66 Steel Technical Report, No.

China 25, pp.

A Novel 66－72, (2012)
Temperature Diagnostic System for Stelmor Air-Cooling of Wire Rods

A Novel Temperature Diagnostic System for Stelmor

Air-Cooling of Wire Rods

CHUN-JEN FANG* and YI-YING LIN**

*New Materials Research & Development Department
**Metallurgical Department
China Steel Corporation

A novel online and movable temperature diagnostic system was designed to capture the surface temperature
profile of wire rods during the Stelmor air-cooling process. It consisted of an infrared pyrometer, a data
logger, a cooling system, 3-axis steering gear, and a movable mechanical platform. A series of experiments had
been conducted to investigate the relationship between the cooling rate and the mechanical properties of
NLP(Non-Lead Patenting) wire rods. The temperature of the wire rods during the Stelmor air-cooling process
was more accurately controlled through the use of this novel system and the variation in the tensile strength
of NLP wires was reduced from 40% to 16%.
Keywords: Stelmor, NLP (Non-Lead Patenting) Wire Rods, Temperature Diagnostic

1. INTRODUCTION Figure 1 shows the Stelmor conveyor in the rod

mill of CSC. The total length of the Stelmor conveyor
Forced air-cooling processes such as the “Stelmor”
in the rod mill is 100 m, and it can be divided into ten
cooling conveyor are generally used as cooling methods
cooling zones. Zone Nos. #1 to 5 are the fast air-cooling
in a hot rolled wire rod mill. The strength and ductility
areas, which are used to cool the high carbon steel wire
of NLP(Non-Lead Patenting) wire rods can be controlled
rods rapidly. The zone Nos. #6 to 10 are the slow
in the rolling line with these processes. However,
air-cooling areas used for alloy steel wire rods cooling.
because forced cooling is performed with an air blast,
The wire rod coils are arranged to pass over several
the cooling capacity of the forced air-cooling process is
zones in order to obtain the desired microstructure and
often inadequate. Generally speaking, the air flow from
corresponding material properties. Each zone is
the plenum chamber cannot be accurately controlled for
equipped with fans to provide the forced air for cool-
the required cooling rate of wire rods because of the
ing, and each fan is connected via a plenum chamber to
lack of a temperature measurement system in the
the bottom of the conveyor deck as shown in Fig.2. Air
air-cooling process for processes such as the Stelmor in
flows through a series of nozzles on the deck and up
the Rod Mill at China Steel (CSC).
through the coils to complete the air-cooling process.

Fig.1. The Stelmor conveyor in the rod mill at Fig.2. Assembly of a single plenum Stelmor air
China Steel. conveyor system.
 Chun-Jen Fang and Yi-Ying Lin 67

Hence, variable-speed fans provide the flexibility for tions for the temperature measurement of the wire rod
controlling the air flow rate. Meanwhile, two baffles coils during the Stelmor process are shown in Fig.3.
are installed within the plenum and provide the func- There are three major positions (P1, P2 and P3) for
tion to direct the air flow to the side or center regions of taking temperature measurements. Both side positions
the conveyor deck through adjusting the baffle angles. (P1, P3) and the center position (P2) for the tempera-
From Fig.1, it is clear that the wire rod coils pass ture measurement of the coils are 15cm and 60cm from
over the cooling nozzles on the Stelmor conveyor and the conveyor wall, respectively.
are not in a static state, but always move forward with
the rotation of the cooling conveyor equipment.
Accordingly, a surface temperature profile of the wire
rods measured with a contact thermocouple is not
suitable for Stelmor air-cooling process. Kawasaki
Steel Corporation(1) has been using the wire rod coils
attached to a pre-welded thermocouple to assess the
Stelmor conveyor cooling performance, but this
measurement method can only be applied to offline
experimental equipment, and cannot be applied to the
temperature measurements of an actual on-line Stelmor
air-cooling production process. When Maanshan Iron
& Steel Company(2) were revamping their Stelmor
conveyor equipment, they tried to add a series of
infrared temperature measurement systems above the
Stelmor conveyor and connect these to the process
computer in order to more precisely control the cooling
rate of the wire rods. Baosteel Corporation(3) also
Fig.3. The positions (P1, P2 and P3) taken for
developed an online cooling control system through the
temperature measurement.
use of improved Stelmor equipment. From the existing
literature survey, although there have been many recent
Figure 4 shows the temperature measurement
studies of the Stelmor air-cooling process, almost all
results using the dual-color pyrometer. The center posi-
the studies have focused on air-cooling control. No
tion for the temperature measurement of the coils had
novel temperature diagnostic system or equipment can
more gaps between the wire rods, so the diagnostic
be found in the existing literature.
signal from wire rod coils was unstable and fluctuant.
In this study, we developed a wire rods temperature
However, at both the side positions for temperature
diagnostic system for the Stelmor process in the rod
measurement the coils were more compact and the
mill. The system combined an infrared pyrometer with
diagnostic signal was more stable and repeatable.
a mechanical platform capturing the surface tempera-
ture of wire rods immediately for the various districts
and gathered the temperature information for feedback
to the process computer in order to adjust the air flow
in the Stelmor air-cooling process.
2. EXPERIMENTAL METHOD
Recently, the more commonly used instruments for
wire rods temperature measurement on the Stelmor
conveyor is the infrared pyrometer.
In order to capture the surface temperature profile
of wire rods during the Stelmor air-cooling process, the
present investigation utilized both a single color and a
dual-color pyrometer to test which type is better for
wire rods temperature measurement. The size of the
area (spot size) to be measured determines the distance Fig.4. Both side positions were stable and repeatable.
between the pyrometer sensor and the wire rods. The
spot size must not be larger than the wire rods coil. The The dual-color pyrometer(4) utilizes two wave
pyrometer sensor should be mounted so that the mea- lengths in the continuous spectrum to measure the
sured spot size is smaller than wire rods coil. The posi- intensity of radiation from the wire rods and deduces
68 A Novel Temperature Diagnostic System for Stelmor Air-Cooling of Wire Rods

