JMEPEG ÓASM International

https://doi.org/10.1007/s11665-018-3664-4 1059-9495/$19.00

Effect of Scale Spallation During Coiling

on the Electrochemical and Pickling Behavior
of a Hot-Rolled Dual-Phase Steel
S. Choudhary, K. Mondal, S. Mukherjee, and Sundara Bharathy R.

(Submitted April 6, 2018; in revised form June 25, 2018)

Tertiary oxide scale consists of magnetite and hematite, and in dual-phase steel strips were observed after
the exit of final rolling stand in a thin slab casting rolling (TSCR) mill. Microscopic observation confirmed
that these scales cracked and spalled off from many places of surface during coiling causing a non-uniform
oxide layer on the steel surface. The present work discusses the synergistic effect of such non-uniform oxide
layer and run-out-table cooling water on the electrochemical behavior of the dual-phase steel. The elec-
trochemical response was assessed with the help of dynamic and linear polarization tests considering
cooling water as electrolyte. The presence of non-uniform scale produced high cathodic (scaled zone) to
anodic (spalled-off zone) area ratio that eventually led to extensive pitting corrosion in the present steel.
Furthermore, the pickling behavior of the steel was also studied in detail. It was observed that the steel
surface, with spalled-off area of about one-tenth of total surface area, corroded preferentially 3-4 times
faster at spalled-off area than that of the uniformly pickled surface. The characterization study was
performed using scanning electron microscopy, confocal laser 3D scanning microscopy, x-ray diffraction
and Raman spectroscopy in order to establish the reasons for pitting corrosion due to the scale spallation.

promotes the formation of secondary oxide scale. Descaling of

Keywords corrosion, electron microscope, steel, tertiary oxide
scale sheet is again carried out before entering to the finishing mill.
However, complete removal of scale is not possible before
coiling as the strip is still exposed to high temperature and
water after finishing mill. Thus, it reacts with oxygen and water
vapor and generates tertiary oxide scale on the surface of the
steel sheet. Researchers reported that thickness of secondary as
1. Introduction well as tertiary oxide scale is around 10-15 lm (Ref 5, 6).
Tertiary oxide scale is adhesive to the surface even after the
During hot rolling of steel, slabs are normally reheated to a cooling down of the coil to room temperature. Therefore, final
temperature range of about 1100-1300 °C. The hot slab is then product of the hot rolling process is the steel sheet covered with
transferred to roughing mill followed by finishing mill, where it thin tertiary oxide scale layer.
is deformed at temperature around 1050-850 °C. Subsequently, The tertiary oxide scale formed at the exit of the finishing
the strip is cooled in run-out-table (ROT) using laminar water mill comprises of three different oxide layers, namely a thick
jet and finally coiled at a temperature range of 700-200 °C wustite (FeO) layer adjacent to the steel substrate, an interme-
depending on the requirement of desired properties (Ref 1). diate magnetite (Fe3O4) layer and a thin outermost hematite
Inevitably, oxide scale is generated on the surface of strip (Fe2O3) layer (Ref 2). However, wustite is unstable below
during various stages of hot rolling due to synergistic effect of 570 °C and it decomposes to magnetite with fine a-iron
high temperature, the presence of atmospheric oxygen and particles (Ref 7).
water vapor (Ref 2). Surface quality and downstream process-
ing of the hot-rolled coil, such as acid pickling, forming, 4FeO ) a-Fe þ Fe3 O4
coating, largely depend on the mechanical and electrochemical This transformation increases the scale adhesiveness to the
properties of the scale (Ref 3). Initially, primary oxide scale steel substrate (Ref 8). However, it is not possible to avoid
with thickness of about 2-6 mm is generated during reheating cracking/spalling-off tendency of scales, when they are sub-
stage (Ref 4). Descaling of the slab is carried out by a high- jected to bending during coiling, due to strain mismatch
pressure water jet just before its entrance to roughing mill. between scale layer and the steel substrate. Later on, pickling
Furthermore, during successive reduction in roughing stand, performance would also depend on the occasional cracking/
hot metal sheet is exposed to work roll cooling water, which spalling-off of the oxide cover formed during hot rolling
process. The outer Fe2O3 layer is known to dissolve only
slowly in acid, while the lower FeO and Fe3O4 layers dissolve
S. Choudhary and S. Mukherjee, Research and Development, Tata faster in comparison (Ref 9). Therefore, scale adhesiveness also
Steel Limited, Jamshedpur 831001, India; K. Mondal, Department of
depends on the volume fraction of different constituents in the
Materials Science and Engineering, IIT Kanpur, Kanpur 208016, India;
Sundara Bharathy R., Flat Product Technology Group, Tata Steel scale.
Limited, Jamshedpur 831001, India. Contact e-mail: Sanjay. Scale adhesiveness is one of the critical properties of the
Choudhary@monash.edu. hot-rolled dual-phase steel products. According to world

