simulations of 1/4th scaled down tundish of Rourkela Steel Plant (RSP) with different flow modifiers. RSP ... molten steel flow characteristics, inclusion floatation.
FLOW OPTIMIZATION IN THE TUNDISH WITH THE DIFFERENT COMBINATION OF FLOW MODIFIERS
Sujata Devi, R. K. Singh, A. Paul, T. K. Pratihar &
s. K. Srivastava
ABSTRACT The efficiency and optimization of continuous casting operation require a close control of the molten steel flow characteristics within the tundish. Therefore, to improve the flow characteristics in the tundish, different flow modifiers such as dams, weirs, baffles, turbostop etc. are used. The present study consists of mathematical simulations of 1/4th scaled down tundish of Rourkela Steel Plant (RSP) with different flow modifiers. RSP tundish of 30T capacity is presently having striker pad and baffle as flow modifiers. Fluid flow has been simulated using k- turbulent model with species transport for residence time distribution (RTD) analysis and discrete phase model to investigate inclusion floatation. With the existing condition of the plant tundish, results in a recirculatory motion with intense mixing to the region below the inlet at the inlet region, due to presence of baffle, which leads more dead zone, higher turbulence on the free surface of the melt and higher shear stress on the walls of the tundish. Use of turbostop with existing condition of tundish hampers tundish function resulting in higher dead volume and less mean residence time. Similar types of results are found when turbostop is used with the baffle, whereas tundish with turbostop, gives better performance with respect to higher mean residence time, less dead zone, higher dispersed plug volume and mixed volume. The turbostop also provides more surface directed flow with lesser turbulence on the free surface of the melt which improves percentage of inclusions removal and also generates less shear stress on the walls of the tundish which increases tundish life, Keywords: CFD, Tundish, Fluid flow, RTD analysis INTRODUCTION
flow characteristics. In this method of transfer of molten steel to the mold, molten steel remains in the
Earlier, the tundish in a continuous casting operation
tundish for the time;
was used as a reservoir and distributor of molten steel only but now its role has considerably expanded to
1:
deliver metal of desired cleanliness and composition. The efficiency and
optimization
Tundish Volume
= ---------
(1)
Volumetric Flowrate
of continuous
require a close control of the
This time is known as theoretical average residence
molten steel flow characteristics, inclusion floatation
time'. With the availability of the residence time,
and separation,
adjustment
tundish provides an excellent opportunity to perform
within the tundish. If the flow of metal in the tundish
some metallurgical treatments like inclusion separation
is not properly controlled, it may even deteriorate the
and floatation, alloy trimming of steel, calcium
'quality' of steel produced
in the ladle. Thus,
induced inclusion modification etc., during the process
tundishes, in terms of their shape and use of the flow
of continuous casting. The ideal flow requirements of
control devices (dams, weirs, baffles, striker pads,
steel melt depend
turbostop, etc.) are designed to provide optimum
treatment. For example, plug flow is the best condition
casting operation
and the composition
Research & Development STEEL INDIA
on the type of metallurgical
Centre for Iron & Steel, Steel Authority of India Ltd., Ranchi-834002, 109
India VOL. 34 NO. 2 September
2011
Flow
110
Optimization
for the separation of non-metallic inclusions but is unfavourable
for carrying out dissolution of the
additive and the distribution of the dissolved products into the tundish. Therefore, a detailed knowledge of the behaviour of the steel melt flowing in the tundish is necessary so that the conventional tundish can be made
amenable
to perform
the
metallurgical
treatments during the process of continuous casting. To assess the effectiveness of a given tundish design,
in
the Tundish with the Different
MATHEMATICAL
Mathematical tundish
Combination
MODELlNG
modelling
metallurgy
of Flow Modifiers
STUDIES
of various
has
been
aspects
carried
of out.
Considerable efforts have been made through these studies
to design
a turbostop
to increase
the
efficiency of the continuous casting tundish system. For the convenience
of the present
discussion,
mathematical model studies have been summarised in th is section under three main headings namely;
researchers have simulated the metal flow either mathematically or physically, before actually using the design in actual industrial production. Mathematical
Governing
Equation
Fluid Flow:
modelling has been used by many researchers for flow The flow profile of the tundish is calculated using
predictions inside the tundish.
standard
turbulent
equations
of the k-s model
SCOPE OF WORK
expressed in their three-dimensional
The present work deals with the study of steel flow in
standard wall functions.
the
tundish
under
isothermal
condition.
version with
The (2)
mathematical simulations have been carried out for 1I4th scaled down tundish of Rourkela Steel Plant (RSP) with different flow modifiers. RSP tundish of
(3)
30T capacity presently having striker pad and baffle as flow modifiers. Fluid flow has been simulated using k- turbulent model with species transport for residence discrete
time distribution phase
model
to
(RTD) analysis investigate
inclusion
floatation. In this study different combination of flow modifiers
(striker pad, baffle, and turbostop)
evaluated
and
optimum
combination
modifiers has been determined
Residence
Time Distribution
Residence
Time Distribution
and
is
of flow
for RSP tundish.
