materials

Article
Effect of Lance Structure on Behavior of
Coherent Jet in EAF Steelmaking Process
Fuhai Liu 1,2, *, Rong Zhu 2 , Guangsheng Wei 2, * and Shiliang Fan 3
1 National Center for Materials Service Safety, University of Science and Technology Beijing,
Beijing 100083, China
2 Key Laboratory of Fluid Interaction with Material, Ministry of Education, University of Science and
Technology Beijing, Beijing 100083, China; zhurong1206@126.com
3 Xining Special Steel Co., Ltd, Qinghai, Xining 810005, China; fanman2006@126.com
* Correspondence: liufuhaisteel@126.com (F.L.); weiguangsheng@ustb.edu.cn (G.W.)




Received: 5 December 2019; Accepted: 18 February 2020; Published: 26 February 2020


Abstract: During the electric arc furnace steelmaking process, the coherent jet technology was widely
used to protect the kinetic energy of the supersonic oxygen jet and achieve better mixing effects.
Comparing with the conventional oxygen lance, the coherent lance could increase the surface area of
impaction cavity, resulting in a better stirring effect and higher reaction rate. However, there was
limited research about the effect of restriction structure for the coherent lance tip on the flow field
characteristic of the main oxygen jet. In this research, three kinds of restriction structures have
been investigated by numerical simulation and combustion experiment at room and high ambient
temperature conditions. Then an optimum restriction structure would be tested in a 75 t electrical arc
furnace steelmaking process to verify its metallurgical property.

Keywords: coherent lance; electric arc furnace; numerical simulation; combustion experiment

1. Introduction
The oxygen lance plays an important role in the electric arc furnace (EAF) and basic oxygen
furnace (BOF) steelmaking processes, which delivers the oxygen into the molten bath under supersonic
condition [1,2]. Before the supersonic jet reaches the molten bath, its velocity keeps reducing, because
of an entrainment phenomenon between the jet and the surrounding ambient gas, which lowers its
impaction ability and oxidation rate [3,4]. To suppress this phenomenon, a shrouding gas flow has
been adopted to reduce the environmental resistance for the supersonic oxygen jet. The shrouding gas
flow, including gas/liquid fuel and oxygen, forms a combustion flame within the high-temperature
environment of the furnace, and then a low-density zone is generated at the center of the flame.
As a result, the potential velocity core for the supersonic oxygen jet will be prolonged, and this method
to maintain the velocity of the supersonic oxygen jet is defined as the coherent jet technology, and the
oxygen lance using the coherent jet technology has been defined as the coherent oxygen lance [5–7].
Nowadays, the coherent jet technology is widely used in the EAFs to control the oxygen supply process,
and its metallurgical and operational benefits have been well reported [8,9].
As coherent lance is one of main apparatuses used in the EAF steelmaking process, there has
been a lot of research regarding the effect of the shrouding nozzle arrangement, the process control,
and the initial working parameters on the behavior of the coherent jet. During the research process,
the application of coherent lances is hard to analyze by water experiment or multi-phase flow simulation,
because these methods do not easily reproduce the combustion process. Therefore, the numerical
simulations, combustion experiments, and industrial applications are always used to analyze the flow
field characteristic of the coherent jet [10,11]. There are two main methods which are currently adopted

Materials 2020, 13, 1043; doi:10.3390/ma13051043 www.mdpi.com/journal/materials

 Materials 2020, 13, x FOR PEER REVIEW 2 of 15
Materials 2020, 13, 1043 2 of 15

45 currently adopted to improve the stirring ability of coherent jets. The first one involves augmenting
46 theimprove
to total energy by adding
the stirring more
ability methane jets.
of coherent or byTheincreasing
first onethe initial augmenting
involves temperature of thethe input
total gas by
energy
47 flow. The
adding moresecond
methane oneorisbyoptimizing
increasing the the initial
arrangement
temperature for shrouding
of the inputnozzles,
gas flow.orThe selecting
second one an is
48 appropriate characteristic curve for the Laval nozzle to suppress the generation
optimizing the arrangement for shrouding nozzles, or selecting an appropriate characteristic curve for of the shock waves
49 for the
the main
Laval oxygen
nozzle [12,13]. the generation of the shock waves for the main oxygen [12,13].
to suppress
50 Liu
Liu et
et al.
al. [14]
[14] have
havedesigned
designedaashrouding
shroudingnozzlenozzleusingusinga asupersonic
supersonic structure
structure forfor
thethecoherent
coherent
51 lances by means of simulation and experimental research, and the result
lances by means of simulation and experimental research, and the result showed the supersonic showed the supersonic
52 structure used
structure used inin this
this method
method improved
improved the thepotential
potentialcorecorelength.
length. Odenthal
Odenthal etetal.
al.[15]
[15]has
hasexamined
examinedthe
53 the development of coherent jet technology for EAF processes, and has analyzed
development of coherent jet technology for EAF processes, and has analyzed the advantages of having the advantages of
54 having an SIS (Siemag Injection System) coherent lance in terms of reducing the natural gas
an SIS (Siemag Injection System) coherent lance in terms of reducing the natural gas consumption
55 consumption and increasing supersonic oxygen jet impaction ability. Klioutchnikov et al. [16] had
and increasing supersonic oxygen jet impaction ability. Klioutchnikov et al. [16] had analyzed the
56 analyzed the flow field of coherent jet in hot-temperature environment, and achieved a formula to
flow field of coherent jet in hot-temperature environment, and achieved a formula to describe the
57 describe the length of velocity potential core for the coherent jet when the Mach number for both the
length of velocity potential core for the coherent jet when the Mach number for both the main and the
58 main and the shrouding jet was same. Hu et al. [17] has proposed a mixed injection method for the
shrouding jet was same. Hu et al. [17] has proposed a mixed injection method for the shrouding jet to
59 shrouding jet to reduce its methane consumption, and has specified an appropriate mixing rate for
reduce its methane consumption, and has specified an appropriate mixing rate for N2 and CH4 in the
60 N2 and CH4 in the control process. However, few studies have examined how to design an effective
control process. However, few studies have examined how to design an effective structure to channel
61 structure to channel the main and shrouding jet for increasing the impaction ability of the coherent
the main and shrouding jet for increasing the impaction ability of the coherent jet, after they passed
62 jet, after they passed through the nozzles.
63 through the nozzles.
In this research, we first determine the positioning and flow rate for both the main and
64 In this
shrouding jets.research,
Then, wewe first determine
proposed threethe positioning
kinds and flowstructures
of the restriction rate for both thecoherent
at the main and shrouding
lance tip,
65 jets. Then, we proposed three kinds of the restriction structures at the coherent
and compare their performance to a conventional lance (supersonic oxygen lance) by a series of lance tip, and compare
66 their performance
numerical to a conventional
simulations and combustion lance (supersonicThe
experiments. oxygen
axiallance) by aand
velocity series of numerical
static temperature simulations
of the
67 and
maincombustion
oxygen jet at experiments.
the centerline Theareaxial velocity
measured inand static temperature
combustion experiment,oftothe mainthe
verify oxygen
accuracy jet at
ofthe
68 centerline are measured in combustion experiment, to verify the accuracy of the
the simulation data. Based on the results, a proper restriction structure is tested in an industrial simulation data. Based
69 on the results,
application a proper the
by assessing restriction
steelmakingstructure
time, is tested in an industrial
dephosphorization application
rate, and by assessing
mass fraction of FeO inthe
70 steelmaking
the resulting time,
slag. dephosphorization rate, and mass fraction of FeO in the resulting slag.

