ISIJ International. Vol. 35 (1995), No. 7, pp.

886-891

Prevention of Red Scale Formation during Hot Rolling of Steels

HaruhikO ISHIHARA. Atsuki OKAMOT0.1)

Hikaru OKADA.Tomoki FUKAGAWA.1) Masatoshi
AZUMA2) and Yukio MATSUDA3)
lron and Steel Research Laboratories, SumitomoMetal Industries, Ltd., Kashima-cho, Kashima-gun, Ibaraki-ken, 31 Japan. 4
1) Iron and Steel Research Laboratories, SumitomoMetal Industries, Ltd., Fuso-cho, Amagasaki, Hyogo-ken, 660 Japan.
2) Wakayama Steel Works, SumitomoMetal Industries, Ltd., Minato, Wakayama, Wakayama-ken,640 Japan.
3) Tokyo Head Office. SumitomoMetal Industries, Ltd., Ootemachi. Chiyoda-ku. Tokyo, OOJapan. I
(Received on December15. 1994; accepted in final form on March lO. l995)

Red scale defects usually observed in high Si hot rolled strip were reproduced in a laboratory 3 stand
tandemmill. The effects of hot rolling and descaling conditions on the strip surface color and scale structure
were examined.
lrrespective of Si content in steel. the hot rolled strip surface becamered whenthe scale thickness before
rolling
was above 20,tm and the rolling temperature was below 900'C. It was found that surface part of
the scale (mainly FeO) was broken to powder by hot rolling at the temperature below 900"C. The red scale
of Fe.O, was formed by the oxidation of powdered scale during cooling. Thick scale formed during slab
soaking was completely removed by hydraulic descaling before rolling in low Si steel, whereas that was
not removedin high Si steel. This remained scale caused the red sca]e defects after rolling and cooling. The
application of obtained results to the hot strip mill production of red scaleless strip was discussed.

KEYWORDS:
scale; red scale; Si steel; descaling, hot rolled product rolling; oxidation.

Moreover, even when Si content is very low level, a

1. Introduction
similar red scale problem arises sometimesin commercial
lrregularlystriped red scale defects, sometimes called production.
tiger stripes, are often formed on hot rolled high strength This work was carried out, using a laboratory size 3
steel strips containing
more than 0.5 o/o Si.1 ~4) Since the tandem mill,5) to clarify the mechanismof red scale
red scale defects cause inhomogeneousor dirty appear- formation and engineering measuresto prevent red scale
ance even after pickling, several hot rolling conditions defects on low and high Si steels.
have been proposed to ~revent the red scale problem.
Control of slab heating temperature2,3) and strengthen- 2. Experimental Condition
ing of water jet descaling2) are someof those proposed
processes. However, these conditions were not sufficient
2.1. Hot Rolling Mill
in preventing red scale defect in the commercial pro- Layout of laboratory 3
tandemrolling mill is given in
djlction process of hot rolling. According to previous Fig. The
1. mill consists of two electric furnaces, an edger,
reports, the cause of red scale is the existence of Fayalite 3
a water jet descaler, a tandemrolling mill and a water
(Fe2Si04) which firmly bonds steel substrate and iron spray. The maximum
heating temperature of the electric
scale.4) However, the
reason why conventional descaler furnace is
1300'C. A gaseous mixture of N2. C02 and
can not remove the scale before rolling is not clear. H20 was used to simulate the conditions inside a
760 1800 11OO 1ooo No.3 Mill
No 1 Mill
.
No 2
.

Mi Il
DescaIer
Water spray
o
Edger D DD DO
Mani pu Iator

cr o cro

Fig. 1.

Experimental apparatus.

