High Temp. Mater. Proc.
2019; 38: 1–7
Feng-guang Li* and Jian-liang Zhang
Research on the Influence of Furnace Structure
on Copper Cooling Stave Life
https://doi.org/10.1515/htmp-2017-0040
Received March 22, 2017; accepted July 23, 2017
Abstract: In this paper, a blast furnace gas flow distribu-
tion model with variable furnace structure was founded
based on CFD (computational fluid dynamics) theory,
and the gas velocity distribution near the surface of the
copper staves in different areas of the BF is calculated
under different conditions of variational structure para-
meters like Bosh angle, shaft angle, and the newly pro-
posed “equivalent Bosh angle.” Based on the calculation,
the influence rule of the BF structure on the service life of
copper stave and the corresponding operation measures
were obtained. The result shows that the increase of the
Bosh angle and the decrease of the shaft angle will incur
increasing of the gas flow velocity near the surface of the
copper staves, which is harmful to the cooling stave life;
the variation of the equivalent Bosh angle has a most
significant influence on the cooling stave life, and the
increase of the equivalent Bosh angle will cause a sharp
increase of the gas flow velocity, which will damage the
copper staves seriously; adopting long tuyeres and min-
ishing the equivalent Bosh angle will reduce the washing
action of the gas flow and ensure the stability of slag
hanging to achieve a long service life of copper staves.
Keywords: blast furnace, furnace structure, copper
cooling stave, gas flow distribution, numerical
simulation
The blast furnace structure has an important influence on
burden movement, flow of coal gas, and utilization of
chemical and thermal energy in blast furnace production.
Besides, rational furnace structure could also avoid the
abnormal erosion of refractory and prolong the service
time of blast furnace [1, 2]. Domestic furnace structure
has the following development trend characters [3]:
*Corresponding author: Feng-guang Li, School of Materials Science
and Engineering, Hubei University of Automotive Technology,
Shiyan, Hubei 442002, China; School of Metallurgical and
Ecological Engineering, University of Science and Technology
Beijing, Beijing 100083, China, E-mail: lemon2006@126.com
Jian-liang Zhang, School of Metallurgical and Ecological
Engineering, University of Science and Technology Beijing, Beijing
100083, China
Open Access. © 2019 Walter de Gruyter GmbH, Berlin/Boston.
This work is licensed under the Creative Commons Attribution 4.0 Public License.
(1) Value of height/diameter diminishes, which means
furnace structure becomes from “high and thin” to
“low and fat”;
(2) Shaft angle β and Bosh angle α both diminish and
become close to each other. For modern large-scale
blast furnaces, the shaft angle normally ranges from
79 ° to 81 °, while for Bosh angle, the range is
between74 and 80 °.
Before application of copper cooling stave, the brick
in Bosh, belly, and shaft areas used to be thick, and after
blow-in of the furnace the lining of the furnace would be
eroded into different shapes according to various furnace
operations; thus, the operating furnace profile was
formed. Rational operating furnace profile could main-
tain good production indexes and prolong service life of
the furnace [4–7]. Due to this reason, furnace operation
could make up improper designing of the traditional
thick bricking furnace. However, after application of
cooper cooling stave, the structure of “thin bricking fur-
nace” was formed for which only 120–150 mm insert
bricks existed at hot surface of cooling stave, so the
influence of insert brick erosion on final furnace profile
was limited and the designing furnace profile was also
the final furnace profile [8]. It could be summarized that
after application of cooper cooling stave, the requirement
for furnace design becomes stricter. But in fact after
application of cooper cooling stave, no practical improve-
ment was presented in furnace designing area.
After investigation of cooper cooling stave damage
cases in recent years, many scholars considered irra-
tional furnace structure as an important reason.
Scholars holding this standpoint thought that for most
of blast furnace they all had overlarge Bosh angle and
undersize shaft angle, which would lead to fierce brush
of coal gas on hot face of cooling stave and cause fast
damage [9].
In order to investigate the influence of furnace
structure on lifetime of cooling stave, commercial soft-
ware Fluent was used in this paper to establish a coal
gas distribution model in blast furnace. Coal gas dis-
tribution near the surface of cooling stave under con-
ditions of various Bosh angles and shaft angles was
calculated to ascertain the influence of furnace
Unauthenticated
Download Date | 6/19/19 7:39 PM
2 F.-G. Li and J.-L. Zhang: Research on the Influence of Furnace Structure

