---

Available online at www.sciencedirect.com

@#

”*5” ScienceDirect
JOURNAL OF IRON AND STEEL RESEARCH, INTERNATIONAL. 2008, 15(1) : 70-74

Non-Metallic Inclusion Distribution in Surface Layer of IF Steel Slabs

ZHANG Qiao-ying’ , WANG Li-tao’ , WANG Xin-hua’ , LI Hong’ , WANG Wan-jun’
(1. School of Metallurgical and Ecological Engineering, University of Science and Technology Beijing ,
Beijing 100083, China$ 2. Shougang Iron and Steel Corporation Ltd, Beijing 100041, China)

Abstract: Non-metallic inclusion distribution in the surface layer of IF steel slabs during unsteady casting was investi-
gated using the original position statistic distribution analysis (OPA) method. It was found that most non-metallic
inclusions larger than 10 pm existed in the subsurface layers of 0. 5 - 3. 5 mm from the slab surfaces and very few
large non-metallic inclusions were found in the inner regions ( 2 4 . 5 mm from slab surfaces). In addition, it was
found that at high casting speed level (1.4 m/min), even a slight change of casting speed could result in a remarkable
increase of the non-metallic inclusions. Thus, at high casting speed, changing the casting speed should be avoided or
considerably lower speed changing rate must be used.
Key words: IF steel: transient casting; slab surface layer: non-metallic inclusion

The interstitial-free steel (IF steel) is widely steelmaking plants, the methods of “down grade” or
used in automotives. Since the sheets, especially for “defect slab” are adopted, i. e. the slabs that are
exposed panels require good surface quality, the cast during the casting speed change period are used
control of the non-metallic inclusions, especially for producing lower grade products or even consid-
large sized inclusions in I F steels is very strict. In ered as scraped slabs. When this method is simply
recent years, the surface defects of I F steel sheets used, some good slabs may be “down grade” or
caused by steelmaking, and secondary refining, etc. scraped and some defect slabs may escape as good
have been significantly decreased owing to the pro- slabs. T o improve the quality control of I F steel in
gress and improvement of the steelmaking and refi- Baosteel , investigation on the non-metallic inclu-
ning technologies and equipment. However, the ra- sions distribution in the surface layers of I F steel
tio of the defects owing to the mold powder entrap- slabs was carried out.
ment during continuous casting has increased”-’].
1 Experimental Method
Mold powder entrapment mainly occurs in the
so-called “transient casting” periods, e. g. the time The experiments were carried out at the No. 2 BOF
of casting start, casting end, ladle exchange, mold shop of Baosteel. After the desulphurization pretreat-
width change, SEN change, etc. During the un- ment, the hot metal was blown in a 250 t top and bot-
steady casting periods, usually, the casting speed V, tom combined blowing furnace and the liquid steel
is largely varied, which disturbs the flow of the liq- after BOF steelmaking was further refined in an R H
uid steel in the mold and results in severe level fluc- degasser. There are two slab casters in the No. 2
tuation. Hence, mold powders are easy to be carried BOF shop and their main specifications are shown in
into the liquid steel and some are entrapped by the Table 1. In the experiments, slabs of 1 150 mm
solidified shell and exist as non-metallic inclusions in width and 230 mm thickness were produced.
slabs. Normal, increasing or decreasing the casting
Casting speed variation must be avoided during speed were intentionally arranged in the experiments
casting. However, in practical casting production, to investigate the influence of the V, change on the
casting speed change is still inevitable. In most non-metallic inclusions distribution in the surface layer

Biography:ZHANG Qiao-ying(l976-), Female, Doctor Candidate! E-mail: dyn19781223@163. comi Revised Date: April 14, 2006
Issue 1 Non-Metallic Inclusion Distribution in Surface Layer of IF Steel Slabs 71