C 1995 ISIJ 886

 ISIJ Intemational, Vol. 35 (1995). No. 7
commercially operating slab heating furnace. An edger from the furnace, and fixed to manipulator, two samples
of 200 diameter is attached to simulate vertical scale
mm were separated. After cooling for 5
to 50 s in air to fixed
breaker (VSB) in front of the water jet descaler. Samples temperatures (750, 800, 850, 900, OOO'C), the one I
were transferred at 0.5m/s by manipulator from the 3
sample washot rolled with passes to 4.0mmthick and
edger to the entrance of rolling mill through a water jet the other sample was immediately water quenched to
descaler. Specification of descaler is as follows; pressure measurethe scale thickness before rolling. Each reduc-
of 15 MPa, discharge rate of I .8 l/s, spray angle of 25', tion ratio of the rolling was 20 o/. and rolling speed
was
lead angle of lO' to the foreword, vertical distance 1.67 m/s. After rolling, the samples
were cooled by water
between nozzle and sample of 250 mm. A radiation spray in air.
thermometer is attached at the entrance of Ist mill and
the exit of 3rd mill to measuretemperature of samples.
2.3.2. Test 2
The effect of atmosphere of water spray cooling zone
Rolling mills are lined up in the following order 4Hi,
was examined by using low Si steel. After being heated
2Hi and 4Hi whose work roll diameters are 200, 220 to temperatures between I OOOand 1280'C for 30min,
and 220 mm,respectively. Water cooling zone apparatus the low Si steel samples (8 mmt)were air cooled to 900'C
with controlled atmosphere is installed at the exit of 3rd for the secondary scale formation and then rolled with
rolling mill.
3passes, as wasdone in test I After rolling, the samples
.

2.2. Materials were cooled by water spray in N2 atmosphere.

Chemical compositions of test samples are shownin
Table 1. Steel I is a high strength steel containing 0.5 olo
2.3.3. Test 3
In order to examine the effect of descaling before
Si, while steel
2 is a low carbon mild steel of very low rolling, the high Si steel and the low Si steel (30mm
Si content. The samples (30mmt, mmtx 70mm"
8 x 200 thickness) were heated to temperatures between OOO I
mml) were machined from production slab (200 mmtx
1
and 280'C for 30min as wasdone in test l. After being
l 200mm~), after eliminating IO from surface and mm cooled from separating temperature at about I 050 to
20 from the center.
mm 970'C for 40 s in air, the samples were descaled by high
2.3. Test condition 3
water jet followed by pass rolling and then cooled by
The test rolling pass and descaling condition are
water spray in air. The rolling start temperature was
schematically gl'ven in Fig. 2. approximately 900'C.

2.3.1. Test l 2.3.4. Test 4

In order to examine the effect of the primary scale on
The two samples of 8 thick were placed one on
mm red scale formation, the primary thick scale usually
top of the other sample and then heated to temperatures
between 820 and 1280'C for 30 min in N2gas. Scale was formed on slab surface in commercial production was
simulated by heating samples of 30mmthick in an
not formed during this heating. Af~er being taken out
atmosphere of 77.lo/,N2+14.30/.H20+8.6'10C02 at
temperature between I 150 and 1300 'C for between 20
Table 1. Chemical composition of test samples. (masso/.)
and 120 min. After being taken out from the furnace, the
samples were passed through the edger and descaled by
Steel C Si
Mn P S sol A~
water jet for removal of the thick scale. After being rolled
1 0.09 0.54 1.46 0.009 0.014 O.019 High Si with I pass, samples were cooled for 40 s in air for
2 0.05 0.005 0.15 O.009 0.013 O.024 Low Si secondary scale formation, followed by seconddescaling,

N0'1 N0'2 N0'3

Test 1
Cover secondary
sca Ie
irif
r~;+
ijWa~ter ~
~6.'t~l:~- 5~)/4t
SPecimen 8 SPray

Cover secondary
No 1 N0'2 No'3
~t
'