71 2. Structural Design
2. Structural Designand
andValidation
Validation

72 The Figure
The Figure 11 showed
showedthree
threekinds
kindsofofcoherent
coherentlances
lancesand
andoneone kind
kind ofof conventional
conventional supersonic
supersonic
73 oxygen
oxygen lance. For the
lance. For the three
three kinds
kinds of
of coherent
coherent lance
lance structures,
structures, the
the positioning
positioning of
of their
their arrangements
arrangements for
74 main oxygen and shrouding nozzles were same, as shown in Figure 1d. The key
for main oxygen and shrouding nozzles were same, as shown in Figure 1d. The key distinction distinction between
75 them was the shape of the restriction structure for the coherent lance tip. Hereafter,
between them was the shape of the restriction structure for the coherent lance tip. Hereafter, the the coherent
76 lance depicted
coherent in Figurein1a,b
lance depicted would
Figure 1a,bbe referred
would to as a flap
be referred to asand an and
a flap openanrestriction structure
open restriction coherent
structure
77 lance, respectively,
coherent or simplyor
lance, respectively, assimply
a ‘flap coherent
as a ‘flap lance’ andlance’
coherent ‘openandcoherent
‘open lance’ for lance’
coherent short, for
respectively.
short,
78 The coherent The
respectively. lancecoherent
depictedlance
in Figure 1c,e would
depicted in Figurebe addressed
1c,e wouldasbe a conventional
addressed ascoherent lance and
a conventional
79 acoherent
conventional lance,
lance and respectively.lance, respectively.
a conventional

(a) (b)
80 Figure 1. Cont.
 Materials 2020, 13, 1043 3 of 15
Materials 2020, 13, x FOR PEER REVIEW 3 of 15

(c) (d) (e)

81 Figure
Figure 1. 1. Cross-sectional and
Cross-sectional andfront
frontview
viewof the coherent
of the lanceslances
coherent with various structures.
with various (a) The cross-sectional
structures. for
(a) The cross-
82 a flap restricted
sectional for a flapcoherent lance;
restricted (b) cross-sectional
coherent lance; (b)for an open restricted
cross-sectional coherent
for an openlance; (c) cross-sectional
restricted for a
coherent lance;
83 (c) conventional
cross-sectionalcoherent
for alance; (d) front view
conventional of the whole
coherent lance;coherent
(d) frontlances;
view(e)ofcross-sectional for a conventional
the whole coherent lances;
84 lance.
(e) cross-sectional for a conventional lance.

85 TheThe design
design Mach
Mach number
number ofof
thethe mainoxygen
main oxygenflow
flowwas
was2.0,
2.0,with
withaathroat
throatdiameter
diameter ofof 24.6
24.6 mmmm
86 and an exit diameter of 31.8 mm. The shrouding nozzles were respectively distributed
and an exit diameter of 31.8 mm. The shrouding nozzles were respectively distributed around three around three
87 concentric
concentric circles,
circles, as as presented
presented inin Figure.
Figure 1. 1.The
Theinner
innerand
andouter
outernozzles
nozzleswere
were injected
injected an
an OO22 flow,
flow,
88 and the intermediate nozzle injected a CH 4 flow. The diameter of these concentric rings, from inside
and the intermediate nozzle injected a CH4 flow. The diameter of these concentric rings, from inside to
89 to outside,
outside, were49.5,
were 34.0, 34.0,and
49.5, andmm,
128.0 128.0respectively.
mm, respectively. The exactly
The exactly sameparameters
same nozzle nozzle parameters were
were applied
90 applied
in both in both combustion
combustion experimentexperiment and the numerical
and the numerical simulation.simulation. Theliteratures
The previous previous literatures have
have provided
91 provided the details of the experimental equipment and its control process used in this
the details of the experimental equipment and its control process used in this paper [12–14], and only paper [12–14],
92 and only
a briefly a briefly explanation
explanation has been
has been provided inprovided
this paper.in this paper.
93 The pitot tube and thermocouple were fixed at specific locations to measure the pressure and
The pitot tube and thermocouple were fixed at specific locations to measure the pressure and the
94 the static temperature of the main oxygen jet. The pitot tube was designed with a water-cooled
static temperature of the main oxygen jet. The pitot tube was designed with a water-cooled structure.
95 structure. When the dynamic and static pressure were measured by the pitot tube, the Equations (1)
When the dynamic and static pressure were measured by the pitot tube, the Equations (1) and (2) were
96 and (2) were used to calculate the axial velocity of the coherent jet [18]
used to calculate the axial velocity of the coherent jet [18]
t2 RT   p! ( 1)/ 
v
Vaxial = 2RTγ p0 0(γ−1 )/γ
 1 , Ma＞0.3
=   1  p 
(1)
Vaxial − 1, Ma > 0.3 (1)
γ−1  p 
s
(pp−p0p)0 )
22(
Vaxial==
Vaxial , Ma < 0.3
, Ma<0.3 (2)
(2)
ρ
where, p and p0 are the dynamic and static pressure of the supersonic jet, respectively. Vaxial and ρ are
97 thewhere, p and p0and
axial velocity are the dynamic
density and
of the static
gas flow.pressure
R and Tofare
theideal
supersonic jet, respectively.
gas constant and staticVtemperature,
axial and ρ are

98 respectively.
the axial velocity
γ is theand
ratiodensity of the
of oxygen gas flow.
specific heatRbeing
and T1.4.
are ideal gas constant and static temperature,
99 respectively. γ is the ratio of oxygen specific heat being 1.4.
3. Numerical Simulation
100 3. Numerical Simulation
3.1. Governing Equations
101 3.1. Governing Equations
In this paper, the following governing equations were used for the simulation process to integrate
102 In this paper, the
the Reynolds-averaged following governing
Navier–Stokes equations were used for the simulation process to
(RANS) [19]:
103 integrate the Reynolds-averaged
Continuity equation: Navier–Stokes (RANS) [19]:
104 Continuity equation: ∂ρ →
+ ∇(ρ v ) = 0 (3)
∂t

 (  v )  0 (3)
t
Materials 2020, 13, 1043 4 of 15

Momentum equation:

∂ →→ → →
(ρvi ) + ∇(ρ v v ) = −∇P + ∇[µ(∇ v )] + ρ g (4)
∂x j

Energy equation:

∂(ρE) →
X → → →
+ ∇[ v (ρE + ρ)] = ∇(ke f f ∇T − hi Ji + (τe f f · v )) + Sh (5)
∂t
j

→
where, ρ and µ are the density and kinetic viscosity for fluid, respectively; Ji and hi are the diffusion
→
flux and enthalpy of species i; v is the instantaneous velocity. keff is the effective thermal conductivity;
→
Sh represents the heat of chemical reaction; R and g are the ideal gas and acceleration of gravity
constant being 8.31 and 9.81, respectively; E is the total energy.
When the main oxygen jet passed through the Laval nozzle, its high pressure energy was
transformed into its kinetic energy, and then its temperature and density also changed. Thus, the gas
phases should be defined as the ideal gas for the simulation process, and their parameters could be
calculated by the equation [20]
PM
ρ= (6)
nRT
where, M, P, and n are the mass, total pressure, and mole number of the gas, respectively.
The realizable k-ε turbulence model was adopted to calculate the simulation data. The following
equations were used to calculate turbulence kinetic energy (k) which was used to construct its dissipation
rate (ε) [19]
∂ ∂ µt ∂k
(ρkui ) = [(µ + ) ] − Gk + Gb − ρε − YM (7)
∂xi ∂x j σκ ∂xi
∂(ρεui ) ∂ µt ∂ε ε2 ε
!
= µ+ + ρC1ε Sε − ρC2ε √ + C1ε C3ε Gb (8)
∂xi ∂x j σε ∂xi κ + uε κ
where, C1ε , C2ε , C3ε , σk , and σε are the constants, with values of 1.44, 1.92, 0.09, 1.0, and 1.2,
respectively [19,21]. Gk and Gb are the generation of turbulence kinetic energy due to the mean velocity
gradient and buoyancy. YM is the contribution of the fluctuating dilatation in the compressible
turbulence process, as defined by the equation

k
YM = 2ρε (9)
γRT

The eddy dissipation concept (EDC) model with the multi-step chemical kinetic mechanisms
(GRI Mech, 3.0, c/o Gas Research Institute, Chicago, IL, USA) was used to simulate the combustion
reaction. The transport equation of mass fraction for each species in the chemical mechanism could be
presented as [22]:
∂ → →
(ρYi ) + ∇(ρ v Yi ) = −∇ Ji + Ri (10)
∂t
where Yi is the local mass fraction of each species i, and Ri is the net rate of species i production by
chemical reaction, and Ji is the diffusion flux of each species i.