Journal of Materials Engineering and Performance

autosteel, majority of autobody parts is designed using dual- scale characterization and electrochemical experimentation, and
phase steels (Ref 10). Moreover, many components, such as the coupons with dimension 20 mm 9 10 mm 9 3.9 mm were
body of an excavator, a chassis of a pick-up truck, are made used for cross-sectional microstructural analysis. All the
with the dual-phase steel sheet without removal of oxide scale. specimens were carefully immersed in acetone to remove oil,
A proper bonding between the scale and the steel is necessary grease and dust on the surface followed by distilled water
in these applications. On the other side, poor scale adhesion is immersion and drying. Run-out-table (ROT) water was col-
required for cold-rolled and coated dual-phase steels, which are lected for conducting the electrochemical experiments of the
used for cabin structural parts. In these applications, hot-rolled samples. Initially, chemical analysis of the water was carried
coil is sent for pickling followed by cold rolling (Ref 11). out to determine pH and concentration of Cl and SO2 4 ions.
However, non-uniform scale on the surface during acid pickling Chloride ion concentration was determined using VolhardÕs
may lead to excess removal of metal from many places on the method (Ref 17), while inductively coupled plasma-atomic
surface (Ref 12). Similar localized corrosion attack due to scale emission spectrometry (ICP-AES) was used to determine the
spallation is also possible for many hot-rolled coils, which are sulfate ion concentration (Ref 18). VolhardÕs method used the
directly used without pickling. Hot-rolled dual-phase steel is back-titration with titanium thiocyanate to determine the
coiled at the temperature below 300 °C. Low-temperature chloride concentration in a solution.
coiling leads to incomplete removal of water from the surface Microstructures of the rolling surface as well as the cross
of the steel sheet as temperature varies from 300 °C to below section were analyzed using scanning electron microscope
100 °C (normal water vaporization point). Coiling temperature (Quanta FEG 650). Energy-dispersive spectroscopy analysis of
above the vaporization point can also lead to incomplete the oxide scale was carried out to identify the morphology and
removal of water as water vapor in inner wrap will be possible composition of different oxide layers across the scale
condensed below 100 °C. This water will act as an electrolyte thickness. The scale was also characterized using grazing angle
and can lead to localized corrosion attack on the surface due to x-ray diffraction and Raman spectroscopy. X-ray diffraction
scale spallation. was carried out using XÕPert PRO PANalytical system with a
While scale spallation mechanism has been extensively step size of 0.016°/s and changing the grazing angle from 0.5°
studied in the context of strain mismatch between scale and the to 20° to investigate the structure of the oxide layer in the scale.
steel substrate (Ref 13-16), the studies on the electrochemical Raman spectroscopy was carried out in an Acton SpectraPro
behavior of steel surface due to scale spallation would also have SP-2500 system. Scale surface was excited with the 532-nm
huge practical implication and importance. However, it is not Nd: YAG laser. Raman intensity ratio could also be used to
studied in detail in the open literature. Hence, it is a necessity to calculate the concentration of the components in a sample.
study the scale protectiveness in run-out-table water environ- However, this method might be prone to errors as the intensities
ment due to scale spallation and concurrent effect on the are affected by various experimental factors, such as laser
electrochemical behavior of the surface. In addition, the present power, sample homogeneity, sample positioning (Ref 19).
work analyzes the effect of cracked/spalled oxide layer at Electrochemical tests were carried out using a PARSTAT
localized regions of a hot-rolled dual-phase steel (DP600 steel 2263 system in a flat-bottom cell consisting of a specimen, a
sheet) on the pickling of the scale and surface appearance. saturated calomel electrode with saturated calomel
(E° = + 0.2444 V versus standard hydrogen electrode (SHE))
and a platinum wire mesh as working, reference and counter
electrode, respectively. All the electrochemical experiments
2. Experimentation were performed at room temperature ( 30 °C) in 300 ml ROT
water mentioned above. Initially, open-circuit potential (OCP)
The material under investigation was a hot-rolled dual-phase was measured after 1 h of its stabilization, followed by linear
steel with the tensile strength of 600 MPa. Chemical compo- polarization and potentiodynamic polarization. Linear polar-
sition and mechanical properties are given in Table 1. This hot- ization was carried out between electrode potential rang-
rolled dual-phase steel was produced in TSCR (thin slab casting ing ± 20 mV versus OCP at the scan rate of 0.166 mV/s.
and rolling) mill at Tata Steel Limited, Jamshedpur (India). However, potentiodynamic polarization scan was done from
Thin slab with thickness of around 60 mm was reheated at  250 mV versus OCP to 1.6 V at the scan rate of 0.5 mV/s.
1200 °C for 30 min followed by hot rolling. Roughing mill All the samples were tested without disturbing the scale layer.
entering temperature and finishing mill temperature were 1015 Thereafter, the same electrochemical experimentation was done
and 815 °C, respectively. Coiling was done at temperature with the acid pickled sample to compare the electrochemical
range of around 200-250 °C with the final thickness of 3.9 mm. importance of scale spallation with completely descaled
Samples were collected just after the coiling to minimize the surface. Pickling process involved passing the scaled surface
primary mechanical and electrochemical damage. Test speci- from five different acid baths having increasing hydrochloric
mens were cut into 20 mm 9 20 mm 9 3.9 mm coupons for acid concentration (6, 9, 12, 14 and 16 wt.%) followed by