structure on coal gas distribution. For coal gas distri- (4) Different layers of coke, ore, and cohesive zone are
bution in blast furnace, many scholars have done distinguished according to different sets of
abundant work and good results were obtained. Based porosities.
on transfer phenomenon among phases of gas, solid, (5) Coal gas is treated as incompressible Newtonian
and fluid, Hatano and Kurita established a coal gas fluid.
distribution model and also took chemical reaction
and heat transfer into account [10]. Austin et al. estab- Based on these assumptions, mathematical formulas
lished a model considering four phases of gas, solid, for the established model through software “Fluent” are
fluid, and particle, which could accurately calculate as follows.
the gas and temperature distribution [11, 12], and
based on this some scholars further developed five-
phase model considering coal gas, solid burden, dust,
Gas distribution model
slag, and hot metal [13]. Through finite-difference
method, Zhang et al. [14] established a coal gas distri-
Gas distribution model consists of equation of continuity,
bution model in which parameters of burden diameter,
momentum equation, turbulence model equation, turbu-
resistance coefficient, stack angle, and burden thick-
lence energy dissipation equation, and Ergun equation.
ness could be considered and calculation was more
Under two-dimensional rectangular coordinate system,
suitable to practical furnace.
the above equations are presented as follows:
However, aforementioned models are more focused
(1) Equation of continuity
on transport phenomena of heat and mass between dif-
ferent phases of solid burden and gas, and the influence ∂ρui
=0 (1)
of furnace structure is less considered. In this study, the ∂xi
influence of furnace structure on coal gas distribution is
In the formula, ρ is gas density, kg·m−3; ui is gas flow
the main focus, so the interaction among different phases
speed, m·s−1.
in the furnace is neglected, and shaft angle and Bosh
(2) Momentum equation
angle are chosen as the main influence factors. Besides,
  
“equivalent Bosh angle” concept is proposed, and a thor- ∂ðρui uj Þ ∂P ∂ ∂ui ∂uj
=− + μ + (2)
ough analysis of the influence of different factors on coal ∂xj ∂xi ∂xj eff ∂xj ∂xi
gas flow rate near cooling stave surface is finished and
the influence of furnace structure on lifetime of cooling In the formula, ui and uj are speed of gas along directions
stave is summarized. of i and j, respectively, m·s−1; xi and xj are coordinate
figure of directions i and j; ρ is density, kg·m−3; P is
pressure, Pa; μeff is effective coefficient of viscosity,
Pa·s, which is ascertained by turbulence model.
Foundation of coal gas distribution (3) k–ε turbulence model
 
model in the blast furnace ∂ μeff ∂k
ρui k −  = Gk − ρε (3)
∂xi σ k ∂xi

Assumption and simplification of the model In the formula: k is turbulent kinetic energy, m2·s−2; Gk is
item derived from turbulent energy; ε is dissipation ratio
Considering the axial symmetrical property of blast fur- of turbulent energy, m2·s−3.
nace, the following assumptions are made: (4) Turbulent energy dissipation equation
(1) The model is two-dimensional steady-state model,  
∂ μeff ∂ε
and blast furnace is considered as axial symmetrical ρui ε −  = ðC1 εGk − C2 ρε2 Þ=k (4)
∂xi σk ∂xi
structure, and half of longitudinal section is selected
as the computational area; In the formula：
(2) Particle size change of ore and coke are considered,  
and lowering the rate of solid burden is fixed as ∂uj ∂ui ∂uj
Gk = μt  + (5)
constant; ∂xi ∂xj ∂xi
(3) Position and shape of cohesive zone are considered,
and heat transfer is neglected; μeff = μl + μt (6)

Unauthenticated
Download Date | 6/19/19 7:39 PM
 F.-G. Li and J.-L. Zhang: Research on the Influence of Furnace Structure 3

k2
μt = ρCμ (7)
ε

It the formula: μt is turbulence viscosity coefficient, Pa·s;

μl is laminar flow viscosity coefficient, Pa·s; C1, C2, Cμ, σk,
σt are empirical constants, which are ascertained through
recommended values of Launder and Spalding [15], as
shown in Table 1.
(5) Ergun equation

ΔP ð1 − εÞ2 μ 1−ε ρ
= 150 μ + 1.75 3 μ 2 (8)
H ε3 ðϕdp Þ2 A ε ϕdp A

Table 1: Empirical constant of the kinematics dissipate equation.

