See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/306019119

Modeling of Steel Slab Reheating Process in a Walking Beam Reheating Furnace

Conference Paper · July 2016

DOI: 10.1115/HT2016-7282

CITATIONS READS
2 1,326

8 authors, including:

Guangwu Tang Tyamo Okosun

Purdue University Purdue University Northwest
58 PUBLICATIONS   72 CITATIONS    21 PUBLICATIONS   25 CITATIONS   

SEE PROFILE SEE PROFILE

Chenn Zhou Dengqi Bai

Purdue University Ssab Ab
243 PUBLICATIONS   788 CITATIONS    6 PUBLICATIONS   14 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Electric arc furnace modeling View project

plate rolling View project

All content following this page was uploaded by Guangwu Tang on 10 August 2016.

The user has requested enhancement of the downloaded file.

 Proceedings of the ASME 2016 Heat Transfer Summer Conference
HT2016
July 10-14, 2016, Washington, DC, USA

HT2016-7282

MODELING OF STEEL SLAB REHEATING PROCESS IN A WALKING BEAM REHEATING FURNACE

Guangwu Tang, Arturo Saavedra, Tyamo Okosun, Bin Wu, Chenn Q. Zhou

Center for Innovation through Visualization and Simulation (CIVS), Purdue University Calumet
2200 169th Street, Hammond, IN 46323
Phone: 219-989-2665
Email: czhou@purduecal.edu

Dengqi Bai, Yufeng Wang, Rick Bodnar

SSAB Americas R&D, 1755 Bill Sharp Blvd, Muscatine, IA 52761

ABSTRACT
Slab reheating is a very important step in steel product
manufacturing. A small improvement in reheating efficiency
can translate into big savings to steel mills in terms of fuel
consumption and productivity. Computational fluid dynamics
(CFD) has been employed in conducting numerical simulations
of the slab reheating furnace operation. However, a full
industrial scale three-dimensional (3D) simulation of a slab
reheating furnace, while comprehensive, is not an efficient way
to conduct broad studies of the slab heating process. In this
paper, a comprehensive two-dimensional (2D) numerical heat
transfer model for slab reheating in a walking beam furnace
was developed using the finite difference method. The 2D
heat transfer model utilizes the heat transfer coefficients
derived from a 3D reheating furnace CFD model which was
validated by using mill instrumented slab trials. The 2D heat
transfer model is capable of predicting slab temperature
evolutions during the reheating processes based on the real time
furnace conditions and steel physical properties. The 2D
model was validated by using mill instrumented slab trials and
production data. Good agreement between the model
predictions and production data was obtained.
Key words: Reheating Furnace, CFD, Zone Method, Heat
Transfer, Transient Heating Characteristics
INTRODUCTION
Slab reheating is an important step in the production of
high quality plate products. A good reheating practice ensures
not only high furnace throughput but also a uniform
temperature distribution within reheated slabs. While an over
heated slab could have surface defects and coarse initial as-
heated austenite grain sizes, an under heated slab can cause
high mill load and shape problems during the subsequent
rolling [1-2]. Therefore, achieving an aim bulk temperature
with good uniformity both along the slab length and through
the slab thickness is critical for reheating furnace operation
control.
The reheating process is a complex physical and chemical
process which involves combustion, heat exchange among
furnace walls, flames, skids and steel slabs [3]. A well-
controlled reheating furnace should have high thermal
efficiency, and produce low emissions and a well heated slab
[4-5]. Much research has been done on numerical modeling
of reheating furnaces. The scope of simulating a reheating
furnace can vary widely, from furnace combustion, heat
transfer, and skid mark effect to slab residence time
optimization, environmental effects [6-7], and slab surface
scale effects [8-9]. Although CFD modeling has been proven
to be an effective tool to simulate the reheating process [10-13],
the computation time is usually long and requires high
computing power. Therefore, simplified methods such as
dividing a reheating furnace into several zones and assuming a
constant temperature in each zone are usually used [14-16]. A
simplified slab moment in a reheating furnace is also used for
modeling [3, 17, 18]. However, in a real reheating process,
the dynamic phenomena such as fuel input variation, changing
of slab travel speed, and different steel grades all can affect
heat transfer characteristics. In order to obtain the details in
simulation, a multiscale grid structure has to be used which will