Table 1 Chemical composition and mechanical properties of dual-phase steel

Chemical composition, wt.% Mechanical properties

C Mn Si Al P V Nb Cr YS, MPa UTS, MPa %El

0.053 1.173 0.03 0.034 0.022 0.002 0.024 0.704 453 660 23

Journal of Materials Engineering and Performance

water rinsing. It was ensured that sample was immersed for 10- thermal expansion with respect to steel (Ref 13-16). This results
14 s in each bath, while the baths were held at 80°C. Just after in compressive shear through-thickness cracks in oxide scale,
all the electrochemical measurements, surfaces were analyzed leading to scale spallation during coiling. Scale thickness has
using SEM and confocal laser 3D scanning microscope (OLS been measured from the cross-sectional image (Fig. 1b) and
4000). found to be uniform ( 10 lm).
Figure 1(c) shows that phase fraction of martensite near the
surface ( 50 lm depth) is substantially low. However,
martensite is uniformly distributed at the center with phase
3. Results and Discussion fraction of around  15% (Fig. 1d). This microstructural
gradient evolution could be attributed to the decarburization
Chloride concentration, sulfate concentration and pH of the (Ref 24-26) and smaller prior austenite grain size (Ref 27) at the
ROT water were 30 ppm, 91.4 ppm and 8.9, respectively. surface of the steel during hot rolling. It is well known that
Although water is basic in nature, pH 8.9 is not enough to form decarburization is unavoidable during high-temperature rolling.
a stable protective oxide layer on the surface (Ref 20, 21). Although this kind of microstructure gradient during hot rolling
Researchers have reported that most waters have less than is not reported in open literature till date, many researchers
50 ppm chloride; however, concentration as low as 3 ppm has observed similar microstructural evolution due to depletion of
an immediate effect on corrosion of steel (Ref 22). Further- carbon during decarburization. Jing et al. (Ref 25) have studied
more, Niu et al. (Ref 23) have found that 100 ppm sulfate ions the decarburization mechanism during hydrogen reduction
have adverse effect on pitting. At higher potential, pit growth descaling of hot-rolled strip steel. They have observed equiaxed
rate is mainly controlled by diffusion of cations (Fe2+). While ferritic grain with depletion of pearlite near the surface due to
diffusing out from inside the pit, these cations interact with decarburization. Castaneda et al. (Ref 26) have also reported
anions (Cl, SO42) and form a salt cover over the pit mouth. similar microstructural changes of hot-rolled steel due to
As a result, pit growth rate decreases due to suppression of the decarburization.
cation diffusion. However, an increase in concentration of Decarburization leads to the diffusion of free carbon from
SO42 from 50 to 100 ppm (in the presence of 100 ppm Cl) the bulk to the surface, which subsequently reacts with the
increases the solubility of the salt cover. This further leads to wustite (FeO) at high temperature and forms CO (Ref 11). This
the fracture of salt film, facilitating the diffusion of metal gas layer at the interface between scale and the steel substrate
cations and accelerating the rate of pit growth. Therefore, helps in descaling. On the other hand, surface layer of steel is
present water could adversely affect the pitting behavior of the subjected to shear force and roll chilling effect leading to
steel surface. reduction of the prior austenite grain size in the surface region
Figure 1 shows the SEM micrographs of the rolling surface (Ref 27). Therefore, faster kinetics of ferrite formation at the
(Fig. 1a) and cross section of the sheet (Fig. 1b, c and d). surface could lead to lower amount of martensite. However,
Figure 1(a) reveals that the tertiary scale is discontinuous detailed study is required to establish the exact reasons behind
throughout the surface and it has been chipped off from many this microstructural gradient at the surface. This microstructural
locations on the surface (as shown by arrows). Oxide scale is variation leads to undesirable electrochemical and mechanical
under compression during cooling in ROT due to differential properties of the surface (Ref 28).

Fig. 1 SEM micrographs of hot-rolled dual-phase steel after coiling: (a) rolling surface, (b) cross section: oxide scale, (c) full cross section:
decarburization near the surface, (d) cross section: microstructure of dual-phase steel

