Biaxial Stretching Behavior of a Copper-Alloyed Interstitial-Free

Steel by Bulge Test

R. RANA, S.B. SINGH, W. BLECK, and O.N. MOHANTY

Biaxial stretching behavior of a promising high-strength copper-alloyed interstitial-free (IF)

steel has been investigated under various processing conditions using bulge tests. Hill theory and
von Mises yield criterion have been used to analyze the results. It is revealed that copper-alloyed
IF steel in continuous-annealed (CA) condition exhibits the highest equivalent strain at fracture
and largest limiting dome height (LDH) among all the processing conditions. However, these
values are lower in copper-alloyed interstitial steel than in traditional interstitial-free–high-
strength (IF-HS) steels due to the presence of solute copper and copper precipitates in the
former.

DOI: 10.1007/s11661-010-0198-2
 The Minerals, Metals & Materials Society and ASM International 2010

I. INTRODUCTION (YS  460 MPa; UTS  560 MPa) IF steels, with pro-
vision for using the high formability of IF steels.[8–12]
INTERSTITIAL-FREE (IF) steels constitute a major The basis of this strengthening can be traced to the
class of high formable steels for the automobile and classical precipitation hardening by copper in a-iron.[13–19]
white goods sectors.[1–3] The specialty of these steels is It should be noted that copper is the only commercially
characterized by an extremely low level of interstitial feasible substitutional element to give rise to the age
content (typically, C and N £ 40 ppm each) achieved hardening effect in the Fe-based system. It is now
through the use of modern degassing techniques. These generally agreed that from the supersaturated solid
remaining interstitials are further removed from the solution, initially, copper precipitates out as bcc
matrix as various types of carbide, sulfide, carbosulfide, copper-rich clusters and copper precipitation in ferritic
and carbonitride by addition of small amounts of iron generally follows the following precipitation
stabilizing elements such as Ti and/or Nb.[4–6] Thus, sequence:[19]
these steels are made ‘‘interstitial free,’’ rendering them
nonaging and ensuring a high formability (strain hard- supersaturated solid solution ! bcc Cu ! 9R Cu
ening exponent, n ‡ 0.22; normal anisotropy factor, ! e CuðfccÞ
rm ‡ 1.8). In spite of their high formability, IF steels,
however, lack severely in strength properties (yield In the peak-aged condition, copper precipitates are in
strength, YS £ 220 MPa; ultimate tensile strength, the form of nanometric (2 to 3 nm), spherical, bcc
UTS £ 360 MPa), which restrict their vast applicability. copper-rich clusters.[19]
Even with the application of various metallurgical IF-Cu steels in peak-aged condition are far stronger
concepts such as solid solution hardening (by substitu- than traditional IF steels.[20] However, interestingly,
tional elements such as P, Mn, and Si), bake hardening, IF-Cu steels are soft and ductile and should have
or Hall–Petch hardening, a maximum UTS of adequate deep drawability and stretchability (high rm
~450 MPa has been achieved so far in IF steels.[7] and n values) in the as-continuous-annealed condition,
To the contrary, copper-alloyed IF (IF-Cu) steels are indicating easy press forming. It has been suggested that
discussed in recent times as potential high-strength the advantage of high formability of IF steels would be
used by press forming the IF-Cu steel in the as-annealed
condition when it is soft and formable.[8–12] Then, the
R. RANA, Institute Research Scholar, formerly Department of high strength would be achieved during a short-time age
Metallurgical and Materials Engineering, Indian Institute of Technol- hardening treatment of the easily formed annealed steel.
ogy Kharagpur, Kharagpur – 721 302, India, and DAAD Scholar,
Department of Ferrous Metallurgy, RWTH Aachen University, However, all these results were obtained from uniaxial
Aachen – 52072, Germany, is Researcher, Corus Research, Develop- tensile tests.[20,21] In sheet metal forming, the stress state
ment and Technology, 1970 CA IJmuiden, The Netherlands. Contact is biaxial and rarely the one of uniaxial tension.[22] The
e-mail: [email protected]; [email protected] bulge test provides extended stress-strain curve beyond
S.B. SINGH, Associate Professor, is with the Department of
Metallurgical and Materials Engineering, Indian Institute of Technol- the point of necking under biaxial stress conditions.[23]
ogy Kharagpur. O.N. MOHANTY, formerly Tata Research Profes- This information is very useful for the evaluation of
sor, Department of Metallurgical and Materials Engineering, Indian formability and selection of these materials for use in the
Institute of Technology Kharagpur, is retired. W. BLECK, Professor automotive sector. Therefore, bulge tests of IF-Cu steel
and Head, is with the Department of Ferrous Metallurgy, RWTH
Aachen University.
under various industrially important processing condi-
Manuscript submitted May 26, 2009. tions have been carried out, and their various features
Article published online March 25, 2010 have been discussed in this investigation. In addition, all

