACE distance of the first node point p from the wall.[10]
The bottom electrode in this furnace will be solid with
To build a model of the processes in liquid metal, the
constant temperature during all operation times accord-
parameters of the industrial DC EAF with the bottom
ing to bottom electrode construction.[11]
electrode are taken into account.[4] The geometrical
For all types of analysis on the axis of symmetry of the
arrangement of the furnace has been shown in Figure 1.
calculation domain, the following conditions were used:
The main parts of the configuration in Figure 1 are
1—fettle, 2—liquid metal, 3—top and bottom elec- @u @w @k @e
trodes, and 4—slag. The axial symmetry allows calcu- ¼ 0; ¼ 0; ¼ 0; ¼0
@y @y @y @y
lating the half of the cross-section area. Its main
parameters are as follows: Furnace capacity is 100 t,
direct current load is 80 to 100 kA, the mainlines voltage While modeling the following physical parameters for
is 500 to 1000 V, power of current consumption is 40 to liquid steel, electrodes and fettle were used: Electric
100 MW, and polarity is «+» on bottom electrode. The conductivity of liquid steel does not change significantly
formulated problem was solved with the corresponding with the temperature increase, it changes only by 9 pct
boundary conditions that are defined in Figure 1 as B1 from 1823 K to 3003 K (1550 °C to 2730 °C). It was
to B9 and presented in Table I. taken as r1 = 7.12 106 (OhmÆm)1 at 2023 K (1750 °C);
The current density on the boundary with the normal electric conductivity of electrodes r2 = 0.2 106
cross-section of electrode jn =j0 =I=S, where S is cross- (OhmÆm)1; relative permeability and relative conduc-
section of electrode. For the electric and magnetic fields, tivity of the media, electrodes, and liquid steel l = 1 and
the conditions of continuity are used. The heat bound- e =1; specific heat for 1773 K to 2273 K (1500 °C to
ary conditions are represented by constant temperature 2000 °C) is Cp = 750 J/kgÆK; and thermal conductivity
of the electric arc, the bottom electrode, the boundary for 1773 K to 2773 K (1500 °C to 2500 °C) is a =
with slag, and fettle.[1,2,4] 32 W/MÆK.[12] The density and viscosity were used as set

METALLURGICAL AND MATERIALS TRANSACTIONS B

 Table I. Correspondence of Boundaries to Boundary Conditions

Boundary/Conditions B1 B2 B3 B4 B5 B6 B7 B8 B9
Electric j0 jn ¼ 0 j0 jn ¼ 0 j0
Magnetic Es1 ¼ Es2 ; Dn1 ¼ Dn2 ; Bn1 ¼ Bn2 ; Bs1 ¼ Bs2
Heat – – – – – T2 = 1900 K T3 = 1900 K T1 = 3300 K
(1627 °C) (1627 °C) (3027 °C)
Hydrodynamic – – – – – u¼0
~