C C Cμ σk σk
1-burden surface; 2-ore; 3-coke; 4-softing and cohesive zone;
. . . . .
5-coke window; 6-dead man; 7-tuyere raceway

Figure 1: Gas flow distribution model of the blast furnace.

In the formula: μA is gas flow rate through packed bed 1, burden surface; 2, ore; 3, coke; 4, softing and cohesive zone; 5,
calculated according to cross-sectional area, m·s−1; dp is coke window; 6, dead man; 7, tuyere raceway.

particle diameter, m; μ is fluid viscosity, Pa·s; ϕ is shape

factor of particles.
For Eb:

Physical model of gas distribution 1 Vb 4Vb

Based on the above formulas, according to structure
parameters of a domestic 1,780 m3 blast furnace (shown
in Table 2), gas distribution model for practical blast
furnace is established, as shown in Figure 1.
Influence of raceway and deadman on gas distribu-
tion is considered, and the size of them is ascertained by
the following formula:
LR
LR = 0.118 × 10 − 3 Eb + 0.77, = KR
HR
Table 2: Structure parameters of the model.
Eb = ρ b ½ ×
Tb P0 2
 (10)
2 gn πnðdb Þ2 273Pb
In the formula, LR is the depth of raceway, m; HR is the
height of raceway, m; KR is the raceway shape coefficient
and is dimensionless, which lies between 0.6 and 1.17
according to different furnace volumes; Eb is the blast
kinetic energy, kg·m·s−1; Vb is the blast volume, m3·s−1;
N is the tuyere number; G is the acceleration of gravity,
m2·s−1; db is the diameter of tuyere, m; Tb is the blast
(9) temperature, K; Pb is the blast pressure, Pa; P0 is the
atmosphere pressure, Pa.

Items Values
Boundary conditions
BF hearth diameter d/mm ,
BF hearth height h/mm , (1) Wall surface: boundary condition for wall touching
Distance from tuyere level to tap hole centerline hf/mm , gas part is set as non-slipping;
BF Bosh height h/mm ,
(2) Inlet: inlet of the model is tuyere and is set as
BF belly height h/mm ,
BF Bosh diameter D/mm ,
velocity boundary condition, which is calculated
BF shaft height h/mm , from conversion of blast volume;
BF throat diameter d/mm , (3) Outlet: outlet of the model is the up edge of furnace
BF throat height h/mm , throat and is set as the pressure boundary condition,
BF Tuyere Diameterr/mm  which is equal to top pressure;

Unauthenticated
Download Date | 6/19/19 7:39 PM
4 F.-G. Li and J.-L. Zhang: Research on the Influence of Furnace Structure

(4) Surface of liquid metal and slag: the bottom of the considering one factor, the values of other variables are
model is slag–metal surface and is set as constant, set as their characteristic value.
∂u
∂y = 0;
(5) Axial symmetry boundary: along the axis of symme-
try the radial velocity is 0, ∂u
∂x = 0. Results and discussion

Calculation parameters Method for analysis

and modeling scheme
Parameters of the practical blast furnace are used for
calculation, which are shown in Table 3.
Table 3: Parameters for calculation.
As shown in Figure 2, in order to analyze the gas dis-
tribution under different conditions, along 10 mm near
the wall surface from bottom of Bosh to the lower area
of shaft, some points are chosen and the gas velocity of
these points is used as the surface velocity of gas. The
chosen points are scattered along the red line in Figure 2.

Items Values

Ore porosity/dimensionless value .

Coke porosity/dimensionless value .
Deadman porosity/dimensionless value .
Blast velocity/m·s−  30
Top pressure/kPa 
Coke gas density/kg·m− .
Weight of ore batches/kg ,
Weight of coke batches/kg ,
Coke layer thickness/mm . 20
Ore layer thickness/mm .