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 41A, JUNE 2010—1483

the results obtained for IF-Cu steel have been compared strips were cold rolled into 0.8-mm-thick sheets (80 pct
with those for traditional interstitial-free–high-strength thickness reduction) without any intermediate annealing
(IF-HS) steels for a better assessment of the new IF-Cu of the material. This is similar to the usual industrial
steel. rolling practice for IF grade steel. The cold-rolled sheets
were annealed inside salt bath according to the follow-
ing schedules and quenched in water: (a) batch anneal-
II. EXPERIMENTAL DETAILS ing (BA) at 1023 K (750 C) for 4 hours and (b)
continuous annealing (CA) at 1123 K (850 C) for
A. Material and Processing 1 minute. These annealing treatments resulted in fully
The IF-Cu steel was melted in a 50-kg laboratory recrystallized microstructure in the steel sheets.[20] The
induction furnace and cast in situ, both under the optical microstructures of the annealed IF-Cu steel are
protective atmosphere of Ar. Ti-stabilized IF steel scrap shown in Figure 1. To achieve the peak strength due to
was used as the charge material, and electrolytically pure copper precipitation, these fully recrystallized materials
copper was added as the major alloying element. The were aged at 823 K (550 C) in salt bath and quenched
composition of the investigated IF-Cu steel is listed in in water. The batch-annealed (BA) IF-Cu steel peak-
Table I. After casting, the ingots were reheated at aged (PA) in 2 hours, while the continuous-annealed
1473 K (1200 C) for 1 hour under protective atmo- (CA) IF-Cu steel took 1.5 hour.[20] A summary of these
sphere of Ar. These were then hot rolled from 30- to annealing and peak aging (PA) treatments is provided in
4-mm thickness with a finish rolling temperature of Table II.
1173 K (900 C) in a laboratory hot rolling mill and It can be seen from Table I that IF-Cu steel contains a
then air cooled to room-temperature. The hot-rolled somewhat higher amount of carbon (0.008 wt pct) than
traditional IF steels (<0.004 wt pct). However, it can be
processed into a completely stabilized IF steel, as can be
Table I. Chemical Composition of the IF-Cu and IF-HS shown with the help of the stabilization mechanism
Steels in Weight Percent applicable to such steels where C:S:N  1:1:1.[4] The
stabilization map[4] corresponding to the IF-Cu steel
Steels composition is shown in Figure 2. It reveals that the
Elements IF-Cu IF-HS-1 IF-HS-2
interstitial atoms are fully stabilized in IF-Cu steel
through the formation of Ti4C2S2 (H phase) and
C 0.008 0.004 0.003 epitaxial and free standing carbides of Ti ((TiC)e and
Mn 0.123 0.570 0.350
S 0.0065 0.005 0.008
P 0.011 0.031 0.042
Table II. Various Processing Conditions of the Steels
Si 0.011 0.010 0.011
Al 0.013 0.042 0.043
Steel Treatment Treatment Variables
Cu 1.21 0.015 0.028
Cr 0.031 0.025 0.017 IF-Cu BA 1023 K (750 C), 240 min
Ni 0.017 0.023 0.013 CA 1123 K (850 C), 1 min
Nb <0.005 0.0041 0.024 BA + PA BA + (823 K (550 C), 2 h)
Ti 0.062 0.0828 0.023 CA + PA CA + (823 K (550 C), 1.5 h)
N 0.0019 0.0017 0.0026 IF-HS-1
Fe balance balance balance IF-HS-2 industrial CA

Fig. 1—Microstructure of IF-Cu steel in (a) BA and (b) CA conditions.