the temperature from the intensity ratio. This led to

some measurement variations from the gaps between
the wire rod coils during the switching of the two wave
lengths, as shown in Fig.4.
Figure 5 shows the temperature measurement
results of a single color pyrometer, clearly demonstrating
that the temperature data for the center position (P2) of
the coils was more stable than the data measured from
dual-color pyrometer, and showing a similar temperature
trend for both the side positions (P1, P3). Therefore, in
this study the single color pyrometer was used for the
setup and installation of the wire rods temperature
diagnostic system in the Stelmor air-cooling process.
Fig.6. The novel online and movable temperature
diagnostic system for Stelmor.

Fig.7(a). The wire rod temperature measurement

for a fast air-cooling process (W/O Encopanels).
Fig.5. The temperature measurement results of the
single color pyrometer.

Besides, there are also other mechanical equip-

ments and electrical systems that need to be developed
and designed for the novel online and movable
temperature diagnostic system in the Stelmor
air-cooling process. There are eight major components
of the temperature diagnostic system designed for wire
rods temperature measurement in Stelmor as follows,
and as shown in Fig.6: (1) Single Color Pyrometer; (2)
Movable Mechanical Platform; (3) 3-Axis Steering
Gear; (4) The Encode; (5) Data Logger; (6) Water
Cooling System; (7) Air Cooling System; and (8) The
Extended Line. The temperature diagnostic system can Fig.7(b). The wire rod temperature measurement for
be used for wire rods temperature measurement in dif- a slow air-cooling process (W/ Encopanels).
ferent manufacturing processes including fast and slow
air-cooling processes even for the case of within the 3. RESULTS AND DISCUSSION
encopanels on the Stelmor conveyor as shown in Fig.7.
Figures 7(a) and 7(b) show the situations of wire 3.1 Wire Rods Surface Temperature Distributions
rod temperature measurement for fast and slow Figure 8 presents the surface temperature distribu-
air-cooling process in Stelmor, respectively. tions in cross-section of NLP coils at different locations
 Chun-Jen Fang and Yi-Ying Lin 69

(Zone No.#1, Zone No.#2, Zone No.#6 and Zone Figure 9 shows the local air flow rate distributions in a
No.#10) along stock. The wire rod surface temperature cross-section of conveyor at different locations (Zone
distributions measured in terms of the self- developed No.#1, Zone No.#2, Zone No.#4 and Zone No.#6)
novel online and movable temperature diagnostic along stock and with different fan-loadings (100%,
system reveals that the surface temperature distributions 85%, 80% and 75%, respectively). The results reveal
in the cross-section of NLP coils are not uniform, so the that the center region (P2) of coils having a similar air
variation in tensile strength of NLP wires are about flow rate with the both side regions of coils (P1, P3).
40% before improving. Besides, the surface temperature The same phenomenon also can be found at different
distributions are symmetrical within the center region locations. Actually, the total heat capacities in the
(P2) of the coils and show a similar temperature trend both side regions of coils are more than the center
with the both side regions of the coils (P1, P3), this region, so the cooling capacities of the forced air-cooling
phenomenon also can be found at different locations. process for the both side regions are insufficiently.
Because, the center region (P2) of the coils has more
gaps between wire rods, the surface temperatures are
about 60~90℃ lower than at both the side regions of
coils (P1, P3).

Fig.9. Local air flow rate distributions in a cross-

section of the conveyor at different zones.