Journal of Materials Engineering and Performance

 The scale at the surface comprises of different iron oxide making it impossible to give the bulk data (Ref 33). Therefore,
layers, i.e., wustite (inner layer), magnetite (intermediate) and data obtained from Raman spectroscopy mainly confirm the
hematite (outer most layer). Although wustite is not stable be- outer layer as hematite.
low 570 °C, it transforms into magnetite and free a-Fe particles Figure 5 shows the linear polarization (a) and potentiody-
making magnetite as inner most layer (Ref 7). This layering namic polarization curves (b) of the samples with scale and
phenomenon can be confirmed from the elemental mass pickled surface, respectively, in ROT water environment. It is
fraction analysis using SEM–EDS (Fig. 2). Mass fraction of important to mention that the analysis
 ofthe pickling has been
Fe has been found out to be around 73 and 70 mass% in the mentioned in later sections. Slope dðDEÞ
at zero overpotential
inner and the outer scale layers, respectively. Theoretically, di
magnetite and hematite contain 73 and 70 mass% Fe, respec- (or zero current) gives the linear polarization resistance (Rp),
tively. Average thickness of the magnetite layer has been found and it is inversely related to the corrosion current density (Ref
out to be 8 lm, while that of the hematite layer is 2 lm thick. 34).
However, hematite layer was not compact and continuous. It is dðDEÞ
very clear that the hematite layer has wide cracks, which could ¼ Rp
di DE¼0
act as sites for electrolyte access. Man et al. (Ref 29) have
analyzed the oxide scale with the help of focused ion beam It is important to mention that linear polarization test was
(FIB). They have also reported similar scale morphology carried out after 1 h wait for the requirement of stabilization of
including microcracks. These cracks are generated during OCP. At stable OCP, both cathodic and anodic reaction rates or
cooling due to surface stresses caused by differential thermal cathodic and anodic current densities are the same as per the
expansion coefficient between different oxides within the oxide principle of mixed potential theory. Therefore, net current is
scale. zero at OCP. This linear plot is obtained when a small
Grazing angle XRD result also confirms the presence of overvoltage is applied. The slope dðDEÞ
di was calculated by fitting
magnetite and hematite (Fig. 3). Figure 3(a) shows the XRD straight line through a set of 5-7 data points including (0, 0) as
peaks of the scale with increasing grazing angle from 0.5° to the center point. Rp is the resistance for the charge transfer
20°. Grazing angle is directly related to the depth of penetra- between sample surface and the electrolyte. However, corrosion
tion, and therefore, grazing angle can be increased to analyze current density cannot be determined only using linear
inner most layer (Ref 30, 31). Intensity of the peaks of polarization method.
magnetite increases with increasing grazing angle (Fig. 3b) Tafel extrapolation method (ASTM G3-14) was used to
confirming the result of SEM–EDS. However, quantitative measure the corrosion current density and corresponding
analysis from XRD analysis is difficult due to layered structure corrosion potential. FaradayÕs law (ASTM G102-89) has been
of the scale as fraction of scale constituent is changed with used to calculate the corrosion rate:
depth of penetration. Figure 4 shows the Raman spectrum of
the steel, and it shows mainly hematite peaks and few lower 3:27  icorr  E:W:
Corrosion Rate ðmm/yearÞ ¼
intensity peaks have been identified as magnetite (Ref 32). This q
is due to the low penetration depth of laser because low-
where icorr is the corrosion current density (mA/cm2), E.W. is
intensity 532-nm laser has been used. Researchers have shown
equivalent weight in gram (= 27.92 for carbon steel), and q
that 532-nm laser reaches only up to around 0.7 lm depth

Fig. 2 SEM-EDS analysis of scale layers

Journal of Materials Engineering and Performance

Fig. 3 (a) Grazing angle XRD spectrums of the scale layer and (b) magnified peak in the range of 2h = 34°-37° (H hematite, M magnetite)

Fig. 4 Raman spectrum of the scale layer (H hematite, M magnetite)

denotes the density of the material in g/cm3 (= 7.85 for carbon tively. It is important to mention that Fig. 6(a) and (b) is not the
steel). Values of the electrochemical parameters for the samples same spot as it is difficult to identify the same spot after
with the scale as well as pickled surface are given in Table 2. electrochemical test. However, more than three spots were
Linear polarization resistance with the scale surface is analyzed in both the cases for sound comparison. Line profiles
higher than that of the pickled surface (Fig. 5a and Table 2). show that depth of the spalled-off zone from scaled surface
This is the indication of the surface protectiveness due to the increased from around  10 to  60 lm after potentiody-
scale formation. This can also be seen from the potentiody- namic polarization test.
namic polarization results that the corrosion rate of the pickled However, surface covered with scales is unaffected. It shows
surface is 3-4 times higher than that of the sample with scale that the localized spalled zones are electrochemically more
surface (Fig. 5b and Table 2). However, the sample with scale active than the scaled area. It also reflects that the dissolution is
surface gets intensely pitted only from spalled-off areas. only concentrated in the spalled-off zone, increasing the
Confocal 3D micrographs with the scale surface, before and possibility of intense pitting on these regions. Confocal 3D
after polarization test, are shown in Fig. 6(a) and (b), respec- micrographs of pickled surface, before and after polarization

Journal of Materials Engineering and Performance

Fig. 5 (a) Linear polarization and (b) potentiodynamic polarization curves with scale and pickled surface

Table 2 Electrochemical properties with scale and pickled surface from the linear and potentiodynamic polarization
curves
Electrochemical properties With scale surface Pickled surface