3.2. Simulation Details

Figure 2 depicted a 3D computational domain grid of the flap coherent lance used in this
study. In this research, all the cell structures were used hexahedral mesh, and its orthogonal quality
(checked in the Fluent) was 8.9–9.1. The main oxygen jet and shrouding jet flowing zones were used as
the computational domain, and the exit diameter of the main oxygen Laval nozzle was defined as 1 De.
128 chemical reaction, and Ji is the diffusion flux of each species i.

129 3.2. Simulation Details

130 Figure 2 depicted a 3D computational domain grid of the flap coherent lance used in this study.
131 In this research,
Materials all the cell structures were used hexahedral mesh, and its orthogonal quality (checked
2020, 13, 1043 5 of 15
132 in the Fluent) was 8.9–9.1. The main oxygen jet and shrouding jet flowing zones were used as the
133 computational domain, and the exit diameter of the main oxygen Laval nozzle was defined as 1 De.
The main computational domain ranged from the tip of the Laval nozzle to 85 De downstream in
134 The main computational domain ranged from the tip of the Laval nozzle to 85 De downstream in an
an axial direction and 12.5 De in a radial direction. Under the initial condition, the air filled the whole
135 axial direction and 12.5 De in a radial direction. Under the initial condition, the air filled the whole
computational domain. Tables 1 and 2 showed the boundary conditions and the gas thermophysical
136 computational domain. Table 1 and Table 2 showed the boundary conditions and the gas
properties used in the simulation model, respectively.
137 thermophysical properties used in the simulation model, respectively.

138
139 Figure
Figure 2. 2. Structuregrid
Structure grid of
of the
the computational
computational domain.
domain.

140 Table1.1.Boundary
Table conditions.
Boundary conditions.

Nameof
Name of Boundary TypeType of Boundary Conditions
of boundary Values
Values
boundary conditions
Mass flow rate 0.9921 kg/s
Main oxygen inlet Mass flowMass rate fractions 0.9921 = 100%
O2 kg/s
Main oxygen Oxygen temperature 298 K
Mass fractions O2 = 100%
inlet Mass flow rate 0.1190
Oxygen
Shrouding gas inlet temperature 298 K kg/s
Mass fractions O2 = 100%
Shrouding (Ogas2) Mass flow rate 0.1190 kg/s
Gas temperature 298 K
inlet Mass fractions O2 = 100%
Mass flow rate 0.0595 kg/s
(O2)
Shrouding gas inletGas temperature 298
Mass fractions CH4K= 100%
Shrouding(CHgas4 ) Mass flow rate
Gas temperature 0.0595 298
kg/sK
inlet Mass fractions
Static pressure CH4 =101325
100%Pa
(CH4)Outlet Gas temperature
Mass fractions = 23%,
O2 298 K N2 = 77%
Ambient
Static pressure temperature 300 K,
101325 Pa 1700 K
Outlet Mass fractions O2 = 23%, N2 = 77%
Table 2. Thermo-physical
Ambient temperature properties for gases.
300 K, 1700 K
O2 CH4 Air
141
Density/(kg·m−3 ) Ideal gas Ideal gas Ideal gas
142 Cp/(J·kg−1 ·K−1 ) Piecewise-polynomial Piecewise-polynomial Piecewise-polynomial
Molecular weight/(kg·kgmol−1 ) 31.999 16.043 28.966
Standard state enthalpy/(J·kgmol−1 ) 0 −7.490 × 107 -
Standard state entropy/(J·kgmol−1 ·K−1 ) 2.050 × 105 1.864 × 105 1.934 × 105

A pressure-based solver using a steady-state model was employed to calculate the RANS
equation. The standard spatial discretization approach was used to calculate pressure values, and other
variables, including energy, turbulent kinetic energy, and dissipation rate were solved using a QUICK
(Quadratic Upstream Interpolation for Convective Kinematics) scheme. The walls were defined as the
standard wall at the non-slip condition. The discrete ordinate (DO) model was considered for radiation
phenomenon to achieve noticeable effects on the results. Moreover, the weighted sum of gray gas
(WSGG) model was used to calculate the total emissivity of mixture gases as the function of their
temperatures and pressures.
146 other variables, including energy, turbulent kinetic energy, and dissipation rate were solved using a
147 QUICK (Quadratic Upstream Interpolation for Convective Kinematics) scheme. The walls were
148 defined as the standard wall at the non-slip condition. The discrete ordinate (DO) model was
149 considered for radiation phenomenon to achieve noticeable effects on the results. Moreover, the
150 Materials
weighted 2020,
sum13, 1043 6 of 15
of gray gas (WSGG) model was used to calculate the total emissivity of mixture gases
151 as the function of their temperatures and pressures.
152 Solution convergence
Solution convergence for forthe
thenumerical
numericalmodel
modelwere determined
were when
determined whenthe the
energy residual
energy was
residual
153 <10 −7 and residuals for other variables was <10−5, with consecutive iterations for temperature and
was <10 and residuals for other variables was <10 , with consecutive iterations for temperature
−7 −5
154 velocity
and at the
velocity at outlet of the
the outlet of computational domain
the computational being
domain <1.0<1.0
being K and 1.0 m/s,
K and respectively.
1.0 m/s, respectively.

155 3.3. Mesh

3.3. Mesh Independency Test
Independency Test
156 To insure
To insure the
the accuracy
accuracy ofof simulation
simulation data,
data, the
the grid
grid independency
independency of of the
the simulation
simulation model
model has has
157 been tested using three kinds of grid levels: coarse grid (322,643 cells), medium grid (616,837
been tested using three kinds of grid levels: coarse grid (322,643 cells), medium grid (616,837 cells), cells),
158 and fine
and fine grid
grid(856,715
(856,715cells).
cells).Figure
Figure3 3depicted
depictedthe
the axial
axial velocity
velocity distributions
distributions of the
of the coherent
coherent jets jets at
at the
159 thecenterline,
jet jet centerline, using
using those
those gridgrid levels.
levels.

160
161 Figure 3. Axial
Figure velocity
3. Axial distribution
velocity distributionof
of the coherentjets
the coherent jetsusing
using different
different mesh
mesh levels.
levels.