functions of temperature that are given as tabulated

values.[12] The k  e turbulence model constants
Cl ¼ 0:09; Ce1 ¼ 1:44; Ck2 ¼ 1:92; rk ¼ 1:0 and re ¼
1:3 are chosen.
The computer modeling of vortex flows is considered
to be a multiphysics problem, and the strategy of
solution consists of the following steps: first step is
solving electromagnetic fields, the second step is solving
heat processes, and the third step is solving electro
vortex flows taking into account heat exchange and
convection. For its software realization in ANSYS
Multiphysics, ANSYS CFX (ANSYS Inc., Canonsburg,
PA) and COMSOL (COMSOL Inc., Burlington, MA)
are chosen.
The preliminary analysis provided the optimal divi-
Fig. 2—The vector and contour field of Lorentz force near the bot-
sion into elements, as well as their form.[13–15] The tom electrode (1—fettle, 2—liquid metal, 3—electrodes, and
domain area was split into elements unevenly: In the 4—slag).
area of the electrodes and boundary with the fettle,
where the large gradients of all parameters are situated,
the elements were densely located and were of small size, can be seen Figure 3, the maximum temperature value is
approximately 0.01 of the electrode radius. At the other located right near the cathode (top electrode) at the
parts of the calculation domain, the elements were not distance about the electrode radius. In temperature
densely located and were of large size (approximately arrangement, the radial gradient produces the convec-
10 times larger). The methods of building the calcula- tive flow in electrovortex movement as is shown in
tion mesh that take into account the specialty of the Reference 16.
processes were taken from Reference 13, in which the At the next stage, according to the solution strategy,
calculations were made for plane axial symmetry (two- the hydrodynamic processes in liquid metal were mod-
dimensional [2-D]) as well as for three dimensions (three eled taking into account the electrovortex and convec-
dimensional [3-D]). At well-corresponding meshes, the tion flows in axial symmetry formulation.[16] In
differences of the results obtained at 2-D and 3-D Figure 4, the hydrodynamic fields of velocity vector,
formulation ranged between 1 and 3 pct. But the time of contour, and streamlines are given.
the calculation for 3-D formulation increased sev- The figure demonstrates that the higher intensity of
eral times, as compared with 2-D formulation. The vortex movement appears in liquid metal volume. The
good coincidence of obtained results makes it possible to convection flows are in the line with electrovortex flows
carry out calculations in 2-D axial symmetry formula- and electrovortex movement increases. The vortex arises
tion. near the bottom electrode, as Figure 4 shows. The
vortex flow of liquid metal on axis of symmetry is
directed upward. When the flow achieves the upper
IV. RESULTS AND DISCUSSION boundary of liquid metal, it comes along the boundary
and comes down. The maximum value of the vortex flow
Some results of solving the processes proceeding in velocity was located on the axis of symmetry and
liquid steel are given below. Figure 2 demonstrates the reaches 0.5 m/s. The vortex flow velocity value in close
vector and contour fields of the Lorentz force near the proximity to the bottom electrode and the fettle com-
bottom electrode (anode), where 1 is fettle, 2 is liquid prises about 0.3 m/s. According to the figure, at the top
metal, and 3 is bottom electrode. The value of Lorentz electrode area the inverse vortex flow appears. This
force ranged and comprised about 30 pct of volumetric vortex is produced by the uneven distribution of current
gravity force. The results of the calculations prove that density near the top electrode.
the Lorentz force in such furnaces is essential for the According to the general theoretical conceptions and
appearance of electrovortex flow.[13,14] the recent researches in metallurgy,[10] there is a direct
The heat processes in the liquid metal were modeled in connection between shear stress on the fettle area from
axial symmetry formulation. In Figure 3, the tempera- velocity of the liquid and the increased fettle wear. The
ture field throughout liquid metal is demonstrated. As shear stress can be used as a criterion to estimate the

METALLURGICAL AND MATERIALS TRANSACTIONS B

Fig. 3—Contour field and streamlines of the temperature (1—fettle,
2—liquid metal, 3—electrodes, and 4—slag).

Fig. 5—The vector, contour field, and streamlines of vortex flow

velocity with bottom electrode lowered below the fettle by the dis-
tance equal to two radiuses of the electrode (1—fettle, 2—liquid
metal, 3—electrodes, and 4—slag).

Fig. 4—The vector, contour field, and streamlines of vortex flow

velocity with bottom electrode temperature T = 1900 K (1627 °C)
(1—fettle, 2—liquid metal, 3—electrodes, and 4—slag).

negative influence of the liquid metal movement on the

bottom electrode and the fettle near it.
The change of the bottom electrode position is one of
the possible ways to reduce the negative influence of the
liquid metal vortex flow on the increased fettle wearing. Fig. 6—The vector, contour field, and streamlines of vortex flow
To estimate this effect, some simulation of hydrody- velocity with the bottom electrode lowered below the fettle and
namic processes at different positions of bottom elec- expanded in diameter by a half of the electrode radius (1—fettle,
2—liquid metal, 3—electrodes, and 4—slag).
trode has been done.
At the first step, the numerical experiment for the
bottom electrode position dropped below the fettle by electrode, some vortex flows appear in the space above
the distance equal to two radiuses of the electrode was it. They are localized on the surface of the fettle around
conducted. In Figure 5, the vector, contour field, and the axis of symmetry. The vortex at the upper electrode
streamlines of vortex flow velocity for this bottom was extended. Also, the intensity of the movement is
electrode position are shown. As can be seen, intense reduced significantly. Thus, at the level of the sublayer,
vortex motion occurs in the melt. The maximum velocity the shear stress decreases by 1.5 times, and the shear
of the vortex motion is observed on the electrode axis stress on the surface of the fettle is decreased by 10 pct.
and reaches 0.57 m/s. In comparison with the standard Figure 7 shows the results of raising the bottom
provision of the bottom electrode, the increase of electrode above the level of the fettle to the height equal
maximum velocity of the melt is observed. Also, the to the radius of the electrode. The calculations show that
velocity increases near the bottom electrode and fettle the velocity of the melt in this case is reduced almost
around it. The value of the shear stress at the fettle near two times. Some significant change in the structure of
the bottom electrode does not change significantly. It the melt flow can be observed: A vortex with small
should be noted that in the space above the lowered velocity is formed over the bottom electrode. This
bottom electrode structure of the flow is very complex, bottom electrode position leads to the decrease of the
and it forms several vortices of small size and intensity. shear stress on the fettle area by 30 pct.
At the next step, the numerical experiment for the Figures 8 and 9 compare the graphs of the velocity
bottom electrode put lower than the fettle and expanded distribution at the level of the shear sublayer and the
in its diameter to a half of the radius of the electrode was values of the shear stress as a function of the distance
conducted. The results of the numerical simulation for from the axis of symmetry. Graphs show the magnitude
this case are shown in Figure 6. At this position of the of the shear stress in a nondimensional form. The scale