1484—VOLUME 41A, JUNE 2010 METALLURGICAL AND MATERIALS TRANSACTIONS A

(TiC)f, respectively). Some amount of excess Ti is also machine with a strain rate of 3.33 9 104 s1 following
available in this steel after stabilization. However, a the DIN 10002 standard. The stress at 0.2 pct offset
somewhat higher amount of traditional precipitates in strain was considered as the yield strength. To determine
IF steels[4–6] would form in this steel due to its higher the anisotropy factors, samples were prepared in longi-
interstitial content as compared with traditional IF tudinal (0 deg), diagonal (45 deg), and perpendicular
steels. (90 deg) to the rolling direction. The normal anisotropy
Two traditional IF-HS steels (Table I), designated as factor (rm) and planar anisotropy factor (Dr) were
IF-HS-1 and IF-HS-2 steels, were procured in 0.8-mm determined using their well-known formula.[24] The
gage in the industrial CA condition and used for strain hardening exponent (n) was calculated using the
comparison with IF-Cu steel. The IF-HS steels were Hollomon equation.[25] Since the strain hardening expo-
strengthened by substitutional elements P and Mn. The nent value shows a variation with strain,[25] the values
major difference in composition of these two IF-HS for the entire plastic strain range from YS to UTS and
steels is that IF-HS-1 steel is stabilized by Ti only, also for the true strain range of 0.04 to 0.06 (n4–6) are
whereas IF-HS-2 steel is stabilized by Ti and Nb both. reported here. Moreover, the n value for a specific true
strain range is useful in order to compare a variety of
materials with different mechanical properties such as
B. Test Procedure
UTS and uniform elongation.
1. Quasi-static tensile test
Room-temperature quasi-static uniaxial tensile tests 2. Bulge test
on various steel samples with 25-mm gage length and
5-mm width were carried out in a 25-kN tensile testing a. Test setup. The as-received IF-HS-1 and IF-HS-2
steels and IF-Cu steel processed under various condi-
tions were tested for their stretch forming behavior at
room temperature. Bulge tests were conducted on
165-mm-diameter sheet samples with a 100-mm-diame-
ter hemispherical punch in a hydraulic bulging device,
illustrated schematically in Figure 3. The bulging device
was attached to an Erichsen 145/60 universal sheet
testing machine (Erichsen GmbH & Co. KG, Hemer,
Germany). Three specimens were tested for each pro-
cessing condition of the steels and the representative
result is presented. Circular grids of 5-mm diameter were
spray painted on the specimen surface to measure the
local distribution of deformation. The grids consist of
fluorescent color to create high contrast with the metal
surface. A bulged sample with the dot grids and two
laser lines before fracture is shown in Figure 4.
The Erichsen 145/60 universal sheet testing machine is
Fig. 2—IF-Cu steel composition superimposed on the stabilization capable of working in the temperature range of 293 to
map of S-containing, Ti-stabilized ultra-low-carbon steels proposed 673 K (20 to 400 C). The bulging device (Figure 3)
by Hua et al.[4] It shows that the interstitials in IF-Cu steel are fully consists of a blank holder and a stretching die, which
stabilized through the formation of Ti4C2S2 (H phase) and epitaxial
and free standing carbides of Ti ((TiC)e and (TiC)f). Ti# represents
can hold the sheet sample firmly without permitting any
the amount of solute Ti left after combining with N atoms to form material flow. A reservoir underneath the specimen
TiN. contains the hydraulic liquid on which pressure is

Fig. 3—Schematic diagram of the hydraulic bulging device attached to the Ericsen 145/60 universal sheet testing machine.

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 41A, JUNE 2010—1485

Fig. 4—Bulged sample with dot grids and two laser lines.

Fig. 6—Schematic representation of the measurement section used

for the bulge tests in Ericsen 145/60 universal sheet testing machine.

the area between the two laser axes. A digital video

camera records the x and y positions of the grids and
their lateral displacements throughout the test. The z
position of every point is obtained through the profile of
the specimen as measured by the laser ray. The
arrangement of the two laser light section modules and
the camera in the measuring device, which are located
above the testing tool, are shown schematically in
Figure 6. To avoid interference, the laser operates with
Fig. 5—Principle of the laser light section method used in the Ericsen different frequencies. At the point of material failure, a
145/60 universal sheet testing machine for strain measurements. sharp stream of oil (hydraulic liquid) may spurt out of
the cracked sheet hampering the measuring device.
Protection of the measuring device from this possible
exerted by a moving punch resulting in bulging of the damage is given by a detachable compound safety glass
specimen. The deformation is achieved by diminution of plate. In order to insert the sheet sample, the measuring
the sheet thickness. The machine setup is designed for a unit can be moved sideward by an electromechanical
maximum forming load of 600 kN and a blank holding positioning system, which also enables the exact align-
load of 400 kN, which are sufficient for sheet thicknesses ment of the laser ray over the center of the specimen
of up to 1.5 mm of a high strength steel. The punch load before the test.
and displacement are recorded continuously by strain
gages for calculation of the flow curve. The punch speed c. Data analysis. The evaluation of the deformation in
can be regulated from 1 to 800 mm/min and, for the the specimens is concentrated on the center of the bulge
present tests, was set to 30 mm/min to attain a quasi- in a circular area of 25 mm as compared to a bulge
static condition and to complete the test in a reasonable diameter of 100 mm. This is because the largest strains
testing duration of about 2 minutes. are located at the top of the bulge. The strain is averaged
over the measuring area; therefore, the influence of local
b. Measuring section. The shape analysis of the bulged necking is reduced. Practically, only four limiting points
specimen is based on the ‘‘laser light section’’ method and the center of the measuring area are determined in
(Figure 5). A laser ray is projected onto the measuring the machine (Figure 7), facilitating the calculation of the
surface and reflection from the surface is collected by a mean horizontal distance a and the mean vertical
detector. The line of reflection as well as the line of distance h for each principal direction. The bulge radius
incidence on the detector can be changed by changing c, the arc length l, and the logarithmic strain u in the
the position or the shape of the object. For the x-direction are determined by:[26]
comparatively simple geometry of a circular bulged
a2x þ h2x
sheet specimen, two perpendicular laser lines enable a cx ¼ ½1
characterization of the entire surface by interpolating 2hx