Test 2 sca Ie
E~~~~\
SPecimen --~t8 ~:~- 5\~5 ~:;~

DescallerNo'l N0'2 No'3

Test 3
Cover
\9~F30~
I
I~5~_~ I~~~l:~~~,~\~1_,Fr~L6'W~aptreary

~~~:1
SPecI men
\-l
Ist Descall\er No 12nd N0'1 N0'2 No'3

~ Descaller
-~F ~-----
'

Test 4 pr sca Ie
imar~Ly
~~l
~>
SPecimen
30~25t
~::I_
---~25 t _' 15 tHF 10\~/HIF6
water
SPray
Fig. 2.
Schematic illustration of experimental procedures.
-
~~/ 'v'

887 C 1995 ISIJ

 ISIJ International, Vol. 35 (1995), No. 7
rollingand cooling by water spray in the air. The relation betweenscale thickness before rolling and
For samples of test I and test 2, L*a*b* index (JIS8729) a* value whenthe rolling temperature is changedbetween
were measuredby spectrophotometer. The value of a* 750 and I OOO'Cis shown in Fig. 6. Whenthe rolling
indicates Owhenthe visual color is gray. Whenits color temperature is lower than 900'C, and the scale thickness
becomesmore reddish, the a* value indicates higher. before rolling is settled to be more than 20/un, a* value
Cross s~ction micrographs of scale was also observed by becomeshigher than I and red scale wasformed. Onthe
using a scanning electron microscope. Thescale thickness other hand, whenthe rolling temperature is I OOO'Cand
wascalculated from scale weight change though pickling even if the scale thickness is 50 um, the value becomes
in hydrochloric acid with inhibitor, converting the scale lower than 0.5 and the scale was gray. In this way, this
density into 5.7 (g/cm3).6) This calculation result agrees test showsthat the scale thickness before rolling and the
with the result of the cross section observation by rolling temperature are important for the formation of
microscope. red scale.

3. Results
3.2. Test 2
The relation between the scale thickness before rolling
3.1. Test 1 and a* value for low Si steel water spray cooled at
The relation between the air cooling oxidation start different atmosphere is shown in Fig. 7. Whencooled
temperature and the scale thickness at 900'C is shown downin N2, the steel exhibited a low a* value, that is
in Fig. 3. The air cooling
was started by separating the gray surface, irrespective of the scale thickness before
sample, and just after the sample temperature reached rolling, while whencooled in air, the a* value increased
900'C it waswater quenchedand then the scale thickness as the scale thickness increased.
was measured.
In both types of steel, the scale thickness increases as
3.3. Test 3
the air cooling start temperature rises. However, the scale The surface appearance of steels rolled under test 3
of high Si steel is thicker than low Si steel at the same conditions is shownin Fig. 8. Both high and low Si steels
air cooling start temperature. The surface of both steels showedgray surface. Under these test conditions from
the result of Test 1, high and low Si steels formed scale
were gray at any air cooling start temperature. In this
of 40 and 60 pmthick, respectively, at about 900'C of
way, the scale thickness before rolling can be changed
by altering the air cooling start temperature. hot rolling start temperature. Therefore, if the secondary
Twoexamplesof the surface color appearancefor high scale was not removed by descaling, the red scale will
definitely form. Therefore, the present result
Si rolled samples are shownin Fig. 4. When air cooling proves that
the descaling operation by high pressure water descaling