1 Copyright © 2016 by ASME

increase computation time and reduce the computing
efficiency.
In order to create a model which requires less computing power
and can be easily used by a steel mill, a two-dimensional (2D)
heat transfer model has been widely used to simulate slab
temperature evolution in a reheating process. Chen et al. [19]
developed a 2D heat transfer model for slab reheating based on
the finite difference method. Jang and Kim [20] developed a
2D heat transfer model based on the Finite Element Method
(FEM) for billet reheating. In the present paper, a
comprehensive 2D heat transfer model for slab reheating has
been developed. The model was validated by using mill
instrumented slab trials and production data. Compared with
the 3D CFD simulation of a whole reheating furnace, the 2D
heat transfer model used for predicting slab temperature
evolution is much quicker. A reheating process with multiple
operational parameters for a two hour furnace residence time
can be calculated within a few seconds. This is a much more
convenient tool which can be used to perform troubleshooting
of operational issues or to optimize the reheating practices.
MODEL DEVELOPMENT
Governing Equations
When a slab travels through a reheating furnace, the
transient heat conduction process can be expressed by the
following equation:
𝜕𝑇(𝑥,𝑦,𝑡) 𝜕 𝜕𝑇(𝑥,𝑦,𝑡)
𝐶𝑠 (𝑡)𝜌𝑠 = (𝐾𝑠 (𝑡) )+
𝜕𝑡 𝜕𝑥 𝜕𝑥
(1)
where: 𝑇(𝑥, 𝑦, 𝑡) = temperature at time step 𝑡, 𝑥 = slab
longitudinal direction coordinate, 𝑦 = thickness direction
coordinate, 𝐶𝑠 (𝑡) = specific heat of the slab, 𝜌𝑠 = density of
the slab, 𝐾𝑠 (𝑡) = thermal conductivity of the slab.
The temperature boundary conditions were obtained from
28 thermocouple installed in the reheating furnace. The trainset
measurements were taken into account with the moving slab
positions.
Discretization
A two dimensional heat transfer model has been developed
based on finite difference method. In order to solve the
unsteady non-linear heat transfer problem, the explicit method
was applied on the nodes by dividing a slab slice into small
sections as shown in Figure 1. The discretization on the
internal nodes and boundary nodes can be expressed as
Equations (2) and (3).
Internal node:
𝑡+∆𝑡 𝑡 𝑡
𝑇𝑖,𝑗 −𝑇𝑖,𝑗
𝜌𝑠 𝐶𝑠 (𝑡)∆𝑥∆𝑦 = − 𝐾𝑠 (𝑡)∆𝑥 [
∆𝑡
𝑡 𝑡 𝑡
2𝑇𝑖,𝑗 −𝑇𝑖+1,𝑗 −𝑇𝑖−1,𝑗
𝐾𝑠 (𝑡)∆𝑦 [ ] (2)
∆𝑥
where: ∆𝑥, ∆𝑦, ∆𝑡 = increment of 𝑥 coordinate, 𝑦
𝑡
coordinate and time, 𝑇𝑖,𝑗 = node temperature at time step 𝑡,
𝑡+∆𝑡
𝑇𝑖,𝑗 = node temperature at time step 𝑡 + 1.
Boundary node:
𝜕𝑇(𝑥,𝐺, 𝑡)
𝐾𝑠 (𝑡) = 𝑓𝜀𝑠 𝜎[𝑇𝑖𝑡𝑜𝑝 4 − 𝑇(𝑥, 𝐺, 𝑡)4 ] + ℎ𝑖 (𝑇𝑖𝑡𝑜𝑝 −
𝜕𝑦
𝑇(𝑥, 𝐺, 𝑡)) (3)
where: 𝐺 is the thickness of the slab, 𝑇(𝑥, 𝐺, 𝑡) is top
surface (y = G) temperature at time step 𝑡, 𝑓 is view factor, 𝜎
is Stefan-Boltzmann constant, 𝜀𝑠 is emissivity of the slab,
𝑇𝑖𝑡𝑜𝑝 is top chamber furnace temperature ( 𝑖 =1, 2, 3,
representing preheating zone, heating zone and soaking zone,
respectively), ℎ𝑖 is the convective heat transfer coefficients
which is calculated using the results obtained from the
validated CFD simulations reported elsewhere [21].
Simulation Domain
During reheating, the material flow inside the reheating
furnace is shown in Figure 1(a). Slabs travel through three
different zones, preheating zone, heating zone, and soaking
zone, from the furnace charging side to the discharge side.
Slabs are 6” thick and in various lengths. They are charged
into the reheating furnace in an orientation with the width
direction parallel to the charging direction. The computational
geometry used in the 2D model is a slice of the steel slab along
its length as shown in Figure 1(b). The 2D slice is divided
into 1” x 2” rectangles which are also called grids. The nodes
on the divided grids are the points containing temperature
values through the numerical iterations. Boundary conditions
𝜕 𝜕𝑇(𝑥,𝑦,𝑡)
are applied to the boundary nodes such as surface nodes, edge
(𝐾𝑠 (𝑡) ) nodes and corner nodes shown in Figure 1(b). The skid nodes
𝜕𝑦 𝜕𝑦
represent the skid structure which used for supporting the slab
and transporting slab throughout the furnace. Conduction heat
transfer is occur between skids and slabs. And the skids has
effect on the nearby contact region when introducing radiation
heat transfer between combustion flames and slabs, therefore,
the shadow nodes were introduce to taken account this effects.
In this model, special treatments on the skid nodes and the
shadow nodes near the skids are applied by using correction
factors to take into account of the effect of skid temperature
boundary condition on the heat flux in the skid contact area and
its geometric effect on the radiation heat flux near the skid
nodes. During a reheating process, the computational
geometry (the slab longitudinal cross section) moves through
the reheating furnace. As an example to show the effect of the
skids on the slab temperature distribution, the model calculated
temperature profiles along the slab length at different stages
during a two hour reheating are shown in Figure 2.
𝑡 𝑡
2𝑇𝑖,𝑗 −𝑇𝑖,𝑗+1 −𝑇𝑖,𝑗−1
]−
∆𝑦
2 Copyright © 2016 by ASME