Influence of furnace structure on gas distribution is the 10

focus of the study, especially for gas flow in edge area. So
Bosh angle and shaft angle are chosen as variable para-
meters. For practical blast furnace in which furnace struc-
ture is unchangeable, the measure of prolonging tuyere
could be adopted, for which the included angle between
1
horizontal line and line between tuyere front and belly
bottom wall side could be changed. In this study, the
angle is defined as “equivalent Bosh angle” and is also Figure 2: Schematic diagram of the sampling points of the gas flow
set as a variable parameter. velocity.
According to the design parameters and practical
data, variation standard of the variables is ascertained
and shown in Table 4. For specific calculation Influence of Bosh angle on lifetime
of cooling stave

For Bosh angle changing from 74 ° to 79 ° (shaft angle

Table 4: Influence factors and their value level.
78 °, equivalent Bosh angle 76 °), distribution of surface
gas velocity is shown in Figure 3. The 1st to 10th points
Factors Bosh Shaft Equivalent Bosh
angle/° angle/° angle/° are located at Bosh area, the 11th to 20th points are
located at belly area, and 21st to 30th points are located
Variation range  ~   ~   ~  at lower area of shaft. It could be found that from Bosh
Characteristic   
area to the bottom of cohesive zone, gas flow rate
value
decreases quickly. At cohesive zone, due to rectification

Unauthenticated
Download Date | 6/19/19 7:39 PM
 F.-G. Li and J.-L. Zhang: Research on the Influence of Furnace Structure 5

different Bosh angles, gas flow rates are different, and

gas velocities on the surface of copper staves / m s−1

8
smaller the Bosh angle, the lower the gas flow rate
bosh angle
o (Figure4(a)). At the bottom area of Bosh, the influence
74
75
o of Bosh angle on the surface flow rate is limited, and with
6 o
distance away from the Bosh bottom, the influence of the
76
o
77 Bosh angle increases, when further to the bottom of
o
78 cohesive zone, the difference of gas flow rate diminishes
4 o
79
gradually. Take No. 8 point at Bosh area for example,
when Bosh angle changes from 74 ° to 79 °, the gas flow
rate increases by 78 % from 0.97 m·s−1 to 1.73 m·s−1.
2

The above analysis shows that Bosh angle has a

0 significant influence on the wall surface gas flow rate,
and too big Bosh angle could lead to surface flow rate
0 5 10 15 20 25 30
become over high and make brushing effect strength-
No. of sampling points (distance to the lower end of bosh)
ened, which would make adhering slag layer on copper
Figure 3: The velocities of the gas flow on the surface of the cooling cooling stave become unstable and even lead to damage
stave under different conditions of variational Bosh angles. of cooling stave.

effect of cohesive zone, most of gas goes through coke

lawyer and near-wall surface gas flow has a low rate. Influence of shaft angle on lifetime
After cohesive zone, coal gas is redistributed and gas of cooling stave
flow rate near-wall surface increases gradually. At belly
area due to constant of diameter, change of gas flow rate Changes of gas distribution for shaft angle changing from
is limited. At shaft area due to diminishing of diameter, 76 ° to 80 ° are shown in Figure 5. It could be concluded
gas flow rate increases gradually. At the same time, from that, for Bosh and belly area, changes of shaft angle has
Figure 3, it could be summarized that changing of Bosh limited influence on gas flow, but for shaft area, the
angle has a significant influence on the gas flow rate for influence is significant.
region lower than cohesive zone at Bosh area, and for In order to investigate influence of shaft angle on
region higher than cohesive zone the influence is limited. gas flow, from the bottom to the middle of the shaft,
Before rectification of cohesive zone, with increase of 30 points were selected and the flow rates on these
Bosh angle gas flow rate, near-wall surface increases points are shown in Figure 6. It could be found that
gradually. the smaller the shaft angle, the higher the surface flow
In order to accurately analyze flow rate change at rate for lower area of shaft. With distance away from
Bosh region, 30 equidistance points alone Bosh wall sur- the bottom of shaft, influence of shaft angle on gas
face were selected. Surface flow rate under various con- flow rate becomes more and more obvious. Take
ditions are shown in Figure 4. At Bosh region, for the 30th point, for example, when shaft angle is
s−1
s−1