time is short, which meansthe scale is thin
rolling,
,lm) before
the surface of sample is gray. Onthe other hand,
(12 has effectively removedthe secondary scale even in high
Si steel.
when air cooling time is long and the scale is thick
(30 ,am), it becomesred. This tendency is similar to the 3.4. Test 4
case of low Si steel. The surface appearance of steels rolled with test 4
Relation between scale thickness before rolling at conditions is shownin Fig. 9. LowSi steel showedgray
900'C and a* value of the sample after rolling and water surface, while high Si steel showedred surface. In this
cooling is shownin Fig. 5. Scale thickness before rolling
test, very thick scale had been formed during heating to
was obtained from Fig. 3. The result shows that, when 1220'C. But most part of it was removedby twice high
the scale before rolling is thick, a* value becomeshigher,
which meansthe sample surface appears more reddish.
pressure water descaling prior to final 3pass rolling.
Cross section photographs of the scale after Ist and 2nd
No difference in color was observed between the two descaling are shownin Fig. 10. No scale is left on low
types of steel. Si steel, while 4(~l80 /Im scale is left
on high Si steel after
Ist and 2nd descalings.
EJc:o
O:High Si As for high Si steel, the red scale
was formed on the
:~c:(D 60 A entire surface of the steel strip, when it was heated to
-a):J :s A:LowSi
~;~ 1 2
220'C for h. The condition of the red scale was
c>
=as A changedwhenheating condition was changed. The steel
o*
*:~
(D
u)*
40 A strip surfaces after rolling and water cooling at various
0~~
~(D (Q
AA o heating times and heating temperatures are shown in
Figs. 11 and 12. When
~:!~ A O heating time is short, the surface
cot
a)
(D o
'o 20
AAA O
O of the steel strip turns red and is irregular. Whenheating

~o OO temperature is 1
150'C, red scale is not formed. At
oo) ~)OOOCaO l 220'C, it is formed on the entire surface and at 300'C 1
Z t only part of the surface is covered.
-o
(1)oo
Changesof the red scale area ratio with the heating
~
(/)
.~ O900 1000 1100 time and temperature are changedare shownin Fig. 13.
Air cooling Start temperatUre ('C) The red scale is formed over most of the surface at
Fig. 3. Effect of air cooling start temperature on scale thick- l 180 and 1200'C. The higher the heating temperature
ness of steels quenchedfrom 900'C. (Test 1)
is, the less the ratio of red scale is.
Andwhenthe heating

C 1995 ISIJ 888

 ISIJ International, Vol. 35 (1 995). No. 7
~+
cs~:'e'~J ~5 :~~:;
c~
~'~5~:) c~)
~"-'e.
q)e'
v_ s'D~)'

eJ)eO
~'~,
O eOeJ)f~~ ~:;~H
~~ O(.~
c~So ,_
O
'
~~
~F~;~:i'
'::a~~i ~5i~
'~ ~~(•1:'q, O~
O a'h~;
~~ '~ ~; c:50o
c.c,~
1;,F~

~~:::~) ~::Oc~::5 ~~~j~S

-m
1:'
,_*c~a'~'e~~!
-(~ ~
~
~:'
e)
~~ ~O ~O(~1 ~
~-
~:,~)
~:'~:'~:'
~~i-1:'1:~
~)~
+*E~ O~ eJ) -F~F~ ~:l~~;
E,,e)
~~ v5~5y O ~~(~C~ ~~t)
Ohs" t~5 ~:~
:e;~'
Oe,~:'~J
~O
~
~'~)- ^
Lr
~ e~ ~~~~
0_~r~~
e~ (~:: c~~
,J~O~;,
e'~j~~
,D~l~3- ~:~~cead~SeS
~~~ uc:~L)"-
~~ou'~)'L)- ~H--- ~ ~ •- ~~
uo o
F~Qe"
ue'~S
~S~~3~:3
~SQ)e)FSvD c~~'
~)~:1
~~~:'~~3,~)~'
*v'~;
~)0~~
pJ:l~
~:31'~*
C'3eb~~~"~
h0~
c:~0
cl~(~]O
~)(~]\O
c:$~Lr]_~~

ce_ ~:~*
~~~;5 ~5 ~~~:' ~~5'- ~;=~~~_,,*c'SC/)~5C/)~5
~~ ~O
e'
~~l~c~l
~
~~:' ~~ ~s"~C~~~:? '"^'
~ c~~~
~~c~eJ)~0 c~
~F~eJ)?
~'~~),L)
~•*~O~ ~~:1,~J)
~ c~-~ c:~
~_ _,~o
'8
COclS(' ~:F:l .8(J~~~~ *
O ~~10~~:~~ ::8~'g:~~r
~L~~a~ e.'-
cq
'