Table 1 Main specifications of slab casters Subsequently, each specimen was further machined
Machine type Vertical bend totally 5 times and 1 mm of surface layer was ma-
Vertical length 2. 75 m chined away at each time. After each machining, the
Tundish capacity 60 t specimen was analyzed by OPA.
Slab thickness 230 mm, 250 mm OPA is a method based on the techniques of
Mold length 900 mm continuous excitation of spark spectrum with two-di-
Normal casting speed 1.0- 1. 6 m/rnin
mensional moving on specimens, high speed signal ac-
Submerged nozzle Immersion depth: 170 mm
quisition of single spark discharge and the data analysis
Outlet anEle: 15’
~ y s t e m [ ~ - ~ It
’ . can move and scan on large sized
specimens and obtain signals reflecting both the con-
of I F steel slabs. In the experiments of normal cast- centrations and locations of the non-metallic inclu-
ing speed, after casting, specimens were also taken sions on the scanned areas. T h e scanning speed of
on the slabs, which were normally cast at 0. 6 m/min, OPA used in the investigation was 1 mm/s and the
0.8 m/min, 1.0 m/min, 1.2 m/min, and 1 . 4 m/min. interlineation between each scanning was 2 mm.
In the experiments of decreasing casting speed, after T h e diameter of each spark spot was 1 pm and the
V, was maintained at 1. 4m/min for more than 30 excitation depth was 9-20 pm.
min, the casting speed was decreased evenly at the When the non-metallic inclusions on the speci-
speed decreasing rate of 0. 15 m/min2. After cast- men surfaces were detected by O P A , the intensities
ing, the specimens were taken from the slabs, which of the spectrums, which reflect the existence of the
were j u s t in the mold during casting when the cast- non-metallic inclusions, were abnormally increased.
ing speed was decreased t o 1 . 2 m/min, 1 . 0 m/min, T h e net intensities of the spectrums of the abnormal
0. 8 m/min, and 0. 6 m/min, respectively. In the spark signals were correlative to the sizes of the non-
experiments of increasing casting speed, after V, metallic inclusion particles, and the relative occur-
was maintained at 0. 6 m/min for more than 30 min, rence frequencies of the abnormal spark signals were
the casting speed was evenly increased a t the rate of correlative to the contents of the non-metallic inclu-
0. 15 m/min2. After casting, specimens were taken sionsC8’. By analyzing the data of the net spectrum
from the slabs, which were j u s t in the mold when intensities and the relative occurrence frequencies of
the casting speed was increased t o 0. 8 m/min, 1. 0 the abnormal spark signals, the contents and sizes of
m/min, 1. 2 m/min, and 1.4 m/min, respectively. the non-metallic inclusions on the scanned surfaces
Three specimens (of length 200 mm, width 100 mm, can be obtained [Eqn. (1) and Eqn. (Z)].
thickness 25 mm) were respectively cut on the top C=C,F+K (1)
surfaces at 1/8 width, 1 / 2 width, and on the narrow D = ( t z / t l )Z+D, (2)
faces of the slabs ( a s shown in Fig. 1). For each where, C is the content of the non-metallic inclu-
specimen, 0. 5 mm surface layer was first machined sions; C, is a constant; F is the relative occurrence
away. T h e specimen was scanned to obtain the con- frequency of the abnormal spark signals; K is a con-
tent of the non-metallic inclusions on the detected stant; D is the size of the non-metallic particle; t z
surface using the original position statistic distribu- and t , are coefficients; Z is the net intensity of the
tion analysis (OPA) method, which was developed by spectrum of the abnormal spark signal, and Dois the
Beijing NAKE Analysis Instrument Co. The scanning critical value of the size of the particle, which is
area on each specimen by OPA was 60 mmX40 mm. taken as the non-metallic inclusion by OPA.

Casting 2 Experimental Results

Center
Fig. 2 shows the number of non-metallic inclusions
in the subsurface (0.5 - 5. 5 mm from surfaces) speci-
Ccnfcr mens of slabs analyzed by the OPA method: Fig. 2 ( a >
l i C
shows the number of non-metallic inclusions be-
tween 10 pm and 20 pm and Fig. 2 ( b ) shows the
number of non-metallic inclusions larger than 20 pm.
The slabs were normally cast respectively at 0.6 m/min,
Fig. 1 Specimen machining scheme 0.8 m/min, 1.0 m/min, 1 . 2 m/min, and 1.4 m/min
 - 72 Journal of Iron and Steel Research, International Vol. 15

v3- 1-2 mm
32

24

16

0.6 1.0 1.4 0.6 1.0 1.4

Casting speed/(mmin-')