Fig.8. Surface temperature distributions in a cross- In the air-cooling process occurring on the Stelmor
section of the conveyor at different zones. conveyor, the wire rod experiences convective heat
transfer by forced air flow through the array of coil
Moreover, according to the continuous cooling loops. The temperature distribution across the coil array
transformation (CCT) diagram(5) of NLP wires, the can then be evaluated using an equivalent coil section.
optimum phase transformation temperature is between Figure 10 presents the typical laying array of wire rod
500 ~ 550℃. Because, a reheating effect occurs when coil loops, and it can be observed that the overlapping
the wire rods enter the slow cooling zone (Zone of coil loops is more dense in the side regions than in
No.#6), the surface temperature of NLP wires rose to the center region of the wire rod coil array. Hence, it is
about 630℃. Therefore, in order to capture the surface necessary to optimize the air distribution in order to
temperature of NLP wires, a novel online and movable ensure a uniform cooling effect on the wire rod coils. A
temperature diagnostic system was designed for Stelmor simple algorithm was originally designed to evaluate
air-cooling process in this study. Then, the temperature the required air distributions across the conveyor stock
decay, i.e. the air cooling rate from Zone No.#1 to Zone array, allowing the optimum control for the wire rod
No.#5, can be calculated, and the air cooling rate could coil cooling to be determined by adjusting the operating
be controlled by the fan-loadings and baffle angles. baffle angles.
As illustrated in Fig.10, P denotes the pitch
3.2 Local Air Flow Rate Distributions
relevant to different intersections. D is the coil loop
In order to obtain NLP wires with uniform diameter. Meanwhile, N is defined as the distances of
strength, it is necessary to optimize the air flow distri- overlapping intersections from center of one coil loop,
bution for the Stelmor air-cooling conveyor control. D/P.
70 A Novel Temperature Diagnostic System for Stelmor Air-Cooling of Wire Rods

16
(4) 900mm × 3.14 × ( ) × 4pieces = 502mm
360
The ratio of side regions to center region
408  628  502
=( ) × 100% = 130%
1178
So, the length (heat capacity) of the wire rod coils
at the side regions exceed that of the center region by
Fig.10. Typical laying array of wire rod coil.
about 30%. Accordingly, the required air flow rate to
compensate for higher heat capacity at the side regions
The presented simple geometrical model was then
can be obtained.
applied to produce NLP wires in the rod mill of CSC.
The values of P, D and N could be determined by the 3.3 Improving the Quality of NLP Wires
coil diameter and the conveyor roller speed. In this
A series of experiments have been conducted to
study, the coil loop diameter (D) is about 900mm and
investigate the relationship between the cooling rate
the number of overlapping intersections of wire rod coil
and the mechanical property of NLP wire rods. Since
(N) is about 20.
the cooling rate during phase transformations deter-
Therefore, in this case, the areas of the overlapping
mines the scale of microstructure obtained, the cooling
wire rod coils at the side and center regions can be
rate of wire rod coils were accurately controlled via the
divided into four kinds of symmetrical parts as shown
optimum settings of operating fan-loading and baffle
in Fig.11. Each of the four symmetrical parts,1 to 4, in
angle in each plenum. In this study, the wire rod surface
the areas of overlapping wire rod coils has their own
temperature distribution was measured and checked by
circumferential angle of about 75°, 13°, 20°, 16°, respec-
means of the self-developed novel online and movable
tively. Hence, the areas of each part can be calculated
temperature diagnostic system. In the rod mill of CSC,
with a simple mathematical formula as follows:
since the measurements were repeated to confirm the
optimum values of fan-loadings and baffle angles to
obtain the desired air distribution and the required
cooling effect under the conveyor deck, the material
qualities of the NLP wires have been markedly
improved.
A new controlled cooling process function for the
Stelmor system could be established to ensure the
formation of the required microstructure of wire rods.
From the results, shown in Fig.12, when the wire rods
entered the slow cooling zone (Zone No.#6), the
surface temperature of the NLP wires could be
controlled at about 500 oC and meet the optimum phase
Fig.11. Four kinds of symmetrical parts for over- transformation temperature requirement at 500~550oC.
lapping coils at side and center regions. Besides, there are no martensite texture be found in the
core of the wire rods.
Figure 13 presents measured results of NLP wires
The length of wire rod coils at center region surface temperature distributed in different locations
(part no.#1): along stock. A series of experiments had been
75 conducted to investigate the relationship between the
(1) 900mm × 3.14 × ( ) × 2pieces = 1178mm cooling rate and the mechanical property of NLP wire
360
rods. From the established operating database, it has
The length of wire rod coils at side regions: become possible to qualify the cooling effects of NLP
(part no.#2 to 4) Wires along stock of Stelmor conveyor, and therefore,
13 the air-cooling process parameters could be optimized.
(2) 900mm × 3.14 × ( ) × 4pieces = 408mm Results showed that the both side and center region
360
temperature difference of the ring-shaped pile of wire
20 rods on the conveyor reduced to less than 20~40oC.
(3) 900mm × 3.14 × ( ) × 4pieces = 628mm Besides, there are no martensite texture be found in the
360
 Chun-Jen Fang and Yi-Ying Lin 71