'~L'~~s
e'c'~::
c~

e'~)
~)

eb~(:) eb~(~y~:'
*_(~;~~ ~1~~ ebO~)
rhl e:: 1~~)
-~•o-~)
eJ)Hc~
•
'~~o
eOFse'm
'- FS
O 1~~~ ~: '~ o ~:' .- O :S e. hl
~:clD~:;-~)
~l(/)C'I- C')
- -- ~1*J:~~- ~E1(')~~~t-

O e'
~ '-
~~oc')
e)

\ ~ O
O~
c~o
~s_~~
c!)~:t* eJ)
~:"H~~
F~~3e)
c:Sc~c~H

~~
~:s_~~ ~l
>
~e) eJ)

L~O ~') O* >1~:' ~:

eJ)~
~ ~~~F~'~)oc:'
c;~ ~ (D
8 .-
~F~*~5
Vv (~~
,t:: ~~
~~~{-
CQCQCQ
~(b oll~r(~~i
1~~~~~h O
O~i _ ~~s8 - ~~ ';S e,C'H_.~1
vV (e
L~O ~. -7 O) L ~coF:~~;-
(D
c'

-;:::o
vo
t~~O *~)FS
"~,**

a- 11 ~
E ~) o~i~~-H
a)
~~ Cr~;i
e"~~
~5

~
~sh~~~;
v
(D c"OO
~ e o(~~ ~boG~~) O
O
eo~'
-Q)'~;::oo~) ~(L)~:i
~~~O
~:,
F~
(1)

ooe ~~r' ~1
l oo c oasQ,c;v'
:=
~5
:s~!~J)- .-
v)
~SF~c:1>cl)"~~
(~ O(L)~
'~ ~; O
~::_O.e)

OCC~0~~~~ e)O:S
Cs J: ~::

~
o
o
('1
I
rO
\
~ ~;~~
O
t~:c:lF:
O O
\i
~;i
oo o
co o
~ o
(O o
Lo o
~r o
co o a o 5b
c\I
t:~2
C'
r~
eJ)~O
~ .F;
~:,q,
_O~C~~
eb,~e'

o
oo
~::1~:1
(UJ
~)5U!lIOJ ~:
eJo~eq sseu~o!V~ ell30s '-L~~
O O
~I~~IOh- *

tt
~5
-c~
-c~
oo ~
E
:L
e)
~t::~~c:$
,~o'f~
h~'
\ z
cy
~>
o~~
~ :L
~~
eo
~~
oo
o~,
'~ '~ \:
o\
O
LO~ ~ ooo .c:
a)
~~(-
~'
~\ -\l
c: : o-
(O Q)