METALLURGICAL AND MATERIALS TRANSACTIONS B

Fig. 7—The vector, contour field and streamlines of vortex flow
velocity with the bottom electrode raised above the level of the fettle
to the height equal to the electrode radius (1—fettle, 2—liquid metal,
3—electrodes, and 4—slag).
Fig. 8—The comparison of the melt velocity at a shear sublayer at
of the shear stress is taken as a characteristic value at the different bottom electrode positions depending on the distance from
the axis of symmetry.
standard bottom electrode position (s0 ¼ 120Pa)
depending on the distance expressed as the electrode
radiuses (R0 ¼ 0:25m).
It is shown that lifting the bottom electrode above the
surface of the fettle by the electrode radius value leads to
the decrease of the shear stress on the fettle area by
30 pct, whereas putting the bottom electrode lower than
the fettle surface by the electrode radius value and
expending the bottom electrode to the electrode radius
value reduces the stress by 10 pct.

V. VERIFICATION OF THE OBTAINED

RESULTS
Reliability and accuracy of the calculated results were
estimated by comparing these results with the following:
– Theoretical researchers
– Experimental data obtained with the laboratory
installation Fig. 9—The comparison of the shear stress on fettle surface at differ-
– Calculations done by other methods and software ent bottom electrode positions depending on the distance from the
packages axis of symmetry (s0 ¼ 120Pa, R0 ¼ 0:25 m).
– Experimental data based on the increased wear of the
bottom electrode for industrial DC EAF ·10 -2 m/s
m/s
The comparison of the obtained results with theoret- 15
ical researches of liquid metal movement in model tasks
is carried out. The character of the liquid movement
obtained in numerical modeling conforms to the general 12
theoretical assumptions that are given in the physical
statement of the problem and described in References 1, 9
2, 7–9, and 17. According to the theory, when the
current flows through the molten metal, a magnetic field
is generated. The current conductor in the magnetic field 6
is affected by the Lorenz force. Spatial unevenness of the
current leads to the vortex character of the Lorenz force,
according to the geometry of the furnace and the 3 Experimental
electrode location. Under the influence of this force, the
Numerical
elements of liquid move as a whole toward the resultant
force vector, and hence, vortex flow appears. The 0 2 4 6 8
·10 -2 m
maximum intensity of the vortex movement is found
near the bottom electrodes where the value of the Fig. 10—Numerical and experimental results for the laboratory unit
rotational Lorenz force module is maximum. (1, 2—electrodes, and 3—liquid eutectic alloy).

METALLURGICAL AND MATERIALS TRANSACTIONS B

 New After 400 heats

Fettle
Connection
Shear stress between fettle and
electrode
About 400 mm lost
electrode length

Fig. 11—Configuration of the increased wear of bottom electrode and fettle near the bottom electrode.

The turbulent electrovortex movement of liquid metal All of that has proved the reliability of the methods and
in hemispherical volume is calculated based on the approaches, as well as the accuracy of the obtained
developed methods. This movement was studied exper- results.
imentally in the laboratory installation.[17–19] The
obtained results that correlate with the experimental
data within 15 pct limits were received for the k  e VI. CONCLUSIONS
turbulence model, which was used in the further
calculations. The comparisons of the experimental and The physical and mathematical model of processes
numerical results are given in Figure 10. proceeding in DC EAF have been built. To describe the
To verify the obtained results, similar calculations processes in the DC EAF, the model of the magnetic
were done in COMSOL. The comparison of the hydrodynamics is adopted. It takes into account the
obtained results by different software packages and spatial distribution of the current, electric and magnetic
methods showed an insignificant difference and the fields, temperature, Lorenz force, Joule heat, and
value of it is approximately 3 pct.[13–16,18,19] The simi- convection. The strategy of solving the stated conjugate
larity of the calculations done by different methods and problem in standard software packages is worked out.
software packages proves the reliability of the methods Numerical modeling of proceeding processes in liquid
and significance of the results. metal for DC EAF is carried out.
According to the general theoretical conceptions and It is established that the volumetric Lorentz force
the recent research in metallurgy,[10] there is a direct makes up about 30 pct of the volumetric gravity force
connection between the liquid metal movement and the and is essential for the vortex flow appearance, the
increased fettle wear. The increased wear configuration maximum value of which reaches 0.5 m/s. The results
of the bottom electrode and the fettle near the bottom showed the main contribution of electrovortex force and
electrode after 400 heats is given in Figure 8.[5,15] These significant contribution of convection flows with the
data have been obtained since 1996 by checking the maximum heat power of furnace.
actual wear periodically throughout the life of several The results of the calculations in ANSYS Multiphys-
DC EAF: Balboa (Spain) 70 t, Hylsa Planos (Mexico) ics and ANSYS CFX are compared with the experi-
135 t, and ABC Danarc Plus (Italy) 90 t. The increased mental data, calculations in COMSOL, general
wear of the bottom electrode and fettle right near the theoretical conceptions. The similarity of the calcula-
bottom electrode confirms that the flows with higher tions done by different methods and experimental data
velocity and maximum value of the shear stress are proves the reliability of the methods and significance of
located near the bottom electrode at a distance of about the results.
the radius of the electrode. The conic shape of the fettle This article describes the methods for vortex flow
border near the bottom electrode shows that the liquid control in the DC EAF with the bottom electrode based
vortex flow is of conic character. The washout of fettle on the changes of the bottom electrode position. This
right near the bottom electrode indicates that stream- method was proven to increase the stability of the fettle
lines of liquid movement under the Lorenz force are wearing near the bottom electrode. It is shown that
produced and finished on the bottom electrode. This lifting the bottom electrode above the surface by the
characteristic of the fettle wearing confirms that the electrode radius value leads to the decrease of shear
major cause of the increased wear is the liquid vortex stress on the fettle area by 30 pct, while bottom
movement under the Lorenz force. electrode lower than the fettle surface by the electrode
The verification of the obtained results has shown a radius value and expanding it by the same value reduce
good correlation with the general theoretical data the stress by 10 pct. The technique of vortex flow
concerning electrovortex movement, the results control in the DC EAF with the bottom electrode is
obtained by other authors, the experimental data for developed, allowing to reduce the fettle wearing and to
both laboratory installations and industrial DC EAF. optimize the furnace work.