Figure 3 showed that the axial velocity variation between the coarse and the medium grid was
162 Figure 3 showed that the axial velocity variation between the coarse and the medium grid was
found to be 19.1% X/De. This variation between the medium and the fine grid was found to be <1%,
163 found to be 19.1% X/De. This variation between the medium and the fine grid was found to be <1%,
which means that the simulation was not sensitive to their physical properties. The computational time
164 which means that the simulation was not sensitive to their physical properties. The computational
required to calculate using the fine grid was approximately 2.1 times greater than the time required
165 time required to calculate using the fine grid was approximately 2.1 times greater than the time
166 to calculate
required using theusing
to calculate medium grid level,
the medium gridand asand
level, a consequence the medium
as a consequence grid grid
the medium was was
chosen for
chosen
167 further studies.
for further studies.
168 4. Results and Discussion

4.1. Velocity Profile

Figure 4 shows the axial velocity profiles of coherent jets using various restriction structures at
different ambient temperatures. The simulation results are shown by solid and dotted lines, whilst the
measurement data is represented by various symbols (, #, 4, and 5). Room and high ambient
temperature are defined as 300 K and 1700 K, respectively. The results show that the experimental data
is in a good agreement with the simulation results.
The velocity potential core length of the main oxygen jet at room ambient temperature using
a conventional coherent lance, an open coherent lance and a flap coherent lance were 24 De, 27 De,
and 31 De, respectively. At a high ambient temperature, the velocity potential core length of the main
oxygen jet using a conventional coherent lance, an open coherent lance and a flap coherent lance were
29 De, 31 De, and 34 De, respectively. For the comparison, the velocity potential core length of the
main oxygen jet using a conventional lance at room and high ambient temperature were 11 De and
16 De, respectively. The average potential core lengths formed by a coherent lance with a restriction
structure were 1.2- and 2.3-times larger than those generated by a conventional coherent lance and
a conventional lance, respectively. Thus, it can be concluded that a coherent lance with a restriction
structure is able to prolong the velocity potential core length of the main oxygen jet. The coherent
lance with a flap restriction structure is able to further improve this trend, when compared to an open
restriction structure.
170 4.1. Velocity Profile
171 Figure 4 shows the axial velocity profiles of coherent jets using various restriction structures at
172 different ambient temperatures. The simulation results are shown by solid and dotted lines, whilst
173 the measurement data is represented by various symbols (□, ○, △, and ▽). Room and high ambient
174 temperature
Materials are defined as 300 K and 1700 K, respectively. The results show that the experimental
2020, 13, 1043 7 of 15
175 data is in a good agreement with the simulation results.

(a) (b)
176 Figure4.4. Axial
Figure Axialvelocity
velocity distribution of theofcoherent
distribution jets using
the coherent jetsvarious
using restriction structures atstructures
various restriction room and high
at
177 ambient temperature: (a) at an ambient temperature of 300 K; (b) at an ambient temperature
room and high ambient temperature: (a) at an ambient temperature of 300 K; (b) at an ambient is 1700 K.
temperature is 1700 K.
178 The velocity potential core length of the main oxygen jet at room ambient temperature using a
179 The lengthcoherent
conventional of the velocity
lance, potential core for the
an open coherent lancemain
andoxygen jet at thelance
a flap coherent high were
ambient temperature
24 De, 27 De, and
180 using
31 De,a conventional
respectively.lance,
At a ahighconventional coherent lance,
ambient temperature, theanvelocity
open coherent
potential lance
coreand a flapofcoherent
length the main
181 lance were 1.45-, 1.21-, 1.15-, and 1.10-times longer than those were at room
oxygen jet using a conventional coherent lance, an open coherent lance and a flap coherent lance were ambient temperature
182 respectively.
29 De, 31 De, Hence,
and 34although a higher ambient
De, respectively. For thetemperature
comparison,will thegenerally prolong the
velocity potential corepotential
length core
of the
183 length,
main oxygen jet using a conventional lance at room and high ambient temperature were 11 Deflap
it has the greatest impact upon a conventional lance and the smallest impact upon a and
184 coherent lance.
16 De, respectively. The average potential core lengths formed by a coherent lance with a restriction
185 To assess
structure werethe effect
1.2- and the axial velocity
2.3-times larger thanof the
those shrouding
generatedflame, an annulus was
by a conventional created
coherent lancein and
the a
186 simulation model. The annulus was located 100 mm in an axial direction
conventional lance, respectively. Thus, it can be concluded that a coherent lance with a restriction from the exit of the main
187 oxygen Laval
structure is nozzle
able to and had athe
prolong maximum
velocityand minimum
potential coreradius
lengthofof46.0themmmainand 33.8 mm,
oxygen jet. respectively.
The coherent
188 The average
lance with astatic temperature
flap restriction and axial
structure velocity
is able of theimprove
to further gas flowthis
overtrend,
this annulus at room and
when compared to anhigh
open
189 ambient temperature
restriction structure.using different coherent lances are shown in Table 3. The results show the average
190 static temperature
The length of ofthe
thevelocity
gas flowpotential
over thecoreannulus
for the at room and high
main oxygen jetambient temperature
at the high was much
ambient temperature
191 higher than the inlet temperature of the main oxygen jet (298 K). This
using a conventional lance, a conventional coherent lance, an open coherent lance and a flap proves that the gas flow over
coherent
192 the annulus
lance were is the shrouding
1.45-, 1.21-, 1.15-,flame, which is not
and 1.10-times longerthe than
mainthose
oxygen were jet.atThe
room results
ambientalso temperature
show that
193 a respectively.
restriction structure is able toaincrease
Hence, although the axial
higher ambient velocity ofwill
temperature the generally
shrouding flame, the
prolong comparedwith
potential core
194 the conventional
length, it has the coherent
greatestlance.
impact At upon
the same time, a flap restriction
a conventional lance and structure
the smallest alsoimpact
furtheruponincreases
a flap
195 this tendency.
coherent lance.
196 To assess the effect the axial velocity of the shrouding flame, an annulus was created in the
197 Average
Table 3. The
simulation model. static was
annulus temperature
locatedand 100axial
mmvelocity for various
in an axial coherent
direction fromlances.
the exit of the main
198 oxygen Laval nozzle and had a maximum Conventional and minimum radius of 46.0
Open Coherent mm and 33.8 mm, respectively.
Flap Coherent
199 Label
The average static temperature and axialCoherent velocity Lance
of the gas flowLanceover this annulusLance at room and high
200 ambient temperature using different
Average static coherent lances are shown in Table 3. The results show the
201 averageRoom ambient
static temperature of the gas flow over 915the annulus at room 1482 and high ambient 1934 temperature
temperature (K)
temperature
202 was much higher than Average the inletaxial
temperature of64the main oxygen 107 jet (298 K). This proves 223 that the gas
203 velocity
flow over the annulus is the shrouding (m/s) flame, which is not the main oxygen jet. The results also show
Average static
204 that aHigh ambientstructure is able to increase the
restriction temperature (K)
1298axial velocity of 1434the shrouding flame, 1646comparedwith
205 temperature
the conventional coherent lance. At the same time, a flap restriction structure also further increases
Average axial
94 167 300
206 this tendency. velocity (m/s)

4.2. Static Temperature Profile

Figure 5 represents the static temperature profiles of the coherent jets using the various restriction
structures at different ambient temperatures. In this study, the static temperature of the oxygen jet was
only measured at the room ambient temperature, because the thermocouple might be damaged during
the furnace heating process.
208 4.2. Static Temperature Profile
209 Figure 5 represents the static temperature profiles of the coherent jets using the various
210 restriction structures at different ambient temperatures. In this study, the static temperature of the
211 oxygen
Materials jet13,
2020, was1043only measured at the room ambient temperature, because the thermocouple might
8 of 15be
212 damaged during the furnace heating process.