core of wire rods and the average tensile strength of diagnostic system. These coil surface temperature
wire rods up to 126 kg/mm2 high than the normal measurements were repeated to confirm the optimum
cooling process 122 kg/mm2. process parameters used for each plenum and the estab-
lished operating database was utilized to adjust these
process parameters for desired air distribution on deck
corresponding to the required cooling effect of Stelmor
conveyor.
Figure 14 shows the overall surface temperature
distributions of NLP wires during the Stelmor
air-cooling process. The air cooling rate from Zone
No.#1 to Zone No.#10 could be calculated and
controlled by the developed novel online and movable
temperature diagnostic system. The temperature of the
wire rods during the Stelmor air-cooling process was
more accurately controlled through the use of this novel
system and the variation in tensile strength of NLP
wires was reduced from 40% to 16%. Hence, the material
qualities of the NLP wires have been obviously
improved in the CSC No.#2 Rod Mill.

Fig.12. Surface temperature distributions of NLP

wires at different locations.

Fig.14. Overall surface temperature distributions

of NLP wires during the Stelmor air-cooling process

4. CONCLUSIONS
Fig.13. NLP wires surface temperature distribu-
tions with the optimized cooling process parameters The conclusions drawn from the results and
in different locations. discussion are presented below:
(1)The single color pyrometer is more stable for wire
rod temperature measurement than the dual-color
Finally, the Stelmor air-cooling process control pyrometer.
mode established in current work was applied to (2)The surface temperature distributions are symmetrical
improve the quality of NLP wires with uniform in the center region (P2) of the coils and there is a
strength. Measurements for coil surface temperature similar temperature trend in the both the side
distributions were performed to check the cooling regions of the coils (P1 and P3)；moreover this
effect along stock of Stelmor conveyor using the phenomenon also can be found at different loca-
developed novel online and movable temperature tions.
72 A Novel Temperature Diagnostic System for Stelmor Air-Cooling of Wire Rods

(3)In order to obtain NLP wires with uniform strength, ACKNOWLEDGEMENTS

the required air flow rate to compensate for the side
The author would like to thank No.#2 Rod Mill,
regions was about 30%.
Quality Assurance Section, and New Materials R&D
(4)When a new controlled cooling process function for
Department at China Steel for supplying the materials
the Stelmor system can be established, the surface
and facilities to carry out this research into temperature
temperature of NLP wires at Zone No.#6 could be
and air flow rate measurements and to perform the
controlled at about 500℃ and meet the optimum
experiments necessary to this work.
phase transformation temperature requirement at
500~550℃. REFERENCES
(5)The two sides and center temperature difference of
1. Y. Hanada, K. Ueno, A. Noda, H. Kondoh, T.
the ring-shaped pile of wire rods on the conveyor
Sakamoto, K. Mine, “New type Stelmor equipment
was reduced to less than 20~40℃. Besides, there
of wire rod and bar mill”, Kawasaki Steel Technical
was no martensite texture to be found in the core of
Report, 1986, No. 15, pp. 50-57.
the wire rods. The average tensile strength of the
2. Hebin Feng, “Research of Technological Parameters
wire rods was increased to126 kg/mm2, higher than
of Stelmor for High Carbon Wire Rolled at High
the 122 kg/mm2 achieved by the normal cooling
Speed”, Iron and Steel, 2001, Vol. 36, pp. 415-421.
process.
3. Xing'an Zhu, “Exploitation and Application of
(6)A series of experiments had been conducted to
Monitoring System for Stelmor Cooling”, Baosteel
investigate the relationship between the cooling rate
Technology, 2006, Vol. 6, pp. 526-532.
and the mechanical properties of the NLP (Non-
4. Vali Walter and R. Stephen de Voto,“A Two-Color
Lead Patenting) wire rods. The temperature of the
Pyrometer for the Measurement of Temperature
wire rods during the Stelmor air-cooling process was
and Contamination Level in a Spark-Heated
more accurately controlled through the use of this
Hypervelocity Wind Tunnel,” AEDC-TDR-62-174,
novel temperature diagnostic system and the varia-
August 1962.
tion in the tensile strength of the NLP wires was
5. Worcester Morgan, “Morgan Technology”, Morgan
reduced from 40% to 16%.
Construction Company, 1985. □