~o tt (' O ~)wcrl
!:':o~
$::~:i ~I oL~~o*
~C'
e'o
- 'o
~~~
_~'o o
~f ~
* ~:o~,~~
(D
~~ o~r~2
(D
:e~~r~r -~)~2 o
~Q)oc~
(L)hl--
~~ ~
~~:;
o'e)'L)
~c:t

0
o
co
H-
Q)
~:2
u)
(,)
*-~
~~::o~"
e$ ::~~~
c,u'> ~)
~F~~3~i
~
-oo
\\ l

~>
o~~
co
(D

(1)

8~~0~~~!~
L,~:~
-
~5~5
O \rtcoo
,~)
o
C~J~x:
q)c:
oov'o
eJo'~ ~:i O o
(D
c:
~~ ti
-F~ o~) \ c\lJ~:
hcL) ~)~
o~~O o f:~
!t :sl~~~ \CK) e';:~-~)
('~
o*~~- j~~ :;=
F:: ~~i
~:1
~:_.~s~ o:Z~
~f 0 (D
(~io I'i
Q)
(~
,eb
~: a)
(*~)
c\I

Al!o!l~LVoJyO
- o O (/) eb
~: ~ co (\l

(*~) A;!o!~~vJoJyO
- o~
889 C 1995 ISIJ
 ISIJ International, Vol. 35 (1 995). No. 7

Heatingtime O:60minA:120min
iOO AO~;P~~~:~_+
4 A:in Air
~~:
~>:in N=
o
CQ
80
/
/ A ~
~,3
\
~~
/ \
60
A~~A/A
cO
Q) / \A ~5
co
! \ Cls2
:i~
,
O)
CQ 40 E
o
(,)
l
l ~
~)
(D
CC
20
f
I
o
\o_ ol
~:

--~> -----
A
o
1150 1200 1250 1300 O 0.2 0.4 0.6 0.8 1.O
Heating temperature ('C)
Scale poWderWeight (mg/cm2)
Fig. 13. Effects ofheating temperature and time on red scale
Fig. 14. Effect ofscale powderweight and cooling atmosphere
area ratio of high Si steel. (Test 4)
of quenching after rolling at 900'C on value of
time is shortened, the ratio of the red scale chromaticity (a*). (Test l, 2)
is less at high
temperature .

4. Discussion

Result of the present tests are summarizedas follows.

(1) The red scale is formed only whensteel slabs with
scale thicker than 20 plm is hot-rolled at below 900'C,
followed by water spray cooling in air. It is not dependent Fig. 15. Cross section SEMmicrographs oflow Si steel roned
(b) I OOO'C.(Test l)
on the Si content in steel. at (a) 750"c or
(2) The red scale is not formed when water spray
cooling is conducted in N2 atmosphere. than 2I/m.
In the present experiment, red scale powders
(3) Whendescaling is conducted by high pressure were formed on the surface of the steel strip when it
water jet for secondary scale (40-60 ~m)which is formed appeared red. Therefore the powderweight wasmeasured
by air cooling (40 s) prior to hot rolling, red scale is not by extracting the powders from the steel by ultrasonic
formed on steel irrespective of Si content. It implies that cleaning. Relation between the powder weight and a*
the scale is entirely removed by high pressure jet prior value for low Si steel is shownin Fig. 14. In case of air
to hot rolling. cooling, a* value increased with the weight of powder.
(4) When descaling is conducted for the thick This confirms that the cause of red color is the scale
primary scale formed by long time oxidation in the powder quantity. Onthe other hand, in case of cooling
in N2, a* value
heating furnace, red scale is formed only on high Si steel was not influenced by the powder
after hot rolling. quantity.
(5) The red scale coverage for this high Si steel is According to the X-ray diffraction examination, these
larger whenthe heating time is longer. As for the heating scale powders were Hematite (c(-Fe203) in case of air
temperature, the red scale coverage reaches maximum cooling, while they were wtistite (FeO) and Magnetite
at between I 180 and 1240'C and reduces at about (Fe304) in case of cooling in N2' It is knownthat the
scale formed at high temperature is composedof three
l 260'C.
From these results, following two features of red scale distinct layers of Fe/FeO/Fe304/c(-Fe203 and during
formation derived. them the amountof FeOis the greatest. Whenrolling is

a) The red scale formation is influenced by the scale carried out, a part of the scale is broken into powdery
thickness prior to hot rolling, the rolling temperature FeOand Fe304, and which becomered powdery Fe203
and oxidation during cooling, regardless of Si content by oxidation during air cooling. This is the assumed
in steel. mechanism red scale formation.
for
b) As the characteristic phenomenon ofhigh Si steel, 4.2. Effect of Rolling Temperature
primary thick scale formed during heating is not com-
pletely removedby a high pressure jet descaler and this SEM cross section photographs of the pass hot rolled 3
steel are shown in Fig. 15. The scale thickness before
causes the red scale after hot ro]ling. Surface coverage rolling was about 30 ktm and rolling temperature
of red scale is influenced by the slab heating temperature was750
and time. The formation mechanismis discussed as I
or OOO'C.Whenrolled at 750'C, the scale is broken to
follows. I
powders. While, whenrolled at OOO'C,the breakage of
powders is not observed. Ratio of scale powder weight
4.1. Relation between the Scale Powder and the Red measuredafter ultrasonlc cleaning to total scale weight
Scale was calculated for steels rolled at various temperatures.
Previously Asai et al.7) investigated relation between As shownin Fig. 16, the ratio of scale powders decrease
the color of the steel surface and the scale powder and as the rolling temperature rises. Previously Matsunoand
showed that the surface of the steel strip appears red others8) reported that yield strengih of FeOmeasured
when Hematite (oc-Fe203) powder grain size is smaller by 3point bending methoddecreased as the temperature

C 1995 ISIJ 890

 ISIJ International. Vol. 35 (1995). No, 7