(a) (b)
213 Figure
Figure 5. 5.Static
Statictemperature
temperature distribution
distributionofofthe
thecoherent jetsjets
coherent using various
using restriction
various structure
restriction at room
structure at and
roomhigh
214 ambient temperature: (a) at an ambient temperature of 300 K; (b) at an ambient temperature
and high ambient temperature: (a) at an ambient temperature of 300 K; (b) at an ambient temperature of 1700 K.
of 1700 K.
215 During the main oxygen jet passes through the Laval nozzle, its pressure and thermal energy
216 During
would the main oxygen
be transferred jet passes
into kinetic energy,through the in
resulting Laval nozzle,
its static its pressure
temperature and thermal
reducing. Then energy
the main
217 would
oxygenbe transferred
jet would beinto kinetic
heated energy,
by the resulting
ambient in its static
temperature temperature
or the reducing.
high-temperature Thenatthe
flame themain
end of
218 oxygen jet would
the potential be heated
core, which makesby the its ambient
static temperature improve.or the high-temperature
At last, the static flame at the endofofthe
temperature
219 themain
potential
oxygen core,
jet which makes itsapproaching
asymptotically static temperature
the ambientimprove. At last, the
temperature, as static
shown temperature
in Figure 5.of the
220 main oxygen
The pointjet asymptotically
at where the static approaching
temperature the ambient
of a coherenttemperature,
jet starts as shown incan
increasing Figure 5.
be defined as the
221 The point at
temperature where thepoint,
increasing staticTIPtemperature
for short.of Thea coherent
locationjet forstarts increasing
the TIP using acan be definedcoherent
conventional as the
222 temperature increasing
lance, an open coherentpoint, TIP and
lance for short.
a flapThe location
coherent for the
lance wereTIP24using
De, a27conventional
De, and 31 De, coherent lance,
respectively.
223 anAtopen coherent lance and a flap coherent lance were 24 De, 27 De, and
a high ambient temperature, the location for the TIP using a conventional coherent lance, an 31 De, respectively. At a high
open
224 ambient
coherenttemperature,
lance and the a flaplocation
coherentfor the TIPwere
lance using29 a conventional
De, 31 De, and coherent
34 De,lance, an open coherent
respectively. Hence, the
225 lance and afor
location flapthe coherent
TIP waslance verywere
close29toDe, the31 De, andfor
location 34the
De,end
respectively. Hence, potential
point of velocity the location forfor
core thethe
226 TIPmain
was oxygen
very close jet. to the no
With location for the end
entrainment effectpoint of velocity
between the mainpotential
oxygen corejetfor
andthe themain oxygen
ambient gas,jet.
the
227 With noenergy
total entrainment
of the effect
main between
oxygen jet theatmain oxygen jetdid
the centerline andnotthechange,
ambientmaking
gas, theits total energy
static of the
temperature
228 main oxygen
stable. When jetthe
at the
maincenterline
oxygen did not change,
jet reaches the endmaking its static
of potential temperature
core, it will absorbstable.
theWhen the main
heat energy from
229 oxygen jet reaches
the flame, because theofend
theirof relative
potential core, it willgradients,
temperature absorb themaking
heat energy from
its static the flame, because
temperature improve.ofAs
230 their
therelative temperature
main oxygen jet fullygradients,
develops,making its staticits static temperature
temperature improve.
will gradually As the
become maintooxygen
equal jet
the ambient
231 fully develops, its static temperature will gradually become equal to the ambient temperature.
temperature.
232 AsAsshown
shown in in
Figure
Figure 5, the maximum
5, the maximum static temperature
static temperature of of
thethe
mainmainoxygen
oxygen jetjet
at atroom
room ambient
ambient
233 temperature
temperature using
using a conventional
a conventional coherent
coherent lance,
lance, ananopen
open coherent
coherent lance
lance andanda flap
a flap coherent
coherent lance
lance
were 1047 K, 1258 K, and 1180 K in an axial direction of 42 De, 46 De, and 48 De, respectively.
For the high ambient condition, the maximum static temperature of the main oxygen jet at room
ambient temperature using a conventional coherent lance, an open coherent lance and a flap coherent
lance were 1410 K, 1610 K, and 1512 K in an axial direction of 45 De, 48 De, and 50 De, respectively.
Hence, a coherent lance using a restriction structure is able to extend the high-temperature zone of
the flame in the axial direction, when compared to a conventional coherent lance. However, the open
coherent lance achieves a higher maximum static temperature of main oxygen jet and also generates
a shorter potential core, when compared to the flap coherent lance. When the shrouding jet passes
through the flap restriction structure, some of its heat energy of the shrouding flame is transformed into
its kinetic energy, resulting in suppressing its maximum static temperature. This energy transformation
resembles in some respects the behavior of the main oxygen jet, as the main oxygen jet passes through
the Laval nozzle. As a result, the maximum static temperature of a coherent jet using a flap restriction
structure is lower than that using an open restriction structure.
As shown in Figures 6 and 7, the shrouding flame would rapidly expand after it passed through
the shrouding nozzles, and a restriction structure can suppress this tendency. The results show that
the energy of the shrouding flame is transported in both axial and radial directions. In this research,
the total energy value using various coherent lances is the same, because the flow rate of the shrouding
245 the Laval nozzle. As a result, the maximum static temperature of a coherent jet using a flap restriction
246 structure is lower than that using an open restriction structure.
247 As shown in Figures 6 and 7, the shrouding flame would rapidly expand after it passed through
248 the shrouding nozzles, and a restriction structure can suppress this tendency. The results show that
249 the energy of the shrouding flame is transported in both axial and radial directions. In this research,
Materials 2020, 13, 1043 9 of 15
250 the total energy value using various coherent lances is the same, because the flow rate of the
251 shrouding O2 and CH4 is unchanged for any cases. A restriction structure can delay the energy
252 O 2 and CH4 isin
transmission unchanged
the radialfordirection
any cases.atAthe
restriction
coherentstructure can delay
lance tip, the energy
resulting in moretransmission
heat energy in
253 transmission
the in the axial
radial direction at thedirection,
coherent which prolongs
lance tip, the in
resulting high-temperature
more heat energy zone of shrouding
transmission jet in
in the the
axial
254 axial direction.
direction, whichDuring thisthe
prolongs process, the main oxygen
high-temperature zonejet
of using a restriction
shrouding jet in thestructure generates
axial direction. more
During
255 heatprocess,
this energy from the shrouding
the main oxygen jetflame,
using which makesstructure
a restriction its static generates
temperature more higher, as shown
heat energy in the
from
256 Figure 5. flame, which makes its static temperature higher, as shown in the Figure 5.
shrouding

257
258 Figure Static
Figure 6.6.Static temperature
temperature profiles
profiles for coherent
for various various jets
coherent
at roomjets at room
ambient ambient(a)temperature:
temperature: conventional
Materials
(a) 2020, 13, x FOR coherent
conventional PEER REVIEW
lance; (b) open coherent lance; (c) flap coherent lance. 10 of 15
259 coherent lance; (b) open coherent lance; (c) flap coherent lance.

260
261 Static
Figure7.7.Static
Figure temperature
temperature profilesprofiles for coherent
for various various jets
coherent
at high jets at high
ambient ambient(a)temperature:
temperature: conventional
262 (a) conventional
coherent lance; (b) coherent lance;
open coherent (b) (c)
lance; openflapcoherent
coherent lance;
lance. (c) flap coherent lance.