(a) Reheating furnace
(b) Slab cross section
Figure 1 Computational domain
Figure 2 Temperature profiles along the slab length
It can be seen from the temperature profiles presented in
Figure 2 that at the early stages of reheating, skids have a
noticeable effect on the slab temperature. The slab
temperatures at the locations where the slab is in contact with
the skids are lower than the temperatures at the locations where
the slab is in between the skids. This difference becomes
smaller when the slab progressively moves into the soaking
zone. Eventually, the temperatures throughout the slab body
are equalized in the soaking zone.
The flow chart of the slab reheating calculation is shown in
Figure 3. The furnace and slab conditions are the inputs to the
2D heat transfer solver. The furnace temperatures in different
zones can be user-defined values or from the furnace
thermocouple readings. The specific user input interface was
developed to provide users with different options to determine
the furnace conditions as the heat transfer boundary conditions
for calculations. Slab geometry, slab charging temperature,
and steel grades with specific thermal properties, such as
3 Copyright © 2016 by ASME
thermal conductivity and specific heat, can be determined based
on real time production situations. In addition, the model is
also capable of handling non-constant walking speeds of a slab
during reheating. With all these furnace and slab inputs, the
2D heat transfer model will calculate the slab temperature
evolution and slab temperature profile at different locations,
such as slab temperature through the slab thickness and along
the slab length.
Figure 3 Flow chart of the slab reheating simulation
MODEL VALIDATION
Instrumented Slab Trial
In order to validate the numerical model, field
instrumented slab trials were conducted. A full-size slab
about 96” wide and 340” long was prepared for the trials. A
total of six thermocouples in two groups were placed in the
slab at two different locations to cover the areas of over-skid
(thermocouples #1, #2, and #3), and between-skid
(thermocouples #4, #5, and #6) as shown in Figure 4. For
each group, one thermocouple was placed near the slab top
surface, one at the slab mid-thickness, and one near the slab
bottom surface to capture the temperatures through the slab
thickness, as shown in Figure 5(a). All the thermocouples
were inserted about 15” in from the slab narrow face to make
sure that the measurements represent the slab inner
temperatures. Positioning the instrumented slab in the roll
line and alignment with the furnace skid are shown in Figure
5(b). The yellow stripe on the slab top surface marks the
location of the thermocouples which were aligned with one of
the skids in the furnace. The thermocouples were connected
to a data collection device (Datapaq) which was placed in a
water cooled box and the box was wrapped with insulation
material, as shown by the white package in Figure 5(b). The
slab temperatures recorded by the Datapaq unit were extracted
after the trial and compared with the predictions from the 2D
heat transfer model.