12 the descaler, but also descaling position and temperature
o
must be considered, for example, a strong descaling at
_10 o
~~
6) 1
above 173'C for complete removal of Fayalite is one
o~~ 8 of the measureto prevent red scale.
~::v
cri(1) o O
t~ 6
(D(!)
(5)

O
~:'~~
~~
~c~
Q)~~
4 O S ~
~)
(~ O
6. Conclusion

The red scale defects, observed in hot rolled steel

cov
c'(1)_ o C~~ including more than 0.5 Si,
were reproduced in the
o/*
C/) 2 ~) laboratory and the rolling conditions and mechanismof
O700 their formation were examined.
800 900 ~OOO (1) The red scale is formed both in high Si (0.54 olo
Roiling temperature ('C)

Fig.
Si) and low Si (0.005 o/o Si) steels. It is formed whenthe
16. Effect of rolling temperature on the ratio of scale
scale thickness before rolling is thicker than 20 ,Im,
powder to total scale weight.
tandem rolled at below 900'C, and then followed by
cooled by water spray.
rose. Fromthis report, it considered that the scale is
is

soft and not broken into The red scale consists of red powdery ee-Fe203
(2)
powders during
rolling at high
(Hematite). It is thought that the Hematite is formed by
temperature. This maybe the reason why the scale did
oxidization during cooling of FeOand Fe304powdered
not turn red whenrolled at OOO'Cirrespective of the I by rolling below 900'C. The powdering is attributed to
scale thickness. On the other hand, when rolled below
be hardening of FeOat lower temperature.
900'C, the following two factors might have caused the
constant criteria of scale thickness (20 pm) before rolling
(3) Thesecondary scale formed by short time holding
in air did not cause red scale, regardless of Si content.
to from red scale; one is that decrease temperature
increased powdering of the scale and another is that it
That is because the secondary scale is easily removedby
retarded the oxidation of FeOto form Fe203 during water jet descaling before rolling.
rolling and cooling.
(4) On the contrary, the thick scale formed during
heating for long time in an atmosphere of commercially
4.3. Formation of RedScale on High Si Steel operating heating furnace caused red scale in high Si
steel. That is because 40 to 180
As shownin the result of Test 3(Fig. 8), whendescaling
after
pmthick scale remains
by high pressure water jet wasconducted with an interval even water jet descaling because of the strong
between air cooling and rolling, no red scale appeared bonding effect of Fe2Si04 on the sheet. This thick scale
regardless of Si content in steel. However, when the before hot rolling causes red scale.
primary thick scale wasformed during long time heating (5) As for high Si steel, the red scale area ratio reaches
in the furnace in Test 4, the scale becamered in high maximum whenthe heating temperature is between 1180
Si steel, in spite of the descaling
process. From the and 1240'C and decrease whenit is higher than 1260'C.
examination of the scale thickness after rolling, it is (6) Measuresto prevent red scale in commercial hot
thought that, even after descaling process, about 40 to strip mill are proposed. In low Si steel, consistent
180 ,lrn thick scale of FeOremained before rolling by the descaling between rolling passes and finishing at high
strong bonding effect of Fayalite (Fe2Si04) with steel in temperature are effective. In high Si steel, removal of
high Si steel. Whenthe heating temperature rose above Fayalite at high temperature descaling is an additional
1260'C, the red scale area ratio decreased and the surface important factor to prevent red scale problem.
becamefreeform patterns. The irregularly striped red
scale pattern usually observed in commercial production REFERENCES
of Si steel is considered to be equivalent to this condition. l) D. T. Btazevic: Proc, of4th. Jnt. steel Ronconf, vol. 1, (1987),
The freeform patterns of the red scale is thought to be A.38.1.
the result of heterogeneous remGVal of the scale by 2) Y. Ishi, A, Kodoi and I. Wakamatsu:Mech. WorkSteel Process,
descaling at the temperature above the melting point of 30 (1992), 447.
3) Y. Okita. A. Nagai, I.Sinagawaand K. Horinouchi: CAMP-ISIJ,
Fe2Si04 (1 173'C). The details were reported in another
2(1989), 1509.
paper.9) 4) K. Onoda, H. Okabe, R. Nomi and N. Hase: SumitomoMet.,
20 (1968), 1.
5) T. Sasaki, A. Shinjo, K. Kunishige, Y. Okita and T. Ebukuro:
5. Measures to Prevent from the Red Scale during
Commercial Production Process SumitomoMet., 44 (1992), 23.
6) F. Matsuno: ISIJlnt.,20 (1980), 413.
As for low Si steels, consistent descaling during hot 7) T. Asai, T. Nakamuraand T. Inoue: CAMP-ISIJ, (1993), 357.
6
8) F. Matsunoand S. Nishikida: Tetsu-to-Hagan~, 71 (1985), S1282.
rolling passes and finish at high temperature are
rolling
9) T. Fukagawa. H. Okadaand Y. Maehara: ISIJ Int., 34 (1994),
very important factors to prevent red scale problem. The 906.
oxidation after rolling is inevitable, while, in high Si
steels, compl~te descaling of thick scale formed during (Originally published in Tetsu-to-Hagan~, 80 (1994): 849, in Japanese)
slab heating Not only
is
more important. strengthening

891 C 1995 ISIJ