263 4.3. Dynamic Pressure Profile

264 As a major factor in the shape of the impaction cavity of the molten bath, the profiles of the
265 dynamic pressure of coherent jets using various restriction structures at different ambient
266 temperatures are showed in the Figure 8. In this paper, the dynamic pressure profile of the coherent
267 jet at a centerline of X/De = 32 has been selected, based on the distance between the lance tip and
260
261 Figure 7. Static temperature profiles for various coherent jets at high ambient temperature: (a) conventional
262 coherent lance; (b) open coherent lance; (c) flap coherent lance.
Materials 2020, 13, 1043 10 of 15

263 4.3. Dynamic Pressure Profile

264 4.3. Dynamic
As a majorPressure
factorProfile
in the shape of the impaction cavity of the molten bath, the profiles of the
265 dynamic pressure of
As a major factor in coherent
the shape jetsofusing various cavity
the impaction restriction
of thestructures at different
molten bath, the profiles ambient
of the
266 temperatures
dynamic pressureare showed in the
of coherent jetsFigure 8. In thisrestriction
using various paper, the dynamicatpressure
structures differentprofile
ambient of temperatures
the coherent
267 jet
areatshowed
a centerline of X/De8. =In32this
in the Figure haspaper,
beentheselected,
dynamic based on the
pressure distance
profile of thebetween
coherentthe
jet lance tip and
at a centerline
268 molten = 32 has been selected, based on the distance between the lance tip and molten bath.
of X/Debath.

(a) (b)
269 Figure
Figure8.8.Dynamic
Dynamicpressure profiles
pressure of the
profiles of coherent jets using
the coherent variousvarious
jets using restriction structures
restriction at room and
structures high
at room
270 ambient
and hightemperature: (a) at an ambient
ambient temperature: temperature
(a) at of 300
an ambient K; (b) at anof
temperature ambient
300 K;temperature of 1700 temperature
(b) at an ambient K.
of 1700 K.
271 At the ambient room temperature, the maximum dynamic pressure value of the coherent jet
272 usingAt a conventional
the ambient room lance,temperature,
a conventionalthe coherent
maximum lance,
dynamican open coherent
pressure valuelance and
of the a flap coherent
coherent jet using
273 lance, were 2.1 ×lance,
a conventional 104, 11.4 × 104, 21.9 × 10
a conventional 4, and 25.0
coherent × 10an
lance, 4 Pa, respectively.
open At theand
coherent lance high room
a flap temperature,
coherent lance,
274 the
were maximum 4
2.1 × 10 ,dynamic 4
11.4 × 10pressure 4
, 21.9 × value
10 , andof the 4
25.0coherent
× 10 Pa, jetrespectively.
using a conventional
At the highlance,
room a conventional
temperature,
275 coherent
the maximumlance,dynamic
an open coherent
pressure lance
value and a flap
of the coherent
coherent lance,awere
jet using 2.5 × 104,lance,
conventional 21.3 ×a10 4, 22.7 × 104,
conventional
276 and 25.4 × 10 Pa, respectively. The results show the maximum dynamic 4
pressures
coherent lance, an open coherent lance and a flap coherent lance, were 2.5 × 10 , 21.3 × 10 , 22.7 × 10
4 of the 4
coherent jet4 ,
277 at
andthe25.4
high 4
× ambient temperature,
10 Pa, respectively. using
The a conventional
results lance, a conventional
show the maximum coherent
dynamic pressures lance,
of the an open
coherent jet
at the high ambient temperature, using a conventional lance, a conventional coherent lance, an open
coherent lance and a flap coherent lance, were 1.19, 1.87, 1.04, and 1.02 times larger than for the coherent
jet at the room ambient temperature, respectively. Therefore, increasing the ambient temperature
could improve the maximum dynamic pressure. Besides, the ambient temperature has the largest and
smallest effect on the maximum dynamic pressures of the coherent jet using a conventional coherent
lance and a flap coherent lance, respectively.
As mentioned in the preview literatures [9,23,24], a supersonic jet with higher dynamic pressure can
form bigger depth and radius of impaction cavity. Based on the result, the dynamic pressure variation
using the different lance structures is noticeable in the range from 0 to 0.9 De in the radius direction, as
presented in Figure 8, and those areas of field under the curve-line are calculated. When the ambient
temperature is 300 K, the integration areas using a conventional lance, a conventional coherent lance,
an open coherent lance and a flap coherent lance, were 1.7 × 104 , 5.0 × 104 , 6.7 × 104 , and 8.2 × 104 Pa,
respectively. When ambient temperature is 1700 K, the integration areas using a conventional lance,
a conventional coherent lance, an open coherent lance and a flap coherent lance, were 2.0 × 104 ,
6.6 × 104 , 7.0 × 104 , and 8.5 × 104 Pa, respectively. The flap coherent lance can form the biggest
impaction cavity, next the open coherent lance, the conventional coherent lance is in the third place,
and the conventional lance is the smallest both at room and high ambient temperature. Therefore,
the flap coherent lance can further improve the surface area between the supersonic jet and molten
bath, which accelerates the dephosphorization and decarburization rate in the steelmaking process.
 Materials 2020, 13, 1043 11 of 15

4.4. Industrial Application

According to the results of the combustion experiment and the numerical simulation research,
the coherent lance using a flap restriction structure achieves greater impaction ability than the other
coherent lance structures. In order to confirm the metallurgical and operational effects for this coherent
lance structure, both the flap and conventional coherent lances were tested in a 75 t electrical arc
furnace. The data from 160 heats was collected during the smelting process for each kind of coherent
lance, with the control processes remaining identical for each heat. For the purposes of comparative
analysis, we collected data including the steelmaking time, the dephosphorization rate and the mass
fraction of FeO in the slag.
The conditions relating to liquid iron (prior to the steelmaking process) and molten steel (after the
steelmaking process) are shown in Table 4. For the smelting with each of the different coherent lances,
the initial conditions for the liquid iron were fundamentally the same, which means that the initial
conditions had no impact on the results of the industrial application. When compared to a conventional
coherent lance, the average [C][O] and steelmaking time when using a flap coherent lance were reduced
Materials
by 7.7%2020,
and 13, x FOR
1.2%, PEER REVIEW
respectively. 12than
This proves that the impaction ability of a flap coherent jet is better of 15
the conventional coherent lance structure. It also confirms the accuracy of the previous combustion
315 experiment and numericalTable 4. Average
simulation conditions for liquid iron and molten steel
results.

Liquid
Table 4. ironconditions for liquid ironMolten
Average steel steel.
and molten
Steelmaking
Label C Temperature [C][O] Temperature
Liquid Iron
P (%) C(%) P(%) Molten Steel time (min)
Steelmaking
Label (%) (K)
Temperature (10 −4)
[C][O]
(K)
Temperature Time (min)
C (%) P (%) C(%) P(%)
(K) (10−4 ) (K)
Flap coherent lance
Flap coherent lance
3.57
3.57
0.132
0.132
1558
1558
0.076
0.076
0.006
0.006
0.0048
0.0048
1883
1883
49.5
49.5
Conventional
3.57 0.131 1559 0.077 0.008 0.0052 1882 50.1
Conventional
coherent lance
3.57 0.131 1559 0.077 0.008 0.0052 1882 50.1
coherent lance
Figure 9 presents the distributions of the phosphorus in the molten steel using two kinds of
316 Figure 9 presents the distributions of the phosphorus in the molten steel using two kinds of
coherent lances. The results show that the content of phosphorus content in the molten steel had
317 coherent lances. The results show that the content of phosphorus content in the molten steel had a
a distribution of 0.005 to 0.007 wt % using the flap coherent lance and 0.006 to 0.010 wt % using the
318 distribution of 0.005 to 0.007 wt % using the flap coherent lance and 0.006 to 0.010 wt % using the
conventional coherent lance. The average phosphorus contents for flap and conventional coherent lance
319 conventional coherent lance. The average phosphorus contents for flap and conventional coherent
were 0.006 and 0.008 wt %, respectively. Thus, both the average value and the range of distribution
320 lance were 0.006 and 0.008 wt %, respectively. Thus, both the average value and the range of
for P content using a flap coherent lance were lower than those using a conventional coherent lance.
321 distribution for P content using a flap coherent lance were lower than those using a conventional
This proves that a flap coherent lance offers better control performance.
322 coherent lance. This proves that a flap coherent lance offers better control performance.