(a) Inclusions between 10-20 pm; ( b ) Inclusions larger than 20 pm

Fig. 2 Number of non-metallic inclusions in slab subsurface specimens analyzed by OPA (normal casting speed)

without changing V,. specimens, which were 0. 5 mm, 1. 5 mm, 2. 5 mm,

In Fig. 2 , it was found that most of the non-me- and 3. 5 mm from the slab surface. Also, the casting
tallic inclusions larger than 10 pm existed in 0. 5 - speed change is illustrated in the figure; Fig. 3 ( a >
3. 5 mm surface layers of slabs and very few non-me- shows the number of non-metallic inclusions be-
tallic inclusions were detected by OPA in specimens tween 10 pm and 20 p m , and Fig. 3 ( b > shows the
that were more than 4. 5 mm depth from the slab number of non-metallic inclusions larger than 20 pm.
surfaces. In Fig. 3 , it was found that most of the non-me-
As seen in Fig. 2 , the non-metallic inclusions tallic inclusions larger than 10 pm existed in 0. 5 -
detected by OPA were more in slabs that were cast 3. 5 mm surface layers of slabs and very few non-me-
at low casting speeds ( 0 . 6 m/min-0. 8 m/min). It tallic inclusions were detected by OPA in specimens
is also seen in Fig. 2 that when the casting speed was that were more than 4. 5 mm depth from the slab
increased to 1 . 4 m/min, non-metallic inclusions in surfaces.
the slab surface layers tended to increase. In this ex- It is seen in Fig. 3 that in the later casting speed
periment, slabs cast at 1. 0 m/min and 1 . 2 m/min decreasing period when V, was evenly decreased fur-
V, contained the minimum amount of non-metallic ther to 0. 8 m/min, the number of non-metallic in-
inclusions. clusions was the fewest. When the casting speed
Fig. 3 shows the number of non-metallic inclu- changing was stopped after the V, was decreased to
sions in the surface layer specimens of slabs that 0. 6 m/min, a very slight increase of the non-metal-
were cast in the experiment of decreasing casting lic inclusions was found.
speed. The number of non-metallic inclusions at each Fig. 4 shows the number of non-metallic inclu-
V, is the average result of OPA on the surface layer sions in the surface layer specimens of slabs that were

)' A\2-3 mm
3-4 mm
\
24
* \ o-:mm 1-2 m m

0-1 m m
=/,. 12 3-4mm \ /
4-5 mm 6-6 mm
0-03=8L~
0-0
0
1.2 1.0 0.8 0.6 1.2 1.0 0.8 0.6
Casling speed/(mmin-')

(a) Inclusions between 10-20 pmi (b) Inclusions larger than 20 pm

Fig. 3 Number of non-metallic inclusions in slab subsurface specimens analyzed by OPA (decreasing casting speed)
Issue 1 Non-Metallic Inclusion Distribution in Surface Layer of IF Steel Slabs ' 73

0.8 1 .0 1.2 1.4 0.8 1.0 1.2 1.4

Casting speed(m.rnin-')

( a ) Inclusions between 10-20 pm; ( b ) Inclusions larger than 20 pm

Fig. 4 Number of non-metallic inclusions in slab subsurface specimens analyzed by OPA (increasing casting speed)

cast in the experiment of increasing casting speed. 0.6 m/min to 0.8 m/min, 0.6 m/min to 1.0 m/min,
T h e number of non-metallic inclusions at each V, is 0. 6 m/min to 1. 2 m/min.
the average result of O P A on the surface layer speci- Fig. 5 shows several large sized non-metallic in-
mens, which were 0. 5 m m , 1. 5 m m , 2. 5 mm, and clusions found in subsurface specimens by SEM;
3. 5 mm from the slab surface. Also, the casting their chemical compositions analyzed by EDS are
speed change is illustrated in the figure; Fig. 4 ( a > shown in Table 2. T h e distances from slab surface
shows the number of non-metallic inclusions be- in Fig. 5 ( e > and ( f ) are 1. 5 mm and 1. 78 m m , re-
tween 10 pm and 20 p m , and Fig. 4 ( b ) shows the spectively. These large non-metallic inclusions all
number of non-metallic inclusions larger than 20 pm. contain N a z O , which indicates that these were en-
In Fig. 4 , it was found that most of the non-metal- trapped mold powders or the reaction products be-
lic inclusions larger than 10 pm existed in 0. 5-3. 5 mm tween the involved mold powders and the non-metal-
surface layers of slabs and very few non-metallic inclu- lic inclusions, such as Alz03 in liquid steels.
sions were detected by OPA in specimens that were more T h e thickness of the solidified shell in the mold
than 4.5 mm depth from the slab surfaces. can be calculated by the following equation:
It is seen in Fig. 4 t h a t w h e n increasing cast- d= K ( L / V ,)"' (3)
ing speed w a s ended and t h e V, w a s increased to where d is thickness of the solidified shell, mm; K
1. 4 m/min, the number of non-metallic inclusions is solidification coefficient, mm/min"2 ; and L is dis-
was more than others when V, was increased from tance from the liquid steel level in the mold, mm.