4
gas velocities on the surface of copper staves / m
gas velocities on the surface of copper staves / m

4
bosh angle
o
74
3 o 3 point 1
75
o point 8
76 point 15
o
77 point 23
2 o 2
78 point 30
o
79

1 1

0 Figure 4: The gas flow veloci-

0
ties at the Bosh area before
0 5 10 15 20 25 30 74 75 76 77 78 79 “rectifying effect” under the
No. of sampling points (distance to the lower end of bosh) bosh angle / o conditions of different Bosh
(a) (b) angles.

Unauthenticated
Download Date | 6/19/19 7:39 PM
6 F.-G. Li and J.-L. Zhang: Research on the Influence of Furnace Structure

s−1

s−1
12

gas velocities on the surface of copper staves / m

gas velocities on the surface of copper staves / m

20 equivalent bosh angle

10 o
74
shaft angle o
o 75
76 o
15 o 8 76
77 o
o 77
78 o
o 6 78
79 o
10 o 79
80
4

5
2

0 0

0 5 10 15 20 25 30 0 5 10 15 20 25 30

No. of sampling points (distance to the lower end of bosh) No. of sampling points (distance to the lower end of bosh)

Figure 5: The velocities of the gas flow on the surface of the cooling Figure 7: The velocities of the gas flow on the surface of the cooling
stave under different conditions of variational shaft angles. stave under different conditions of variational equivalent Bosh
angle.

changed from 76 ° to 80 °, the flow rate at this point is

changed from 19.42 m·s−1 to 15.27 m·s−1, lowered by region, 30 points were selected and the surface flow rate
21 %. So too small shaft angle could lead to high sur- of these points is shown in Figure 8.
face flow rate at shaft region and cause brushing effect From Figure 8, it could be concluded that the bigger is
of gas, which is bad for longevity of cooling stave. equivalent Bosh angle is, the higher surface flow rate at
belly region is, which is bad for longevity of cooling stave.
At the same time, the closer to bottom of Bosh, the greater
the influence is. Take the break point of hearth and Bosh
Influence of equivalent shaft angle
for example (Point 1), when equivalent Bosh angle is
on lifetime of cooling stave changed from 74 ° to 79 °, surface flow rate at this point
is increased by 5 times from 1.74 m·s−1 to 10.73 m·s−1 . So it
For Bosh angle 76 ° and shaft angle 78 °, when equivalent
could be concluded that it equivalent Bosh angle is too
Bosh angle is increased from 74 ° to 78 ° change of sur-
big, the brushing effect at Bosh region is significant, and
face gas low rate for regions of Bosh, belly, and lower
this is in accordance with the domestic practical situation
area of shaft are shown in Figure 7. It could be concluded
of cooper cooing stave damage. From this aspect, through
that equivalent Bosh angle has a significant influence on
method of lengthening tuyere to lower equivalent Bosh
gas flow at Bosh region, but for region above cohesive
angle could obviously decrease brushing effect at Bosh
zone the influence is limited. In order to investigate the
region and stabilize adhering slag layer, which is good
influence of equivalent shaft angle on gas flow in Bosh
for longevity of cooper cooling stave.
s−1
s−1

20
gas velocities on the surface of copper staves / m

20
gas velocities on the surface of copper staves / m

shaft angle
o
76
o 15 point 1
15 77
o point 8
78 point 15
o
79 point 23
10 o 10 point 30
80

5
5

0
0
Figure 6: The gas flow velocities
0 5 10 15 20 25 30 76 77 78 79 80 at the shaft area after “rectify-
o
shaft angle /
No. of sampling points (distance to the lower end of bosh) ing effect” under the conditions
(a) (b) of different shaft angles.