323
324 Figure
Figure 9. P9.content
P content distributionof
distribution ofmolten
molten steel
steelfor
forthe
thetwo
twokinds of coherent
kinds lances.
of coherent lances.

325 Figure 10 shows the profiles of the average CaO, SiO2, FeO, and P content in the end-point slag
326 after the testing with the different lances. As the same slag basicity being 2.1, the average FeO content
327 using a flap coherent lance was 1.3 wt % lower in the end-point slag than that using a conventional
328 coherent lance. That means a flap coherent lance is beneficial to improvements in the metal yield rate.
 Materials 2020, 13, x FOR PEER REVIEW 13 of 15

Materials 2020, 13, 1043 12 of 15

Figure 10 shows the profiles of the average CaO, SiO2 , FeO, and P content in the end-point slag
after the testing with the different lances. As the same slag basicity being 2.1, the average FeO content
using a flap coherent lance was 1.3 wt % lower in the end-point slag than that using a conventional
coherent lance. That means a flap coherent lance is beneficial to improvements in the metal yield rate.
Moreover, the average P content was increased by 0.07 wt % using the flap coherent lance, due to
a bigger dephosphorization rate.
Materials 2020, 13, x FOR PEER REVIEW 13 of 15

331
332 Figure 10. Distribution of the average CaO, SiO2, FeO, and P content in the liquid slag.

333 Figure 11 presents the average outlet temperature of cool-water profiles for two kinds of
334 coherent lances in steelmaking process. When the flow rate of cool-water for two kinds of coherent
335 lances were both 15.0 t/h, the average temperature of cool-water had a distribution of 318.0 K (44.8 °C)
336 to 324.2 K (51.0 °C) using the flap coherent lance and 316.6 K (43.4 °C) to 322.2 K (49.0 °C) using the
337 conventional coherent lance, as shown in the Figure 11. The average temperature of cool-water for
338 flap and conventional coherent lance were 321.0 K (47.8 °C) and 319.6 K (46.4 °C), respectively.
339 As depicted in the Table 4, the average condition for liquid iron is same for both the flap and
340 conventional coherent lance. That means the initial total energy generated by the oxidation reactions
341 between the oxygen gas and elements in the molten bath (carbon, phosphorus, and silicon). The heat
331
342 productivity per unit time formed by the flap coherent lance is bigger than that by the conventional
332
343 coherent lance,
Figureby comparing
Figure
10. theofsteelmaking
10. Distribution
Distribution of the average time.
the average CaO, The
CaO, SiO
SiO thermal
2, FeO, and P radiation
content in the ofliquid
moltenslag.bath is stronger
2 , FeO, and P content in the liquid slag.
344 using the flap coherent lance, which makes its average outlet temperature of cool-water higher.
333
345 Figurethe
Figure
After 1111presents
presents
50th heat, the
thethe average
average
surface outlet
outlet
structures temperature
temperature
of twoofkinds of ofcool-water
cool-water profiles
coherent profiles
for two
lances havefor
kindstwo
no of kinds of
coherent
significant
334
346 coherent
lances in lances
difference, in steelmaking
steelmaking
comparing process.
with process.
theirWhen
initialthe When the
rateflow
flowThen,
status. of rate of
thecool-water
flow cool-water
rate offor two for
cool-water kindstwo
forofkinds of coherent
coherent
the flap coherent
lances
335
347 lances were
were both
lance both
15.0 t/h,
has been 15.0 t/h, the
the average
increased average temperature
to 15.5temperature of cool-water
of cool-water
t/h, for eliminating had a distribution
had a distribution
the average outlet temperature of 318.0
of 318.0 K K
of (44.8 ◦ C)°C)
(44.8
cool-water to
336
348 to 324.2
324.2
using K
two Kkinds
(51.0 ◦
(51.0C)°C) using
ofusing thethe
coherent flap
flap coherent
coherent
lances. In this lance lance
case, and
and
the 316.6
316.6
flap KK(43.4
coherent (43.4◦ °C)to
C)
lance istostill
322.2
322.2 KK(49.0
working ◦
(49.0properly
°C) using
C) using the
the
after
337
349 conventional
conventional
1000 heats, and coherent
coherent
meets the lance,
lance, asasshown
shown
requirement ofin
in itsthe
the Figure
Figure
service life,11.
11. The The average
average
according thetemperature
totemperature
production of cool-water
of cool-water
safety for
for flap
standards
338
350 flap and conventional
and conventional
proposed coherent lance were 321.0 K ◦
(47.8 °C) and 319.6
coherent lance were 321.0 K (47.8 C) and 319.6 K (46.4 C), respectively.
by the plant. K ◦
(46.4 °C), respectively.
339 As depicted in the Table 4, the average condition for liquid iron is same for both the flap and
340 conventional coherent lance. That means the initial total energy generated by the oxidation reactions
341 between the oxygen gas and elements in the molten bath (carbon, phosphorus, and silicon). The heat
342 productivity per unit time formed by the flap coherent lance is bigger than that by the conventional
343 coherent lance, by comparing the steelmaking time. The thermal radiation of molten bath is stronger
344 using the flap coherent lance, which makes its average outlet temperature of cool-water higher.
345 After the 50th heat, the surface structures of two kinds of coherent lances have no significant
346 difference, comparing with their initial status. Then, the flow rate of cool-water for the flap coherent
347 lance has been increased to 15.5 t/h, for eliminating the average outlet temperature of cool-water
348 using two kinds of coherent lances. In this case, the flap coherent lance is still working properly after
349 1000 heats, and meets the requirement of its service life, according to the production safety standards
350 proposed by the plant.

351
352 11. Average
Figure
Figure outlet
11. Average temperature
outlet temperatureof
ofcool-water profilesfor
cool-water profiles for the
the two
two kinds
kinds of coherent
of coherent lances.
lances.

As depicted in the Table 4, the average condition for liquid iron is same for both the flap and
conventional coherent lance. That means the initial total energy generated by the oxidation reactions
between the oxygen gas and elements in the molten bath (carbon, phosphorus, and silicon). The heat
productivity per unit time formed by the flap coherent lance is bigger than that by the conventional
Materials 2020, 13, 1043 13 of 15

coherent lance, by comparing the steelmaking time. The thermal radiation of molten bath is stronger
using the flap coherent lance, which makes its average outlet temperature of cool-water higher.
After the 50th heat, the surface structures of two kinds of coherent lances have no significant
difference, comparing with their initial status. Then, the flow rate of cool-water for the flap coherent
lance has been increased to 15.5 t/h, for eliminating the average outlet temperature of cool-water
using two kinds of coherent lances. In this case, the flap coherent lance is still working properly after
1000 heats, and meets the requirement of its service life, according to the production safety standards
proposed by the plant.
Based on the industrial application result, the impaction ability of coherent jet using a flap coherent
lance is better than that using a conventional coherent lance, which is consistent with the results of the
combustion experiment and numerical simulation.