1486—VOLUME 41A, JUNE 2010 METALLURGICAL AND MATERIALS TRANSACTIONS A

 with
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
  
3 r0 þ r90
BT ¼ ½9
2 r0 r90 þ r0 þ r90

and
rVðHillÞ ¼ rx BT ½10

where r0 and r90 are the anisotropy factors in directions

parallel and perpendicular to the rolling direction,
respectively; BT is the anisotropy parameter in the Hill
theory; and the term ‘‘Hill’’ in parentheses indicates that
Fig. 7—Determination of the measuring area of the bulged the quantities are according to the Hill theory. The
specimen. values of r0 and r90 and, therefore, BT can be obtained
from simple uniaxial tensile tests.
The strain hardening rate analysis of the flow curve
ax was done by differentiating a fourth-order polynomial
lx ¼ cx arcsin ½2
cx function fitted to the flow curve determined according to
the Hill theory. The limiting dome height (LDH) of the
and bulge-tested samples was measured by a height gage.
lx The height of the base of each sample was measured and
ux ¼ ln ½3 deducted to obtain the absolute LDH value. The
a0
fracture surface of the selected failed bulge test samples
where the suffix x indicates that the values are for x- was studied with a scanning electron microscope (SEM)
direction and a0 represents the initial value of ax. Simi- operated at a voltage of 10 kV.
larly, all the values for the y-direction are also deter-
mined. From the constancy of volume, the strain in
the thickness direction uz is obtained as
III. RESULTS AND DISCUSSION
uz ¼ ux  uy ½4
A. Quasi-Static Mechanical Properties
Plane stress conditions prevail during the bulge test of
thin sheets, because the thickness of the sheet is much The room-temperature quasi-static mechanical prop-
smaller than the bulge radius (t/c  1, t being the erties of the investigated steels under various processing
sheet thickness). According to the theory of circular conditions (Table II) are given in Table III. As can be
sheets, the principal stress in the x-direction on the seen, the IF-Cu steel is softer than the IF-HS steels
bulge pole is[26] (IF-HS-1 and IF-HS-2) in as-annealed (batch and
continuous) conditions, as evidenced by their YS and
pcx d UTS values. IF-Cu steel in as-annealed conditions
rx ¼ ½5 possesses good elongation and normal anisotropy factor
2t
rm, and a higher strain hardening exponent (both n4–6
and in the y-direction, and n), than the IF-HS steels. This IF-Cu steel when
pcx d aged to peak strength is much stronger than the IF-HS
ry ¼ ð2  dÞ ½6 steels due to the presence of nanometric copper precip-
2t
itates.[28,29] These characteristics of quasi-static mechan-
where ical properties support the view that the IF-Cu steel can
cx be press formed in the as-annealed condition when it is
d¼ ½7 soft and formable and then it can be hardened efficiently
cy
during a postannealing aging treatment.[8–12] However, n
and p is the hydrostatic pressure in the oil bath. and rm values of IF-Cu steel here are inferior to those of
The preceding set of equations along with the traditional mild IF steels. A more detailed documenta-
experimentally measured data were used to calculate tion of these quasi-static mechanical properties can be
the principal stress (rx, ry) and principal strain (ux, uy, found elsewhere.[20]
uz) values. These values were then used to calculate the
effective stress and equivalent strain values according to
B. Stretch Forming Behavior
the von Mises yield criterion and Hill theory[27] to
construct the flow curves. The Hill theory accounts for 1. Flow curves
the anisotropic behavior of the materials and proposes a The flow curves obtained from bulge tests showed an
flow curve calculation that is independent of the state of interesting feature in the beginning. The effective stress
stress. The final expressions according to this theory are was very high due to the very high bulge radius in the
x-direction (cx) in the beginning of the test (Eqs. [1], [5],
uVðHillÞ ¼ juz jB1
T ½8 and [10]). In addition, the laser light section cannot

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 41A, JUNE 2010—1487

Table III. Room-Temperature Quasi-Static Mechanical Properties of the Steels Processed under Various Conditions; Tensile Tests
Were Done at a Strain Rate of 3.33 3 1024 s21

Steel Treatment Condition YS, MPa UTS, MPa eu, Pct et, Pct n4–6 n rm Dr
IF-Cu BA 164 317 21.1 22.7 0.29 0.23 1.2 0.2
CA 166 341 21.0 28.0 0.29 0.24 1.0 0.4
BA + PA 376 473 7.1 9.9 0.16 0.08 1.2 0.2
CA + PA 416 516 8.7 9.2 0.16 0.09 1.0 0.1
IF-HS-1 CA 215 357 24.0 38.9 0.22 0.17 1.8 0.2
IF-HS-2 CA 225 374 23.2 36.3 0.22 0.18 1.6 0.0
YS, UTS, eu, et, n4–6, n, rm, and Dr are yield strength, ultimate tensile strength, uniform elongation, total elongation, strain hardening exponent for
the true strain range 0.04 to 0.06, strain hardening exponent for the entire range of strain from YS to UTS, normal anisotropy factor, and planar
anisotropy factor, respectively. The values of YS, UTS, eu, et, and n4–6 are for the longitudinal specimen. The abbreviations BA, CA, BA + PA, and
CA + PA are defined in Table II.