Figure 4 Thermocouple installation
(a) (b)
Figure 5 Thermocouple assembly (a) and trial slab being
charged into the reheating furnace (b)
Under normal production conditions, slabs of different
steel grades with different reheating discharge temperatures run
through the reheating furnace on a daily basis. In the present
trial, the instrumented slab was cold charged into the reheating
furnace with an aim discharge temperature of 2350 °F. The
thermal properties of the steel, which were used in the model
calculation, are shown in Figure 6.
(a)Specific Heat
(b) Thermal Conductivity
Figure 6 Steel thermal conductivity and specific heat
The instrumental slab position in the reheating furnace was
tracked during the trial. Figure 7 shows the walking speed of
the trial slab, which was calculated based on the slab positions
4 Copyright © 2016 by ASME
in the reheating furnace during the two hour reheating process.
It shows that the trial slab traveled through the reheating
furnace at various speeds over the two hour reheating process.
This speed profile was used in the 2D heat transfer model
calculation. It should be pointed that in real production, this
kind of walking pattern inside a reheating furnace happens
since there could be a mill delay or a speed-up. Unlike the 2D
model developed by Chen et al. previously, the current model is
capable of handling a situation like this.
Figure 7 Slab walking speed
The slab temperature profiles predicted by the 2D heat
transfer model and measured by the thermocouples installed in
the intrumented slab are shown in Figure 8.
(a) Top thermocouples
(b) Core thermocouples
(c) Bottom thermocouples
Figure 8 Thermocouple measurements vs. 2D model
predictions
It can be seen from the comparison presented in Figure 8
that the thermocouple measurements and the model predictions
are in good agreement throughout the entire reheating process.
It is also shown that the 2D model is able to capture the details
of the slab temperature evolution such as the transition point on
the temperature profiles shown in Figure 8(b). Based on the
specific heat of the steel shown in Figure 6, there is an abrupt
change in specific heat at approximately 1500 °F due to the
phase change taking place in the steel during reheating. A
comparison of the slab bulk temperatures (the average value of
the temperatures at six different locations) between the
thermocouple measuremnt and the model prediciton is shown
in Figure 9, and a good agreement can be seen.
Figure 9 Comparison of slab bulk temperature
Further comparison between the model predictions and the
thermocouple measurements at discharge are presented in Table
1. The results show that the maximum temperature difference
(T) between the model prediction and the thermocouple
measurement is 27 °F, from the slab bottom side of the
between-skid location. Overall, the slab bulk temperatures at
discharge are in reasonable agreement between the 2D model
prediction and the thermocouple measurement. The
difference is only 18 °F. Both bulk temperatures are also
close to the aim discharge temperature (2350 °F), indicating
that the reheating process produced a fully reheated slab.
Table 1 Comparison of slab discharge temperature between
measurements and 2D heat transfer model prediction
#1 #2 #3 #4 #5 #6
Measurement
2345 2340 2326 2340 2335 2332
(°F)
2D Model
2342 2330 2339 2351 2344 2359
(°F)
T (°F) -3 -10 13 11 9 27
Comparison with Online Model
The reheating furnace is controlled by an online Level II
model which monitors slab temperatures through the
thermocouples installed in the reheating furnace and controls
the slab movement and fuel and air flow rates according to
multiple set points defined under certain operating conditions.
At the end of reheating, slabs will be discharged if the