Linear polarization resistance, X-cm2 9923 2750

Corrosion potential, V  0.415  0.517
Corrosion rate, mm/year 0.029 0.119

Fig. 6 Confocal 3D scanning micrographs of steel surface with scale surface: (a) before potentiodynamic, (b) after potentiodynamic
polarization (shown values are in lm)

Journal of Materials Engineering and Performance

Fig. 7 Confocal 3D scanning micrographs of steel surface with pickled surface: (a) before potentiodynamic, (b) after potentiodynamic
polarization (shown values are in lm)

test, are shown in Fig. 7(a) and (b), respectively. Similar to

Fig. 6, Fig. 7(a) and (b) does not reflect the same spot on the
surface. However, the same sample surface was analyzed
before and after polarization test. Minimum 5 readings were
taken to get the average roughness value. Pickled surface shows
roughness of 1.16 ± 0.16 lm with few shallow points (arrow-
heads). Surface roughness increased to 1.29 ± 0.32 lm after
potentiodynamic polarization test. Depth of few locations has
also increased insignificantly to around 10 lm (arrowheads).
This is the indication of uniform corrosion after pickling as
surface dissolves uniformly. Uniform corrosion of pickled
surface has resulted in higher material loss than that of scaled
surface as corrosion current density is much higher for pickled
surface (Fig. 5b).
The current density used for calculation of corrosion rate
listed in Table 2 represents the current with respect to whole
exposed area during electrochemical tests. However, for scaled
surface, dissolution is concentrated only at the spalled area. Fig. 8 Corrosion rate on the steel surface with scale (WS), pickled
Therefore, current density should be recalculated with respect steel surface and estimated pitting growth rate of steel surface with
to the spalled area in order to determine the ‘‘dissolution rate scaled calculated by Tafel plot
for spalled area’’ or ‘‘pit growth rate.’’
Pit growth rate ¼ calculated corrosion rate captured in Fig. 9(a, b, c, d, e and f). Although micrographs
Whole exposed area shown in Fig. 9 do not show the same spot, micrographs of
 different spots have been added to explain the pitting growth
Spalled off area
mechanism. Figure 9(a) shows the initiation of pitting, which
As shown in Fig. 1(a), approximately one-tenth of total further grows within the spalled-off area (Fig. 9b and c). When
surface area has been spalled off. Therefore, pit growth rate due electrolyte inside the pit becomes saturated with metal cations,
to the scale spallation is approximately ten times higher than crevice corrosion takes place just in the surrounding region
overall corrosion rate with scale surface and 3-4 times higher followed by spallation (Fig. 9d). This provides more space for
than that of pickled surface as shown in Fig. 8. the growth of the pits horizontally (Fig. 9b). Chloride ions in
Nucleation of a pit on the chipped-off area followed by the electrolyte result in vertical growth of any pit, while poor
growth during the course of potentiodynamic polarization is scale adhesiveness leads to horizontal growth (Ref 35).

Journal of Materials Engineering and Performance

Fig. 9 Growth of a pit on steel surface with scale surface during potentiodynamic polarization: (a) nucleation, (b) growth, (c) complete
dissolution of spalled-off area, (d, e) crevice corrosion followed by scale spallation just in the surrounding area, (f) cross-sectional view of
crevice corrosion

Horizontal growth of the pits also allows the electrolyte to (i) Oxidation of scale spalled-off steel surface:
come in contact with subsurface magnetite layer, which are
cathodic sites for cathodic reactions. When surface with scale Fe ) Fe2þ þ 2e
spalled-off area comes in contact with any electrolyte contain- (ii) Combination and hydrolysis reactions
ing oxidizing ions, an electrochemical cell would form due to
the potential difference between spalled-off region and the Fe2+ ions would form FeCl2 and FeSO4 salts, which
scaled area. subsequently hydrolyze to Fe(OH)2 and H+Cl or 2H+SO42
Hematite is non-conductive and therefore will not partici- acids, increasing the pH in the pit region.
pate in the electrochemical reaction during pitting of scale
spalled-off zone. However, this outermost hematite layer is Fe2þ þ 2Cl ) FeCl2
porous, loose and discontinuous, which completes the electro-
chemical circuit between scale spalled-off zone and magnetite Fe2þ þ SO2
4 ) FeSO4
(Ref 36). In the present case also, there are segments, where FeCl2 þ 2H2 O ) FeðOHÞ2 þ 2Hþ 2Cl
oxide scale has spalled off (Fig. 1) as well as the outer hematite
FeSO4 þ 2H2 O ) FeðOHÞ2 þ 2Hþ SO2
4
oxide layer has also wide cracks (Fig. 2). Hence, non-
conducting nature of hematite is not going to prevent electro-
chemical reactions since there is already a contact between the (iii) Cathodic reactions.
subsurface cathodic magnetite with the metal surface via the
sipped in electrolyte. Therefore, electrochemical reactions in O2 þ 2H2 O þ 4e ) 4ðOHÞ
the presence of ROT water electrolyte resulting in concentrated
corrosion comprise of the following redox, combination and This reaction happens in the beginning of the electrochem-
hydrolysis reactions: ical process in ROT water. However, there is generation of H+