Fig. 8—Flow curves, following Hill theory, obtained from bulge Fig. 9—Quasi-static tensile flow curves of the steels under various
tests of the steels under different processing conditions of BA, CA, processing conditions of BA, CA, BA + PA, and CA + PA. The
batch annealing and peak aging (BA + PA), and continuous anneal- flow curves are for longitudinal specimens.
ing and peak aging (CA + PA). uVðHillÞ and rVðHillÞ represent the
equivalent strain and effective stress, respectively, calculated from the
Hill theory.
obtained from the quasi-static tensile test on the
longitudinal specimen is compared with the flow curve
measure the bulge radius properly in the beginning of obtained from the bulge tests using the von Mises yield
bulging.[23] Therefore, the initial values of the effective criterion and Hill theory in Figure 10. It can be
stress in the beginning of the test are not reliable and are observed that the flow curve derived from von Mises
thus discarded. In view of this, the flow curves presented yield criterion is higher than that derived from Hill
here for interpretation start at an equivalent strain of theory due to the anisotropy parameter BT considered in
0.02. the latter. Therefore, for anisotropic materials such as
The flow curves of the samples according to Hill the present ones (Table III), the Hill theory is more
theory are given in Figure 8. The anisotropy parameters useful than the von Mises yield criterion. Further, in
BT calculated for the various samples using Eq. [9] are Figure 10, the flow curve from the bulge test calculated
listed in Table IV. These BT values were used to according to the Hill theory matches closely with the
calculate
 the equivalent
 strain uVðHillÞ and effective quasi-static tensile flow curve, enabling a direct com-
stress rVðHillÞ data according to the Hill theory parison between the two testing methods. This unique-
(Eqs. [8] and [10]). The trend of flow curves for various ness of the bulge test allows an extension of the
materials in Figure 8 matches closely with the trend of measurable data range beyond that of the tensile test.[23]
flow curves of these materials obtained by quasi-static Material data beyond the point of necking are useful for
tensile tests (Figure 9). sheet metal forming and can be obtained from the bulge
The flow curves obtained from uniaxial tensile tests test, but not from the tensile test. For all the materials,
and bulge tests showed some unique characteristics. As a the strain value at fracture from the quasi-static tensile
representative case, the flow curve of CA IF-Cu steel test is at least 50 pct less than that obtained from the

Table IV. Anisotropy Parameter of Hill Theory (BT) for the Bulge Test Samples

IF-Cu IF-Cu IF-Cu IF-Cu IF-HS-1 IF-HS-2

Steel Sample (BA) (CA) (BA + PA) (CA + PA) (CA) (CA)
BT 0.9508 0.9630 0.9516 0.9927 0.9094 0.9287

1488—VOLUME 41A, JUNE 2010 METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 10—Representative comparison of flow curves, following von
Mises yield criterion and Hill theory, obtained from bulge tests with
the flow curve from quasi-static uniaxial tensile test for IF-Cu steel
in CA condition. Fig. 11—Part of the Fe-Cu phase diagram for the IF-Cu steel. The
phase diagram was calculated using ThermoCalc database TCFE4.
Table V. Comparison of the Equivalent Strain at Fracture
(uV(Hill)f) and the Effective Stress at an Equivalent Strain
of 0.3 (rV(Hill)0.3) of the Bulge Test Samples; the Quantities
are According to the Hill Theory

Steel Sample uV(Hill)f rV(Hill)0.3, MPa

IF-Cu (BA) 0.42 414
IF-Cu (CA) 0.49 447
IF-Cu (BA + PA) 0.32 638
IF-Cu (CA + PA) 0.34 657
IF-HS-1 (CA) 0.88 483
IF-HS-2 (CA) 0.93 512