difference between the model predicted slab bulk temperature
and aim discharge temperature is within a specified value.
In order to compare the present offline 2D model with the
online Level II model, a total of 210 reheating cases were
studied, and the results are shown in Figure 10. The selected
cases covered both cold and hot charging conditions with aim
discharging temperatures ranging from 2200 to 2350 °F. For
the offline model calculations, the furnace temperatures
measured by the thermocouples installed in the different zones
of the furnace were used to define the thermal boundary
conditions. With specific slab thermal properties and walking
speeds, slab discharge temperatures at discharge for individual
slabs were calculated. It can be seen from the data presented
in Figure 10 that the differences between the offline and online
model predictions are within ± 15 °F.
2400
2350
±15 °F
2300
2D Model (°F)
2250
2200
2150
2150 2200 2250 2300 2350 2400
Online Model (°F)
Figure 10 Discharge temperature comparison between the
offline and online model predictions
CONCLUSIONS
A comprehensive two-dimensional (2D) numerical heat
transfer model for slab reheating in a walking beam furnace
was developed using the finite difference method. The model
is capable of predicting slab temperature evolution during a
reheating process based on real time furnace conditions and
steel physical properties. The model was validated by using
mill instrumented slab trials and production data. The results
show that the temperature evolution predicted by the model is
in good agreement with that measured by the thermocouples
embedded in the instrumented slab. Compared with 3D CFD
Bulk
simulation of a reheating process, this 2D heat transfer model
used for predicting slab temperature evolution requires less
2336 computing power and can provide results in a few seconds.
This is a very convenient and user-friendly tool which can be
2354
used easily by mill metallurgists in troubleshooting and process
18 optimization.
ACKNOWLEDGMENTS
The authors would like to thank the rolling mill personnel
at SSAB Mobile for their great support in the course of the
project. The authors also want to express their great
appreciation to the SSAB Americas senior management for
their permission to publish this work. The Center for
Innovation through Visualization and Simulation (CIVS) at
Purdue University Calumet is also gratefully acknowledged for
providing all the resources required for this work.
NOMENCLATURE
5 Copyright © 2016 by ASME
 simulation”. International Journal of Thermal Science
𝑇(𝑥, 𝑦, 𝑡) Temperature at time step 𝑡 84(2014) 214-227.
𝑥 Slab longitudinal direction coordinate 10. Jong Gyu Kim, Kang Y. Huh, Il Tae Kim. “Three-
𝑦 Gauge direction coordinate dimensional analysis of the walking beam type slab
𝐶𝑠 (𝑡) Specific heat of the slab reheating furnace in hot strip mills”. Numerical Heat
𝜌𝑠 Density of the slab Transfer, Part A, 38(2000) 589-609.
𝐾𝑠 (𝑡) Thermal conductivity of the slab 11. Chia-Tsung Hsieh, Mei-Jiau Huang, Shih-Tuen Lee, Chao-
Hua Wang. “Numerical modeling of a walking beam type
∆𝑥 increment of 𝑥 coordinate
slab reheating furnace” Numerical Heat Transfer, Part A,
∆𝑦 increment of 𝑦 coordinate
53(2008) 966-981.
∆𝑡 increment of time
𝑡
12. Man Yong Kim. “A heat transfer model for the analysis of
𝑇𝑖,𝑗 node temperature at time step 𝑡, transient heating of the slab in a direct-fired walking beam
𝑡+∆𝑡
𝑇𝑖,𝑗 node temperature at time step 𝑡 + 1 type reheating furnace”. International Journal of Heat and
𝐺 gauge of the slab Mass Transfer 50(2007) 3740-3748.
𝑇(𝑥, 𝐺, 𝑡) top surface (y=G) temperature at time step 𝑡 13. Vinod Kumar Singh, Prabal Talukdar. “Comparisons of
𝑓 view factor different heat transfer models of a walking beam type