Journal of Materials Engineering and Performance

Fig. 10 Corrosion and pit growth mechanism for dual-phase steel after coiling

ions near the pit regions due to hydrolysis of the chloride and ratio and more number of reduction reactions (oxygen reduc-
sulfate salts, which would react with Fe3O4 to give another tion as well as reduction of Fe3O4 in the presence of increased
cathodic reaction, which would further accelerate the dissolu- H+ ions generated due to hydrolysis reaction inside the pit
tion of Fe from the pit region. region). A single oxidation reaction (Fe = Fe2+ + 2e) would
need to supply equivalent amount of electrons for the two
Fe3 O4 þ 8Hþ þ 2e ) 3Fe2þ þ 4H2 O cathodic reactions. Hence, rate of dissolution or growth of pit
Potential difference can also be seen from the potentiody- increases substantially in the exposed section.
namic plots as equilibrium potential of the pickled surface is Figure 10 shows the sequence of events and finally
more negative than that of the surface covered with scale generation of pits and remaining trace of pits after acid
(Fig. 5b). Therefore, descaled area is anodic with respect to pickling. Scale formed after finish rolling called tertiary scale
scaled area. Furthermore, higher cathode/anode area ratio was  10 lm thick. However, scale got cracked and chipped
makes anodic area to dissolve faster, as corrosion current off from many places due to the strain mismatch with steel
density increases with increasing cathode/anode area ratio (Ref substrate during bending during coiling. Decarburization could
37, 38). The larger the cathodic area compared with anodic also have been played a role in descaling as it forms CO gas
area, the more reduction reaction or cathodic reaction can occur bubble at the interface causing blistering (Ref 40). Dual-phase
and, therefore, greater will be the galvanic current. This area steels are processed by very low coiling temperature, and
effect causes the surface to get pitted in the presence of an therefore, quenching water does not vaporize completely from
electrolyte containing anions, such as chloride and sulfate. the surface during coiling. This water acts as an electrolyte and
Cation (Fe2+) concentration increases inside the pit due to makes the surface more susceptible to pitting corrosion due to
oxidation, and it attracts the chloride and sulfate ions to move scale spallation. Scale chipped-off area became anode with
inside the pit. It increases the acidity of electrolyte inside the respect to scaled area (mainly magnetite scale) and surface gets
pit, and therefore, dissolution aggravates and results in rapid pitted extremely due to high cathodic/anodic area ratio. Surface
vertical growth of any pit (Ref 39). Therefore, the present was pickled to eliminate the effect of high cathode/anode area
experimental evidence clearly points at the nucleation and ratio, and corrosion happens by uniform dissolution mode as
growth of pits being governed by high cathodic/anodic area observed in the present case. Therefore, after pickling results
ratio and the presence of chloride and sulfate ions in the show an increase in corrosion rate (Fig. 5b). However, pickled
electrolyte. Oxidizing agents (chloride, sulfate and dissolve surface has insignificant shallow impression than that of with
oxygen) present in the cooling water aggravate the oxidation the scale surface. Therefore, severity of pitting is much higher
reaction of the scale spalled-off zone, and subsequently, in case of non-uniform scaled surface than that of pickled
reduction rate of scale also increases. There are a number of surface. Thus, proper control over coiling temperature is
redox reactions, whereas there is one oxidation reaction of Fe required to avoid the incomplete water removal and subse-
forming soluble Fe2+ ions. There is combination of hydrolysis quently to avoid pitting for dual-phase steels. Moreover, a
reactions which would increase the acidity in the pit region systematic study of scale adhesiveness with respect to chemical
even if the ROT water is basic in nature. This follows basic composition and thermomechanical processing parameters is
mechanism of pit growth and Pourbaix diagram. Hence, pit also necessary to provide a protective atmosphere for the steel
growth is a combined action of large cathodic to anodic area sheets, which are directly used without pickling.