Unauthenticated
Download Date | 6/19/19 7:39 PM
 F.-G. Li and J.-L. Zhang: Research on the Influence of Furnace Structure 7

s−1

s−1
12 12
gas velocities on the surface of copper staves / m

gas velocities on the surface of copper staves / m

equivalent bosh angle point 1
10 o 10 point 8
74
o point 15
75
o point 23
8 76 8
o
point 30
77
o
6 78 6
o
79
4 4

2 2

Figure 8: The gas flow velocities

0 0
at the Bosh area before “recti-
0 5 10 15 20 25 30 74 75 76 77 78 79 fying effect” under the condi-
o
No. of sampling points (distance to the lower end of bosh) equivalent bosh angle /
tions of different equivalent
(a) (b) Bosh angels.

Conclusions permeation suppression (2016CFB293); (3) The Doctoral

Research Start-up Foundation of Hubei University of
(1) Bosh angle has an obvious influence on the surface Automotive Technology (BK201607).
gas flow rate at Bosh region, and a large Bosh angle
could give a surface flow rate at Bosh region and
strengthen brushing effect, which is bad for stability References
of adhering slag layer and could cause damage of
[1] T. Inada, K. Takatani, K. Takata, et al., ISIJ Int., 43(8) (2003)
cooling stave. 1143–1150.
(2) Lowering of shaft angle could lead to increase of [2] M. Ichida, K. Nishihara, K. Tamura, et al., ISIJ Int., 31(5) (1991)
surface flow rate at the shaft region and strengthen 515–523.
brushing effect, which is bad for longevity of cooper [3] Z.Y. Xiang and X.L. Wang, Design of the Blast Furnace: Theory
cooling stave. of Ironmaking Process Design and Practice, 2nd ed.,
Metallurgical Industry Press, Beijing (2014).
(3) Influence of equivalent Bosh angle on lifetime of
[4] X.L. Wang, Q and A of BF Ironmaking, 3rd ed., Metallurgical
cooper cooling stave is the most obvious and Industry Press, Beijing (2013).
increase of equivalent Bosh angle could lead to a [5] Y.J. Cao, J.L. Zhang, H.W. Guo, et al., J. Iron Steel Res., 27(1)
sharp increase of the surface flow rate. If equivalent (2015) 7–11.
Bosh angle is too big, the brushing effect of coal gas [6] C.C. Lin and Z.Y. Xiang, Baogang Technol., 27(2) (2009) 49–53.
[7] Z.R. Yang, Iron. Steel, 50(1) (2015) 31–36.
on cooling stave (especially for lower area) would
[8] W.X. Wang, Ironmaking, 32(4) (2003) 58.
become serious.
[9] F.G. Li, J.L. Zhang, L. Wei, et al., Application status and damage
(4) Lengthening tuyere to lower equivalent Bosh angle reason analysis of copper stave//2012 National Conference on
could obviously decrease the brushing effect of coal BF Long-campaign and High Blast Temperature Technology.
gas at Bosh region and stabilize adhering slag law- Beijing (2012), p. 46.
yer which is good for longevity of cooper cooling [10] J. Yagi, H. Nogami, P.R. Austin, et al., Development of mathe-
matical model and application for super-high efficiency opera-
stave.
tions of blast furnace//Proceedings ICSTI’06, Osaka (2006).
[11] P.R. Austin, H. Nogami and J. Yagi, ISIJ Int., 37(5) (1997)
458–467.
Funding: This paper had been aided by the following [12] P.R. Austin, H. Nogami and J. Yagi, ISIJ Int., 37(8) (1997)
foundations: (1) Natural Science Foundation of China: 748–755.
Study on induced cracking mechanism of BF copper cool- [13] J.A. Castro, H. Nogami and J. Yagi, ISIJ Int., 40(7) (2000)
639–646.
ing stave and constructing technology of hydrogen per-
[14] X.S. Zhang, G.L. Qing, L. Ma, et al., Simulate gas and pressure
meation suppression coating (51604103); (2) Natural flow of BF in mathematic model//2010 National Academic
Science Foundation of Hubei Province: Research on the Conference on ironmaking technology. Beijing (2010), p. 916.
cracking mechanism of BF cooling stave caused by [15] J.F. Zhu, S.S. Cheng, H.B. Zhao, et al., J. Univ. Sci. Technol.
hydrogen and the key technology of hydrogen Beijing, 31(2) (2009) 224–228.

Unauthenticated
Download Date | 6/19/19 7:39 PM