𝜎 Stefan-Boltzmann constant reheat furnace”. International Communication in Heat and
𝜀𝑠 emissivity of the slab Mass Transfer 47(2013) 20-26.
14. Jung Hyun Jang, Dong Eun LEE, Chongmin Kim, Man
𝑇𝑖𝑡𝑜𝑝 convective heat transfer coefficients
Young Kim. “Prediction of furnace heat transfer and its
influence on the steel slab heating and skid mark formation
REFERENCES
in a reheating furnace” ISIJ International, 48(10), 2008,
1. Sang Heon Han, Daejun Chang, Chang Yong Kim. “A
1325-1330.
numerical analysis of slab heating characteristics in a
15. Tiago Morgado, Pedro J. Coelho, Prabal Talukdar.
walking beam type reheating furnace”. International
“Assessment of uniform temperature assumption in zoning
Journal of Heat and Mass Transfer 53(2010) 3855-3861.
on the numerical simulation of a walking beam reheating
2. Sang Heon Han, Seung Wook Baek, Man Yong Kim.
furnace”. Applied Thermal Engineering 76(2015) 496-508.
“Transient radiative heating characteristics of slabs in a
16. Chee Keong Tan, Joana Jenkins, John Ward, Jonathan
walking beam type reheating furnace”. International
Broughton, Andy Heeley. “Zone modeling of the thermal
Journal of Heat and Mass Transfer 52(2009) 1005-1011.
performances of a large scale bloom reheating furnace”.
3. Mingyan Gu, Guang Chen, Xuhui Liu, Cengceng Wu,
Applied Thermal Engineering 50 (2013) 1111-1118.
Huaqiang Chu. “Numerical simulation of slab heating
17. Jose Manuel Casal, Jacobo Porteiro, Jose Luis Miguez,
process in a regenerative walking beam reheating furnace”
Alfonso Vazquez. “New methodology for CFD three
International Journal of Heat and Mass Transfer 76(2014)
dimensional simulation of a walking beam type reheating
405-410.
furnace in steady state”. Applied Thermal Engineering 86
4. Sang Heon Han, Daejun Chang. “Optimum residence time
(2015) 69-80.
analysis for a walking beam type reheating furnace”.
18. Ali Emadi, Ahmad Saboonchi, Mahdi Taheri, Saeid
International Journal of Heat and Mass Transfer 55(2012)
hassanpour. “Heating characteristics of billet in a walking
4079-4087.
hearth type reheating furnace”. Applied Thermal
5. Sang Heon Ham, Daejun Chang, Cheol Huh. “Efficiency
Engineering 63(2014) 396-405.
analysis of radiative slab heating in a walking-beam-type
19. Shaojie Chen, Drew Poshard, Sunday Abraham.
reheating furnace” Energy 36(2011) 1265-1272.
“Modification of reheat furnace practices through
6. Maria T. Johansson. “Bio-synthetic natural gas as fuel in
comprehensive process modeling”, AISTech 2007 -
steel industry reheating furnaces – a case study of
Proceedings of the Iron and Steel Technology Conference,
economic performance and effects on global CO2
Indianapolis, IN, USA.
emissions”. Energy 57(2013) 699-708.
20. Yu Jin Jang, Sang Woo Kim. “An estimation of a billet
7. Flavio A.D. Oliveira, Joao A. Carvalho Jr., Pedro M.
temperature during reheating furnace operation”
Sohrinho, Andre de Castro. “Analysis of oxy-fuel
International Journal of Control, Automation, and System.
combustion as an alternative to combustion with air in
5(1) 2007, 43-50.
metal reheating furnaces”. Energy 78(2014) 290-297
21. Guangwu Tang, Bin Wu, Dengqi Bai, Yufeng Wang, Rick
8. Jung Hyun Jang, Dong Eun Lee, Man Yong Kim, Hyong
Bodnar, and Chenn Zhou, “Numerical Simulation of a
Gon Kim. “Investigation of the slab heating characteristics
Walking Beam Slab Reheating Furnace,” to be published
in a reheating furnace with the formation and growth of
in AISTech 2016 conference proceedings.
scale on the slab surface” International Journal of Heat and
Mass Transfer 53(2010) 4326-4332.
9. Satish Kumar Dubey, P. Srinivasan, “Development of three
dimensional transient numerical heat conduction model
with growth of oxide scale for steel billet reheat

6 Copyright © 2016 by ASME

View publication stats