Journal of Materials Engineering and Performance

4. Conclusions 11. S. Chandra-ambhorn, T. Phadungwong, and K. Sirivedin, Effects of
Carbon and Coiling Temperature on the Adhesion of Thermal Oxide
Scales to Hot-Rolled Carbon Steels, Corros. Sci., 2017, 115, p 30–40
The present work has investigated the electrochemical 12. R.Y. Chen, C. Du, and W.Y.D. Yuen, Factors Affecting Over-Pickling
behavior of hot-rolled dual-phase steel in cooling water due of Low and Ultra-Low Carbon Steels and the Effects of Inhibitor
to scale spallation during coiling, and the following conclusions Additions of Medium Carbon Steels, SEAISI, Q. (South East Asia Iron
Steel Inst.), 2008, 37, p 17–27
are drawn from the above study: 13. J. Min Lee, W. Noh, D.J. Kim, and M.G. Lee, Spallation Analysis of
Oxide Scale on Low Carbon Steel, Mater. Sci. Eng. A., 2016, 676, p
1. Layered tertiary oxide scale of  10 lm thickness was 385–394
observed in the dual-phase steel under investigation. 14. X. Yu, Z. Jiang, J. Zhao, D. Wei, J. Zhou, C. Zhou, and Q. Huang,
2. EDS, grazing angle XRD and Raman spectroscopy con- Local Strain Analysis of the Tertiary Oxide Scale Formed on a Hot-
Rolled Steel Strip via EBSD, Surf. Coat. Technol., 2015, 277, p 151–
firmed that the layered structure consists of inner mag- 159
netite layer ( 8 mm thickness) and outer hematite layer 15. H.E. Evans, Cracking and Spalling of Protective Oxide Layers, Mater.
( 2 mm thickness). The outer hematite layer contains Sci. Eng. A, 1989, 120–121, p 139–146
many microcracks, which later act as access points for 16. X. Yu, D. Wei, X. Wang, and Z. Jiang, Experimental Study on
the electrolyte for an electrical contact between magnetite Adhesion of Oxide Scale on Hot-Rolled Steel Strip, Adv. Mater. Res.,
and metal surface. 2012, 472–475, p 622–625
17. M.J. Pribil, W.I. Ridley, and P. Emsbo, Sulfate and Sulfide Sulfur
3. SEM images confirmed that the tertiary scale got cracked Isotopes (d34S and d33S) Measured by Solution and Laser Ablation
and spalled off sporadically from the steel surface primar- MC-ICP-MS: An Enhanced Approach Using External Correction,
ily due to strain mismatch. Chem. Geol., 2015, 412, p 99–106
4. Lower ( 250 °C) coiling temperature during dual-phase 18. T. Rojanarata, K. Sumran, P. Nateetaweewat, W. Winotapun, S.
steel processing leads to incomplete removal of cooling Sukpisit, P. Opanasopit, and T. Ngawhirunpat, Microscale Chemistry-
Based Design of Eco-Friendly, Reagent-Saving and Efficient Pharma-
water. The non-uniform scale on the surface with high ceutical Analysis: A Miniaturized VolhardÕs Titration for the Assay of
cathodic (scaled zone) to anodic (spalled-off zone) area Sodium Chloride, Talanta., 2011, 85, p 1324–1329
ratio and the presence of an electrolyte (cooling water) 19. C.J. Strachan, T. Rades, K.C. Gordon, and J. Rantanen, Raman
leads to pitting corrosion of the dual-phase steels. Spectroscopy for Quantitative Analysis of Pharmaceutical Solids, J.
5. Scaled surface with spalled-off area of about one-tenth of Pharm. Pharmacol., 2017, 59, p 179–192
total surface area gets corroded preferentially at spalled-off 20. H. Tamura, The Role of Rusts in Corrosion and Corrosion Protection of
Iron and Steel, Corros. Sci., 2008, 50, p 1872–1883
area and 3-4 times slower than that of uniformly getting 21. C. Man, C. Dong, J. Linag, K. Xiao, Q. Yu, and X. Li, Characterization
corroded pickled surface. However, growth rate of pitting of the Passive Film and Corrosion of Martensitic AM355 Stainless
due to formation of microelectrochemical cell was 3-4 Steel, Anal. Lett., 2017, 50, p 1091–1111
times higher than that of corrosion rate of pickled surface. 22. L.C. Rowe and M.S. Walker, Effect of Mineral Impurities in Water on
the Corrosion of Aluminium and Steel, Corrosion, 1961, 7, p 353t–
356t
23. L.B. Niu and N. Kensuke, Effect of Chloride and Sulfate Ions in
Simulated Boiler Water on Pitting Corrosion Behavior of 13Cr Steel,
Corros. Sci., 2015, 96, p 171–177
References 24. R.Y. Chen and W.Y.D. Yuen, Review of the High-Temperature
Oxidation of Iron and Carbon Steels in Air or Oxygen, Oxid. Met.,
1. S.X. Zhou, An Integrated Model for Hot Rolling of Steel Strips, J. 2003, 59, p 433–468
Mater. Process. Technol., 2003, 134, p 338–351 25. Y.A. Jing, Y.M. Yuan, X.L. Yan, L. Zhang, and M.H. Sha,
2. W. Sun, A.K. Tieu, Z. Jiang, H. Zhu, and C. Lu, Oxide Scales Growth Decarburization Mechanism During Hydrogen Reduction Descaling
of Low-Carbon Steel at High Temperatures, J. Mater. Process. of Hot-Rolled Strip Steel, Int. J. Hydrogen Energy, 2017, 42, p
Technol., 2004, 155–156, p 1300–1306 10611–10621
3. Y. Jatuphaksamphan, N. Phinichka, K. Prapakorn, and M. Supradist, 26. E.G. Castaneda, A.S. Rodriguez, R.D. Lara, and F.M. Torres, High
Pickling Kinetics of Tertiary Oxide Scale Formed on Hot-Rolled Steel Temperature Oxidation and Its Effects on Microstructural Changes of
Strip, J. Met. Mater. Miner., 2010, 20, p 33–39 Hot-Rolled Low Carbon Non-oriented Electrical Steels During Air
4. Y. Yu and J.G. Lenard, Estimating the Resistance to Deformation of the Annealing, Oxid. Met., 2015, 83, p 237–252
Layer of Scale During Hot Rolling of Carbon Steel Strips, J. Mater. 27. M.R. Hickson, R.K. Gibbs, and P.D. Hodgson, The Effect of Chemistry
Process. Technol., 2002, 121, p 60–68 on the Formation of Ultrafine Ferrite in Steel, ISIJ Int., 1999, 39, p
5. X.L. Yu, Z.Y. Jiang, X.D. Wang, D.B. Wei, and Q. Yang, Effect of 1176–1180
Coiling Temperature on Oxide Scale of Hot-Rolled Strip, Adv. Mater. 28. X.J. Zhao, J. Guo, H.Y. Wang, Z.F. Wen, Q.Y. Liu, G.T. Zhao, and W.J.
Res., 2012, 415, p 853–858 Wang, Effects of Decarburization on the Wear Resistance and Damage
6. T. Furukawa, H. Okada, and Y. Maehara, Mechanism of Red Scale Mechanisms of Rail Steels Subject to Contact Fatigue, Wear, 2016,
Defect Formation in Si-Added Hot-Rolled Steel Sheets, ISIJ Int., 1994, 364–365, p 130–143
34, p 906–911 29. C. Man, C. Dong, H. Xue, K. Xiao, X. Li, and H. Qi, Influence of
7. X.L. Yu, Z.Y. Jiang, D.J. Yang, D.B. Wei, and Q. Yang, Precipitation Oxide Scales on the Corrosion Behaviors of B510L Hot-Rolled Steel
Behavior of Magnetite in Oxide Scale During Cooling of Microalloyed Strips, Int. J. Miner. Metall. Mater., 2016, 23, p 769–778
Low Carbon Steel, Adv. Mater. Res., 2012, 572, p 249–254 30. H. Yoshida and N. Sato, Grazing-Incidence X-Ray Diffraction Study of
8. A. Kobayashi, K. Seto, T. Urabe, K. Yamada, and K. Sato, Effect of Pentacene Thin Films with the Bulk Phase Structure, Appl. Phys. Lett.,
Scale Microstructure on Scale Adhesion of Low Carbon Sheet Steel, 2006, 89, p 1–3
Mater. Sci. Forum, 2006, 522–523, p 409–415 31. D.G. Neerinck and T.J. Vink, Depth Profiling of Thin ITO Films by
9. A.A. Mouayd, M.E. Orazem, E.M.M. Sutter, B. Tribollet, and A. Grazing Incidence X-Ray Diffraction, Thin Solid Films, 1996, 278, p
Koltsov, Contribution of Electrochemical Dissolution During Pickling 12–17
of Low Carbon Steel in Acidic Solutions, Corros. Sci., 2014, 82, p 32. S.J. Oh, D. Cook, and H. Townsend, Characterization of Rust
362–368 Commonly Formed on Steel, Hyperfine Interact., 1998, 112, p 59–65
10. C.M. Tamarelli, AHSS 101: The Evolving Use of Advance High- 33. F. Adar, Molecular Spectroscopy Workbench—Depth Resolution of the
Strength Steel for Automotive Applications, Steel Market Development Raman Microscope: Optical Limitations and Sample Characteristics,
Institute, Southfield, 2011 Spectroscopy, 2010, 25, p 16–23