5. Conclusions
In this research, the flow fields for coherent jets using various kinds of restriction structure were
investigated at room and high ambient temperatures. Both numerical simulation and combustion
experiments were undertaken, and the data generated by each of them was in good agreement.
The main conclusions can be summarized as follows:

(1) The average potential core length for the main oxygen jet formed by a coherent lance using
a restriction structure was 1.2 and 2.3 times larger than that formed by a conventional coherent
lance or a conventional lance, respectively. Therefore, a coherent lance with a restriction structure
is able to prolong the velocity potential core length for the main oxygen jet, and a flap restriction
structure further improves upon this trend.
(2) When compared to a conventional coherent lance, a coherent lance using a restriction structure is
able to delay energy transmission in a radial direction at the coherent lance tip, which enlarges the
high-temperature zone in a axial direction and increases the axial velocity of the shrouding flame.
For coherent jets using a flap restriction structure, part of its heat energy of the shrouding flame
is transformed into its kinetic energy, when the shrouding flame passes through the coherent
lance tip. As a result, the maximum temperature of the main oxygen jet using a flap restriction
structure is lower than that using an open restriction structure.
(3) Comparison to all the coherent lance structures, the flap coherent lance can increase the dynamic
pressure of the supersonic jet, which further improves the surface area between the supersonic jet
and molten bath, resulting in accelerating the dephosphorization and decarburization rates in
the steelmaking process. Besides, the ambient temperature has the largest and smallest effect on
the maximum dynamic pressures of the coherent jet using conventional coherent lance and flap
coherent lances, respectively.
(4) During the 75 t EAF steelmaking process, the average [C][O] and steelmaking time using a flap
coherent lance was reduced by 7.7% and 1.2%, respectively, and the average FeO content dropped
by 1.3 wt % in the end-point slag, whilst the average dephosphorization rate increased by 1.6%,
when compared to a conventional coherent lance. This is consistent with the results of the
combustion experiment and the numerical simulation.

Author Contributions: Conceptualization, F.L. and R.Z.; Methodology, R.Z.; software, F.L.; Validation, R.Z. and F.L.;
Formal analysis, G.W.; Investigation, F.L. and G.W.; Resources, S.F.; Data curation, S.F. and G.W.; Writing—original
draft preparation, G.W. and R.Z.; Writing—review and editing, F.L. and G.W.; Visualization, F.L.; Supervision,
R.Z.; Project administration, F.L. and R.Z.; Funding acquisition, F.L. All authors have read and agreed to the
published version of the manuscript.
Funding: This research was funded by the Fundamental Research Funds for the Central Universities, grant
number FRF-TP-17-007A1 and the National Nature Science Foundation of China, grant number NSFC 51804028.
Conflicts of Interest: The authors declare no conflict of interest.
Materials 2020, 13, 1043 14 of 15

References
1. Zhou, X.; Ersson, M.; Zhong, L.; Jönsson, P. Numerical simulations of the kinetic energy transfer in the bath
of a BOF converter. Metall. Trans. B 2016, 47, 434. [CrossRef]
2. Wei, G.; Zhu, R.; Cheng, T.; Dong, K.; Liu, R. Modelling on the penetration depth of the coherent supersonic
jet in EAF steelmaking. Ironmak. Steelmak. 2018, 45, 828–838. [CrossRef]
3. Li, Q.; Li, M.; Kuang, S.; Zou, Z. Numerical simulation of the interaction between supersonic oxygen jets and
molten slag–metal bath in steelmaking BOF process. Metall. Trans. B 2015, 46, 1494–1509. [CrossRef]
4. Lv, M.; Zhu, R.; Wang, H.; Bai, R. Simulation and application of swirl-type oxygen lance in vanadium
extraction converter. Steel Res. Int. 2013, 84, 304–312. [CrossRef]
5. Alam, M.; Naser, J.; Brooks, G.; Fontana, A. Computational fluid dynamics modeling of supersonic coherent
jets for electric arc furnace steelmaking process. Metall. Trans. B 2010, 41, 1354–1367. [CrossRef]
6. Anderson, J.E.; Mathur, P.C.; Selines, R.J. Method for Introducing Gas into a Liquid. US Patent 5814125,
29 September 1998.
7. Odenthal, H.J.; Buess, S.; Starke, P.; Nörthemann, R.; Lohmeier, M. The new generation of SIS injector for
improved EAF processes. In Proceedings of the METEC & 2nd ESTAD—European Steel Technology and
Application Days, Düsseldorf, Germany, 15–19 June 2015; p. 69.
8. Alam, M.; Naser, J.; Brooks, G.; Fontana, A. A computational fluid dynamics model of shrouded supersonic
jet impingement on a water surface. ISIJ Int. 2012, 52, 1026–1035. [CrossRef]
9. Meidani, A.R.N.; Isac, M.; Cameron, A.R.A.; Guthrie, R.I.L. Modelling shrouded supersonic jets in
metallurgical reactor vessels. ISIJ Int. 2004, 44, 1639–1645. [CrossRef]
10. Sumi, I.; Kishimoto, Y.; Kikuchi, Y.; Igarashi, H. Effect of high-temperature field on supersonic oxygen jet
behavior. ISIJ Int. 2006, 46, 1312–1317. [CrossRef]
11. Klioutchnikov, I.; Olivier, H.; Odenthal, J. Coaxial jets entering into a hot environment. In Proceedings of the
29th International Symposium on Shock Waves 2, Madison, WI, USA, 14–19 July 2013; p. 913.
12. Liu, F.; Zhu, R.; Dong, K.; Hu, S. Effect of ambient and oxygen temperature on flow field characteristics of
coherent jet. Metall. Trans. B 2016, 47, 228–243. [CrossRef]
13. Liu, F.; Zhu, R.; Dong, K.; Fan, S. Behaviors of Coherent Flow Field with Various Shrouding Nozzles
Arrangement. ISIJ Int. 2018, 58, 496–504. [CrossRef]
14. Liu, F.; Sun, D.; Zhu, R.; Su, R.; Wang, X. Effect of shrouding gas temperature on characteristics of a supersonic
jet flow field with a shrouding Laval nozzle structure. Metall. Trans. B 2018, 49, 2050–2062. [CrossRef]
15. Odenthal, H.J.; Bui, P.; Reifferscheid, M.; Hovestädt, E.; Nies, J.; Klioutchnikov, I.; Olivier, H. The optimized
SIS injector for EAF applications. In Proceedings of the AISTech 2014—The Iron & Steel Technology
Conference and Exposition, Indianapolis, IN, USA, 5–8 May 2014; p. 1.
16. Foelsch, K. The analytical design of an axially symmetric Laval nozzle for a parallel and uniform jet.
J. Aeronaut. Sci. 2012, 16, 161. [CrossRef]
17. Hu, S.; Zhu, R.; Liu, R.; Dong, K.; Liu, F.; Ma, G.; Su, R. Numerical simulation and experimental measurement
of transport phenomena for coherent jet with CH4 + N2 mixed fuel gas. J. Iron Steel Res. Int. 2018, 25, 28–36.
[CrossRef]
18. Sivells, J.C. Aerodynamic design of axisymmetric hypersonic wind-tunnel nozzles. J. Spacecr. Rocket. 1970, 7,
1292–1299. [CrossRef]
19. Launder, B.E.; Spalding, D.B. Lectures in Mathematical Model of Turbulence; Academic Press: London, UK,
1972; p. 124.
20. Anderson, J.D. Introduction to Flight; McGraw-Hill Education (Asia): Tuas, Singapore, 2013; p. 125.
21. Papamoschou, D.; Roshko, A. The compressible turbulent shear layer: An experimental study. J. Fluid Mech.
1988, 197, 453–477. [CrossRef]
Materials 2020, 13, 1043 15 of 15

22. Mei, Z.; Wang, F.; Li, P.; Mi, J. Diffusion flame of a CH4 /H2 jet in hot low-oxygen coflow. Int. J. Hydrogen
Energy 2011, 36, 9267–9277.
23. Wei, G.; Zhu, R.; Cheng, T.; Dong, K.; Yang, L.; Wu, X. Study on the impact characteristics of coherent
supersonic jet and conventional supersonic jet in EAF steelmaking process. Metall. Trans. B 2018, 49, 361–374.
[CrossRef]
24. Molloy, N.A. Impinging jet flow in a 2-phase system-basic flow pattern. J. Iron Steel Inst. 1970, 226, 943.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).