Fig. 5 Photographs of typical large sized non-metallic inclusions in subsurface specimens

 74 ’ Journal of Iron and Steel Research, International Vol. 15

Table 2 Non-metallic inclusions and their chemical compositions of Fig. 5 by SEM-EDG

No. Al Ca Si K Mg Na Mn S Fe Distance from surfaces
a 4. 42 37.03 18. 75 0. 97 1.08 17. 99 1.50 18.24 0 3-4 mm
b 68. 14 0 0. 30 9. 49 0. 26 5. 44 6. 89 9. 49 0 3-4 mrn
C 0 14.02 0 32.81 10.81 40. 12 0 2. 24 0 2-3 mm
d 22.79 5. 69 31.70 13.28 0 20.54 0 6. 00 0 1-2 rnm
e 11.08 5. 54 12.93 0.71 13.02 9. 34 4. 47 0 42.95 1.5 mm
f 2.35 32.06 23.61 0 2. 26 9. 68 3. 71 0 26.33 1.7 mm

In the experiment, the casting speed was within entrapment calculated is 26 - 54 mm below the level
the range of 0. 6 m/min to 1. 4 m/min, and most of of the liquid steel in the mold.
the large sized non-metallic inclusions caused by the
entrapment of the mold powders were found in the References:
subsurface regions between 0-4 mm below the slab
[l] Emling W H , Waugaman T A , Feldbauer S L, et al. Subsur-
surfaces. T h e solidification coefficient K is usually face Mold Slag Entrapment in Ultra Low Carbon Steel [A].
between 18 rnm/min’/’ and 24 mm/min’’2 for slab ISS, eds. 77th Steelmaking Conference Proceedings [C]. War-
caster. Assuming K = 20 rnm/min’/’ , using Eqn. ( 3 ) , rendale, PA: ISS, 1994. 371.
the maximum depth of the mold powder entrapment [21 Iguchi M, Yoshida J , Shirnizu T. Model Study on the Entrap-
ment of Mold Powder Into Molten Steel [J]. ISIJ International,
calculated is 26-54 mm below the level of the liquid
2000, 40(7): 685.
steel in the mold. [3] Tanizawa K. Influence of the Steelmaking Conditions on Non-
metallic Inclusions and Product Defects [J]. Metallurgia Itali-
3 Conclusions ana, 1992, 84(1): 17.
[4] WAN Xiao-guang, HAN Chuan-ji, CAI Kai-ke. Experimental
( 1)
Most non-metallic inclusions larger than
Research of Submerged Entry Nozzle of Slab Continuous Cast-
10 pm existed in the subsurface layers of 0. 5 - 3. 5 ing [J]. Iron and Steel, 2000, 35(9): 20 (in Chinese).
mm from the slab surfaces and very few large non- [5] Suzuki M, Suzuki M , Nakada M. Perspectives of Research on
metallic inclusions were found in the inner regions High-speed Conventional Slab Continuous Casting of Carbon
Steels [J]. ISIJ International, 2001, 41(7) : 670.
(>4. 5 mm from the slab surfaces).
C6l CAI Kai-ke. Continuous Casting [MI. Beijing: Science Press,
(2) A t high casting speed level (1. 4 m/min) , 1991 (in Chinese).
even a slight change of casting speed could result in a [7] Gupta D, Lahiri A K. Cold Model Study of the Surtace Profile
remarkable increase of the non-metallic inclusions. in a Continuous Slab Casting Model: Effect of Second Phase
Thus, at high casting speed, changing the casting [J]. Metallurgical and Material Transactions, 1996, 27B(8) :
695.
speed should be avoided or considerably lower speed
[8] Jonsson L, Jonsson P. Modeling of Fluid Flow Conditions
changing rate must be used. Around the Slag/Metal Interface in a Gas-Stirred Ladle [J].
(3) T h e maximum depth of the mold powder ISU International, 1996, 36(9): 1127.