Journal of Materials Engineering and Performance

34. S. Choudhary, A. Garg, and K. Mondal, Relation Between Open Circuit 38. C. Arya and P.R.W. Vassie, Influence of Cathodic-to-Anodic Area
Potential and Polarization Resistance with Rust and Corrosion Monitoring Ratio and Separation Distance on Galvanic Corrosion Currents of Steel
of Mild Steel, J. Mater. Eng. Perform., 2016, 25, p 2969–2976 in Concrete Containing Chlorides, Cem. Concr. Res., 1995, 25, p 989–
35. G.S. Frankel, Pitting Corrosion of Metals, J. Electrochem. Soc., 1998, 998
145, p 2186–2198 39. M. Fontana, Corrosion Engineering, 3rd ed., Tata McGraw-Hill, New
36. C. Dong, H. Xue, X. Li, H. Qi, and Y. Cheng, Electrochemical Delhi, 1987, p 67
Corrosion Behavior of Hot-Rolled Steel Under Oxide Scale in Chloride 40. Y. Kondo, H. Tanei, N. Suzuki, K. Ushioda, and M. Maeda, Blistering
Solution, Electrochim. Acta, 2009, 54, p 4223–4228 Behavior During Oxide Scale Formation on Steel Surface, ISIJ Int.,
37. F. Mansfeld, Area Relationship in Galvanic Corrosion, Corrosion, 2011, 51, p 1696–1702
1971, 27, p 436–442

Journal of Materials Engineering and Performance