bulge test with the Hill calculation model (Figures 8

through 10).
Following the Hill theory, the equivalent strain at Fig. 12—Strain hardening rate curves obtained from bulge tests
fracture and the effective stress at a particular equivalent of the steels under different processing conditions of BA, CA,
BA + PA, and CA + PA. Strain hardening rates were calculated
strain of the steels under various processing conditions from the flow curves obtained by applying the Hill theory.
are listed in Table V. The effective stress values of the
steel samples at an equivalent strain of 0.3 (rV(Hill)0.3) and
the equivalent strain values at fracture (uV(Hill)f) both in different processing conditions. The part of the Fe-Cu
agree with the trend obtained for these steel samples phase diagram for the present IF-Cu steel composition
from quasi-static tensile tests (Table III). The highest shown in Figure 11 reveals that the entire 1.21 wt pct
effective stress and the lowest equivalent strain at copper is in solid solution in both annealing conditions
fracture of the annealed and peak-aged (BA + PA and at 1023 and 1123 K (750 and 850 C). This solute
CA + PA) IF-Cu steel are due to the presence of copper copper is believed to reduce the equivalent strain at
precipitates in these processing conditions.[20] The lowest fracture of IF-Cu steel in batch and CA conditions, as
effective stress of IF-Cu steel in annealed conditions compared to IF-HS steels. A higher amount of usual
(BA and CA) is due to the absence of any strong solid precipitates of IF steels[4] in the present IF-Cu steel due
solution strengthening element (such as P and Mn in to its higher C content (Table I) might also be partly
IF-HS steels) or fine copper precipitates (such as in peak- responsible for its inferior values of equivalent strain at
aged IF-Cu steel). The CA IF-Cu steel shows the highest fracture.[30]
equivalent strain at fracture among all the processing
conditions. Further, the overall lower equivalent strain 2. Strain hardening rate
at fracture of IF-Cu steel in annealed conditions as Strain hardening rates obtained from the flow curves
compared with IF-HS steels can be attributed to the calculated using the Hill theory (Figure 8) are shown in
presence of solute copper, as explained subsequently. Figure 12. The strain hardening curves from the bulge
ThermoCalc* calculations using the TCFE4 database tests can be divided into two zones (stages III and IV)
with differing behaviors.[31] Stage III corresponds to the
strong decline of the strain hardening rate in the
*ThermoCalc is a trademark of Thermo-Calc Software AB,
Stockholm, Sweden.
beginning, whereas the zone of almost a constant strain
hardening rate can be considered as stage IV. The initial
strain hardening (ISH) rate (at uV(Hill)  0.02) of IF-Cu
were used to predict the amount of copper in IF-Cu steel steel under all the processing conditions is distinctly

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 41A, JUNE 2010—1489

 Table VI. ISH Rate of the Bulge Test Samples Obtained from the Strain Hardening Rate Curves Determined by Applying the
Hill Theory

IF-Cu IF-Cu IF-Cu IF-Cu IF-HS-1 IF-HS-2

Steel Sample (BA) (CA) (BA + PA) (CA + PA) (CA) (CA)
ISH rate, MPa 1396 1461 1621 1772 1187 1074

Fig. 13—LDH of the steels under different processing conditions of Fig. 14—Correlation between LDH and strain hardening exponent
BA, CA, BA + PA, and CA + PA. (n) of steels under various processing conditions of BA, CA, BA +
PA, and CA + PA.

higher than that of IF-HS-1 and IF-HS-2 steels.

However, the strain hardening rate of IF-Cu steel
samples drops more rapidly than that in IF-HS steels,
showing a crossover of the strain hardening curves of
the two types of steels at an equivalent strain of around
0.1. ISH rates of the samples are listed in Table VI, since
these are difficult to realize in Figure 12. The samples
can be arranged in order of decreasing ISH rate as
IF-Cu steel (CA + PA) > IF-Cu steel (BA + PA) >
IF-Cu steel (CA) > IF-Cu steel (BA) > IF-HS-1 steel
(CA) > IF-HS-2 steel (CA). The higher ISH rate of the
annealed and peak-aged (BA + PA and CA + PA)
IF-Cu steel than that of the annealed (BA and CA)
IF-Cu steel matches with the results reported from the Fig. 15—Strain paths experienced during bulge tests by the steels
uniaxial tensile test of Fe-Cu binary alloy.[19,32] It is under various processing conditions of BA, CA, BA + PA, and
CA + PA.
postulated that the fine bcc copper precipitates present
in annealed and peak-aged IF-Cu steel samples undergo
a stress- or strain-induced martensite-type transforma- because of the presence of fine copper precipitates.[19]
tion. This transformation is thought to be responsible In agreement with the equivalent strain at fracture
for the higher ISH rate in annealed and peak-aged (Figure 8 and Table V), the LDH is lower in the BA
IF-Cu steel. However, a detailed study is necessary to condition than that in the CA condition of the IF-Cu
confirm this hypothesis. steel.
Since the thickness of the steel samples was the same,
3. Limiting dome height the LDH values of these materials should depend on
The average LDH, which is an index of stretch their strain hardening behavior through the strain
formability, of various samples is shown in Figure 13. hardening exponent n.[33] LDH values of the differently
The CA IF-HS steels showed the largest LDH, while the processed steels are correlated with their n values in
IF-Cu steels in the annealed and peak-aged (BA + PA Figure 14. It can be observed that, for differently
and CA + PA) conditions showed the smallest. The processed IF-Cu steels, LDH values increase with n
LDHs of the IF-Cu steel in annealed (BA and CA) values. Thus, the LDH of annealed (BA and CA) IF-Cu
conditions are comparable but less than those of the steels is higher than that of annealed and peak-aged
IF-HS steels. The presence of copper either in solution (BA + PA and CA + PA) IF-Cu steels. However, LDH
or as precipitates appears to have a negative influence on values of CA IF-HS steels are higher, though their n
LDH. Thus, the LDH is lower for the IF-Cu steel than values are lower, than continuous- and batch-annealed
for the IF-HS steels. As far as IF-Cu steel is concerned, IF-Cu steels. This suggests that the composition of the
annealed and peak-aged samples show a lower LDH steels plays a role. It should be noted that investigated

1490—VOLUME 41A, JUNE 2010 METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 16—SEM fractographs of the failed samples after the bulge test: (a) CA IF-Cu (b) CA + PA IF-Cu, and (c) CA IF-HS-1 steel.