METALLURGICAL AND MATERIALS TRANSACTIONS B

 REFERENCES 10. A. Kumar, S.K. Ajmani, and A.R. Pal: Ironmaking Steelmaking,
2010, vol. 37 (1), pp. 9–14.
1. F. Wang, Z. Jin, and Z. Zhu: Ironmaking Steelmaking, 2006, 11. G. Gensini and M. Pavlicevic: U.S. Patent 5651024, 1997.
vol. 33 (1), pp. 39–44. 12. I.K. Kikoin: Table of Physical Value, Information Book, Atomiz-
2. B. Henning, M. Shapiro, and L.A. le Grange: Proceedings: Tenth dat, Moscow, 1976, pp. 1–1008 (in Russian).
International Ferroalloys Congress; INFACON X: ‘‘Transformation 13. O. Kazak and O. Semko: J. Eng. Phys. Thermophys., 2011, vol. 84
through Technology,’’ Cape Town, South Africa, 2004, pp. 565–74. (1), pp. 209–18.
3. S.M. Nekhamin, A.G. Lunin, M.M. Krutyanskii, and A.K. 14. O. Kazak and O. Semko: Ironmaking Steelmaking, 2011, vol. 38
Filippov: Refract. Indust. Ceram., 2005, vol. 46 (1), pp. 37–39. (4), pp. 273–78.
4. A. Michielan, G. Lavaroni, and A. Fior: ABS Rev. Met., France, 15. O. Kazak and O. Semko: Ironmaking Steelmaking, 2011, vol. 38
2000, no. 6, pp. 745–52. (5), pp. 353–58.
5. V.A. Zaycev and L.B. Medovar: SEM, 2009, no. 2, pp. 3–8 (in 16. O. Kazak and O. Semko: Ironmaking Steelmaking, 2012,
Russian). in press.
6. J. Szekely, J. McKelliget, and M. Choudhary: Ironmaking Steel- 17. Yu. Ivochkin, A. Oksman, O. Kazak, I. Teplyakov, and V. Zhilin:
making, 1983, vol. 10 (4), pp. 169–79. 8th International PAMIR Conference on Fundamental and Applied
7. J. Alexis, M. Ramirez, G. Trapaga, and P. Jonsson: ISIJ Int., MHD, Borgo, Corsica, France, 2011, pp. 85–88.
2000, vol. 40 (11), pp. 1089–97. 18. O. Kazak and O. Semko: J. Eng. Phys. Thermophys., 2012, vol. 85
8. M. Ramirez, G. Trapaga, J. Alexis, and P. Jonsson: Electric (5), pp. 1073–83.
Furnace Conference Proceedings, 1999, pp. 751–61. 19. O. Kazak and O. Semko: 8th International PAMIR Conference
9. V. Bojarevics, Ya. Freibergs, E.I. Shilova, and E.V. Shcerbinin: on Fundamental and Applied MHD, Borgo, Corsica, France, 2011,
Electrically Induced Vortical Flows, Kluwer Academic Publishers, pp. 605–09.
Amsterdam, 1989, pp. 1–315.

METALLURGICAL AND MATERIALS TRANSACTIONS B