Material properties of the liquid steel and modelling parameters used in the mathematical simulation are
representation
(RTD) is statistical
of the time spent by an arbitrary
volume of the fluid in the tundish. To theoretically evaluate the process performance
of continuous
casting tundish systems, residence time distributions have been
predicted
under
a wide variety of
conditions by mathematically simulating the pulse injection of an inert tracer in to a tundish. To this end,
given in the Table-I.
the variation of mass fraction or concentration of an Table-L: Physical Properties of liquid steel and Modelling Parameter
function
Property
Value
Density
7000
Unit Kg/m3
Molecular Viscosity
0.00555
Kgm
3500
Kg/rrJl
Inclusion density Parameters Mass flow rate in actual tundish Mass flow rate in model tundish
VOL. 34 NO. 2 September
2011
injected tracer, i, within the tundish is estimated as a
40.28 1.26
dimensional,
convection
three
+ diffusion equation, as
given below
.15.1
a(rm)
a(pum) a(pum) a(pum)
---+----+----+---at ax ay
Kg/s Kg/s
of time by solving a transient,
=
a (r a;;:
(r ayam;) +az-a (r azam,)
am;) +a-y a
effax
dz
eff
eff
(4)
STEEL INDIA
111
Sujata Devi, R. K. Singh, A. Paul, T K. Pratihar & S. K. Srivastava
Inclusion Transport and Separation
formulation utilized segregated steady state solver
The movement of inclusion particles in the liquid
with implicit formulation.
melt is tracked using the Discrete Phase Model. The
standard
Discrete Phase Model in essence is a combined
pressure-velocity
Lagrangian-Eulerian
The
SIMPLE algorithm. The solution is started with
equation for the inclusion transport can be written as:
default under-relaxation parameters, which are later
calculation
procedure.
discretization
For the pressure term, scheme
was used.
coupling is achieved
The
by using
reduced to get the converged solution. Convergence dU/ dt
=-~~C 4
P/
d
criteria of le-06 are fixed for all equations. Solution is
Re (Up -U.) 2 b
d
b ~
I
+
I
'-----v---------' Drag
rr dU
i
_
PP
Pr essure
~2..L(dU,' 2 Pg
dt "----..r---'
and w velocity reached
dU,
dt
dt
)+ (1- ~)g
'--y---
force
Virtual
considered converged when the residuals for u, v,
force
p
~~ mass
force
Bouyancy
below the convergence
criteria. For inclusion transport
600 particles of
different sizes are injected into the tundish from inlet I
P force
and allowed to reflect on the walls. For the residence time distribution model, the solution is considered converged when the residuals for the mass fraction of
Initial and Boundary Conditions
tracer
is below
le-12.
Time step of Is with
The boundary conditions for momentum transfer are
approximately 100 iterations in each time step for a
those of no slip at the solid surfaces, and zero normal
total duration of 2000 s is utilized to generate RTD
velocity gradients on the free surface of the liquid. In
curves.
a similar way, both k and turbulent
are assigned through
intensity (5%) and hydraulic diameter
Design Configuration
correlations2. The initial guess values of velocities in
In the proposed tundish configurations, the location
three dimensions of tundish are estimated from the
of the existing metal feeding and discharge points
inlet velocity.
have been kept unchanged.
For inclusion floatation simulation,
A turbostop has been
particles are allowed to reflect from all the walls of the
designed, and the detailed design of the turbostop
tundish and are allowed to escape from the outlet.
used in the present study is shown in Fig. 1(a-d). The
However, once they touch the free surface, they are
turbostop,
trapped. For residence time distribution, all walls,
shroud. Alldimensions shown in the figure is in mm.
if present, is placed exactly below the
and free surface are considered to have zero diffusive Ut~jGn 01 Turhl)Sll)p
flux. This means, that the tracer is allowed to be injected only at the inlet and can escape only through
I---'~ ---j I
T-='
r-'~---i
~:,;;-r
the outlet.
rzr
Numerical
Solution
of
the
Governing
Equations
~'h--r
.
I T
",,-, i.