IF-HS steels contain P and Mn as major alloying equivalent strain at fracture and lower LDH value
elements (Table I). Thus, the LDH value in the present of this sample in the bulge test (Table V and Figures 8
case can only be correlated with the n value among and 13).
different processing conditions of the same steel.
Further, local inhomogeneities in microstructure might
also have some influence on the LDH value. This is
discussed in Section 5. IV. CONCLUSIONS
It has been shown that the bulge test can be
4. Strain paths successfully applied to IF-Cu steel to obtain mechanical
The strain paths during the bulge test of the various properties beyond the range of uniaxial tensile tests. The
samples, shown in Figure 15, are not consistent with the results of this test give useful insight into the behavior of
anisotropy results from tensile tests (Table III). Irre- the material during forming operations, where the state
spective of the values of normal and planar anisotropy of stress is far from uniaxial tension. For this steel, the
factors (rm and Dr, respectively) obtained from the flow curves obtained from bulge tests using the Hill
tensile tests, the strain paths of all the samples lie near theory are directly comparable with those obtained from
the equi-biaxial strain path (longitudinal strain = uniaxial tensile tests on longitudinal specimens. The
transverse strain) of the plot, implying that the samples following conclusions can be drawn from the present
are deformed to the same extent in rolling and trans- work.
verse directions. It is true even for the materials with
high rm value (IF-HS-1 and IF-HS-2 steels). Since the 1. CA IF-Cu steel performs best among the variously
deformation is constrained in the bulge test, it can be processed IF-Cu steels under biaxial stretching con-
expected that the anisotropy results will not be consis- ditions in bulge tests. It shows the highest equiva-
tent with tensile tests. lent strain at fracture and the largest LDH as
compared to other processing conditions. However,
5. Fracture behavior due to the presence of solute copper and copper
The SEM fractographs of CA and CA + PA IF-Cu precipitates, IF-Cu steel under all processing condi-
steel samples, are compared with that of CA IF-HS steel tions shows lower equivalent strain at fracture and
(IF-HS-1 steel) in Figure 16. It appears that all the LDH than IF-HS steels.
samples failed mostly in a ductile mode, evidenced by 2. The LDH of IF-Cu steels under different processing
the presence of dimples almost in entire part of the conditions increases linearly with the strain harden-
fracture surface. However, the size of the dimples is ing exponent. However, this correlation does not
different according to the heat treatment and type of apply when comparing IF-Cu and IF-HS steels. It
steel. The dimples in CA IF-HS-1 steel (Figure 16(c)) is suggested that the LDH depends on composition
are clear and larger than that in CA IF-Cu steel as well.
(Figure 16(a)). Further, a small part of the fracture 3. The annealed and peak-aged IF-Cu steels show a
surface of CA IF-Cu steel shows flat facets, which are higher ISH rate in the bulge test than the annealed
not present in the fracture surface of CA IF-HS-1 steel. IF-Cu and IF-HS steels. This is believed to be due
This can be due to the presence of local inhomegeneities to a stress/strain induced martensitic transformation
present in IF-Cu steel. These observations correlate with of the copper precipitates.
the effect of copper in reducing the equivalent strain at 4. The fracture surface of failed IF-Cu steel under
fracture in the bulge test (Table V and Figure 8) and the biaxial stretching is mostly ductile type, with a
lower LDH values of IF-Cu steel (Figure 13). The minor presence of flat facets possibly due to local
dimple size in the fracture surface of CA + PA IF-Cu inhomogeneities. Further, the dimple size of the
steel (Figure 16(b)) is even smaller than that in CA fractographs is smaller than that in IF-HS steels,
IF-Cu steel due to the presence of peak-aged copper agreeing with the obtained mechanical properties
precipitates.[19] This fact is also reflected in the lower from biaxial stretching.

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 41A, JUNE 2010—1491

 ACKNOWLEDGMENTS 15. S.R. Goodman, S.S. Brenner, and J.R. Low, Jr.: Metall. Trans.,
1973, vol. 4, pp. 2363–78.
The authors thank Thyssen-Krupp Steel (Duisburg, 16. P.J. Othen, M.L. Jenkins, G.D.W. Smith, and W.J. Phythian: Phil.
Germany) for providing the IF steel raw material for Mag. Lett., 1991, vol. 64, pp. 383–91.
17. F. Maury, N. Lorenzelli, M.H. Mathon, C.H. De Novion, and
steelmaking. One of the authors (RR) expresses his P. Lagarde: J. Phys. Condens. Matter, 1994, vol. 6, pp. 569–88.
gratitude to the German Academic Exchange Service 18. S. Pizzini, K.J. Roberts, W.J. Phythian, C.A. English, and G.N.
(DAAD) for funding his research stay in the Depart- Greaves: Phil. Mag. Lett., 1990, vol. 61, pp. 223–29.
ment of Ferrous Metallurgy, RWTH Aachen Univer- 19. A. Deschamps, M. Militzer, and W.J. Poole: ISIJ Int., 2001,
vol. 42, pp. 196–205.
sity, during the course of this investigation. 20. R. Rana: Ph.D. Thesis, IIT, Kharagpur, India, 2008.
21. R. Rana, S.B. Singh, W. Bleck, and O.N. Mohanty: Metall. Mater.
Trans. A, 2009, vol. 40A, pp. 856–66.
22. W.T. Hosford and R.M. Caddel: Metal Forming Mechanics and
REFERENCES Metallurgy, Prentice-Hall, Englewood Cliffs, NJ, 1983.
1. S. Hoile: Mater. Sci. Technol., 2000, vol. 16, pp. 1079–93. 23. W. Bleck and M. Blumbach: Steel Res. Int., 2005, vol. 76, pp. 125–30.
2. K. Tsunoyama: Phys. Status Solidi (a), 1998, vol. 167, pp. 427–33. 24. W.T. Lankford, S.C. Snyder, and J.A. Bauscher: Trans. ASM,
3. H. Takechi: ISIJ Int., 1994, vol. 34, pp. 1–8. 1950, vol. 42, pp. 1197–1225.
4. M. Hua, C.I. Garcia, and A.J. DeArdo: Metall. Mater. Trans. A, 25. G.E. Dieter: Mechanical Metallurgy, McGraw-Hill Book
1997, vol. 28A, pp. 1769–80. Company, London, 1988, p. 287.
5. X. Yang, D. Vanderschueren, J. Dilewijns, and C. Standaert: ISIJ 26. F. Gologranc: Industrie Anzeiger, 1968, vol. 90, pp. 775–79.
Int., 1996, vol. 36, pp. 1286–94. 27. F. Gologranc: Blech Rohre Profile, 1977, vol. 4, pp. 116–21.
6. B.C. De Cooman and A. De Vyt: Proc. Int. Forum for the Prop- 28. G.M. Worral, J.T. Buswell, C.A. English, M.G. Hetherington, and
erties and Application of IF Steels, H. Takechi, ed., ISIJ, Tokyo, G.D.W. Smith: J. Nucl. Mater., 1987, vol. 148, pp. 107–14.
2003, pp. 249–58. 29. W.J. Phythian, S. Dumbill, P. Brown, and R. Sinclair: Proc. 6th
7. T. Senuma: ISIJ Int., 2001, vol. 41, pp. 520–32. Int. Symp. on Environmental Degradation in Nuclear Power Sys-
8. R. Rana, W. Bleck, S.B. Singh, and O.N. Mohanty: Mater. Lett., tems—Water Reactors, R.E. Gold and E.P. Simonen, eds., TMS,
2007, vol. 61, pp. 2919–22. Warrendale, PA, 1993, pp. 729–37.
9. R. Rana, W. Bleck, S.B. Singh, and O.N. Mohanty: Steel Res. Int., 30. I.L. Dillamore, W.B. Hutchinson, and P.L. Morris: Proc. Effect of
2007, vol. 78, pp. 612–21. Second Phase Particles on the Mechanical Properties of Steel,
10. R. Rana, W. Bleck, S.B. Singh, and O.N. Mohanty: Steel Res. Int., Corporate Laboratories of the British Steel Corporation and the
2007, vol. 78, pp. 622–30. Iron and Steel Institute, Scarborough, United Kingdom, 1971,
11. R. Rana, W. Bleck, S.B. Singh, and O.N. Mohanty: Steel Res. Int., pp. 190–95.
2007, vol. 78, pp. 631–37. 31. X.F. Fang and W. Dahl: Mater. Sci. Eng. A, 1995, vol. A203,
12. K. Kishida: Nippon Steel Tech. Rep., 1995, No. 64, pp. 33–35. pp. 36–45.
13. N. Maruyama, M. Sugiyama, T. Hara, and H. Tamehiro: Mater. 32. A. Deschamps, S. Esmaeili, W.J. Poole, and M. Militzer: J. Phys.
Trans. JIM, 1999, vol. 40, pp. 268–77. IV, 2000, vol. 10, pp. 151–56.
14. E. Hornbogen and R.C. Glenn: Trans. TMS-AIME, 1960, 33. D.H. Kang, D.-W. Kim, S. Kim, G.T. Bae, K.H. Kim, and
vol. 218, pp. 1064–70. N.J. Kim: Scripta Mater., 2009, vol. 61, pp. 768–71.

1492—VOLUME 41A, JUNE 2010 METALLURGICAL AND MATERIALS TRANSACTIONS A